","summary":"The article discusses the

element in HTML, highlighting its features and best practices for usage, including ARIA labeling and styling options.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-23","topics":["html-elements","web-accessibility","aria-attributes"],"diataxisType":"reference","path":"/kb/articles/on-the-dl-5caece01","sourceUrl":"https://simonwillison.net/2026/May/23/on-the-dl/#atom-everything"},{"id":"968a3393-1a27-4f8c-a9c9-b8b2de31c83a","title":"[AINews] All Model Labs are now Agent Labs","summary":"The article discusses the shift from Model Labs to Agent Labs in AI development, highlighting the integration of models with harnesses and workflows as the new product standard.","author":"Swyx","authorId":"swyx","published":"2026-05-23","topics":["ai-agents","model-development","product-design","ai-economics"],"diataxisType":"explanation","path":"/kb/articles/ainews-all-model-labs-are-now-agent-labs-968a3393","sourceUrl":"https://www.latent.space/p/ainews-all-model-labs-are-now-agent"},{"id":"b3aa7c82-138f-4339-a15c-e031489e4a7e","title":"🧠 Community Wisdom: Best AI cold outreach app, advice for college students interested in PM, the state of AI in design, the relevance of QA in 2026, and more","summary":"This week's Community Wisdom discusses the best AI cold outreach apps, advice for college students interested in product management, and the future of AI in design and QA.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-23","topics":["ai-cold-outreach","product-management","ai-in-design","quality-assurance"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-best-ai-cold-outreach-app-advice-for-colle-b3aa7c82","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-best-ai-cold-outreach"},{"id":"c3606a76-1c5d-400c-8c53-6eae42f13a28","title":"The memory shortage is causing a repricing of consumer electronics","summary":"The memory shortage is leading to higher prices for consumer electronics, particularly smartphones, due to increased demand for high-bandwidth memory (HBM) driven by AI growth.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-22","topics":["memory-shortage","consumer-electronics","ai-growth","smartphone-pricing"],"diataxisType":"explanation","path":"/kb/articles/the-memory-shortage-is-causing-a-repricing-of-consumer-elect-c3606a76","sourceUrl":"https://simonwillison.net/2026/May/22/memory-shortage/#atom-everything"},{"id":"2b654654-c581-4b5d-8728-23007250e7f3","title":"FTC to Require Cox Media Group, Two Other Firms to Pay Nearly $1 Million to Settle Charges They Deceived Customers About “Active Listening” AI-Powered Marketing Service","summary":"The FTC has mandated Cox Media Group and two other firms to pay nearly $1 million for misleading customers about their 'Active Listening' AI marketing service.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-22","topics":["ftc","ai-marketing","consumer-deception","data-privacy"],"diataxisType":"explanation","path":"/kb/articles/ftc-to-require-cox-media-group-two-other-firms-to-pay-nearly-2b654654","sourceUrl":"https://simonwillison.net/2026/May/22/ftc-active-listening/#atom-everything"},{"id":"81f3a4f9-dd51-4214-b40c-916068379a60","title":"Goodbye, Lenny's Reads!","summary":"Lenny Rachitsky announces the end of his audio newsletter experiment, reflecting on the journey and thanking listeners for their support.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-22","topics":["audio-newsletters","content-creation","technology"],"diataxisType":"explanation","path":"/kb/articles/goodbye-lennys-reads-81f3a4f9","sourceUrl":"https://www.lennysnewsletter.com/p/goodbye-lennys-reads"},{"id":"0e161fbb-05c4-4e71-bfd8-ce52c8e10b96","title":"[AINews] New AI Infra unicorns: Exa, Modal, TurboPuffer","summary":"The article discusses the latest milestones of AI infrastructure companies Exa, Modal, and TurboPuffer, along with significant advancements in AI research and tools.","author":"Swyx","authorId":"swyx","published":"2026-05-22","topics":["ai-infrastructure","ai-research","developer-tooling","machine-learning"],"diataxisType":"reference","path":"/kb/articles/ainews-new-ai-infra-unicorns-exa-modal-turbopuffer-0e161fbb","sourceUrl":"https://www.latent.space/p/ainews-new-ai-infra-unicorns-exa"},{"id":"e8237e03-34b7-40ff-a57d-dc9167105677","title":"datasette-agent-sprites 0.1a0","summary":"The article introduces datasette-agent-sprites 0.1a0, a tool designed to enhance the capabilities of Datasette with agent sprites for improved data interaction.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-21","topics":["datasette","data-visualization","software-development"],"diataxisType":"reference","path":"/kb/articles/datasette-agent-sprites-01a0-e8237e03","sourceUrl":"https://simonwillison.net/2026/May/21/datasette-agent-sprites/#atom-everything"},{"id":"0b6bf39a-4cdf-4bc9-acde-70bcab3be40d","title":"Datasette Agent","summary":"Simon Willison announces Datasette Agent, an AI assistant for Datasette that provides a conversational interface for querying data and supports extensible plugins.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-21","topics":["ai-assistants","datasette","llm","data-visualization"],"diataxisType":"explanation","path":"/kb/articles/datasette-agent-0b6bf39a","sourceUrl":"https://simonwillison.net/2026/May/21/datasette-agent/#atom-everything"},{"id":"f19f062d-9d70-465c-890f-39a6f5b58ce1","title":"Release: datasette-agent 0.1a3","summary":"The release of datasette-agent 0.1a3 introduces improved SQL query handling and user interface enhancements for the Datasette AI assistant.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-21","topics":["datasette","ai-assistants","sql-queries","software-release"],"diataxisType":"reference","path":"/kb/articles/release-datasette-agent-01a3-f19f062d","sourceUrl":"https://simonwillison.net/2026/May/21/datasette-agent-2/#atom-everything"},{"id":"8fc28d1f-f379-49f6-94f0-ecf582668b7f","title":"Release: datasette-agent-charts 0.1a2","summary":"Simon Willison announces the release of datasette-agent-charts 0.1a2, highlighting its features and inviting sponsorship for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-21","topics":["datasette","llm-developments","open-source"],"diataxisType":"reference","path":"/kb/articles/release-datasette-agent-charts-01a2-8fc28d1f","sourceUrl":"https://simonwillison.net/2026/May/21/datasette-agent-charts/#atom-everything"},{"id":"539304b7-f9fb-413b-a947-a64c7e0f7a23","title":"[AINews] OpenAI GPT-next disproves 80 year old Erdős planar unit distance problem for under $1000","summary":"OpenAI's GPT-next model has disproved the 80-year-old Erdős planar unit distance problem, showcasing advancements in AI's reasoning capabilities.","author":"Swyx","authorId":"swyx","published":"2026-05-21","topics":["ai-research","mathematics","openai","gpt-models"],"diataxisType":"explanation","path":"/kb/articles/ainews-openai-gpt-next-disproves-80-year-old-erds-planar-uni-539304b7","sourceUrl":"https://www.latent.space/p/ainews-openai-gpt-next-disproves"},{"id":"6469fb3c-c1a1-4b80-b88b-cfd57709cf96","title":"Giving Agents Computers — Ivan Burazin, Daytona","summary":"Ivan Burazin discusses the evolution of AI infrastructure at Daytona, emphasizing the need for composable computers for AI agents rather than traditional development environments.","author":"Swyx","authorId":"swyx","published":"2026-05-21","topics":["ai-infrastructure","composable-computers","ai-agents","cloud-computing"],"diataxisType":"explanation","path":"/kb/articles/giving-agents-computers-ivan-burazin-daytona-6469fb3c","sourceUrl":"https://www.latent.space/p/daytona"},{"id":"30e8af9f-eddf-4b1c-995f-57cdf3ab820e","title":"What launched at Google I/O 2026 (30-minute day 1 recap)","summary":"The article recaps the major announcements from Google I/O 2026, including updates on Gemini 3.5, Anti-Gravity 2.0, and new creative tools from Google.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-20","topics":["google-io","ai-tools","gemini","productivity","video-generation"],"diataxisType":"reference","path":"/kb/articles/what-launched-at-google-io-2026-30-minute-day-1-recap-30e8af9f","sourceUrl":"https://www.lennysnewsletter.com/p/what-launched-at-google-io-2026-30"},{"id":"b90eddcf-bde5-4d44-8661-2f722e8a2df8","title":"How fast is 10 tokens per second really?","summary":"Simon Willison discusses the implications of processing speed in generative AI, particularly focusing on the significance of 10 tokens per second.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-20","topics":["generative-ai","llms","ai"],"diataxisType":"explanation","path":"/kb/articles/how-fast-is-10-tokens-per-second-really-b90eddcf","sourceUrl":"https://simonwillison.net/2026/May/20/tokens-per-second/#atom-everything"},{"id":"26492828-077b-4786-a313-1b6f7731a449","title":"A quote from SpaceX S-1","summary":"SpaceX has entered into a $1.25 billion monthly Cloud Services Agreement with Anthropic for compute capacity, supporting its AI applications.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-20","topics":["cloud-services","ai-research","business-agreements"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-spacex-s-1-26492828","sourceUrl":"https://simonwillison.net/2026/May/20/spacex-s1/#atom-everything"},{"id":"d6d5aa87-1468-4953-8fc0-8b6e88bcd44f","title":"Google I/O, Gemini Spark, Antigravity","summary":"The article discusses Google's announcements at I/O, focusing on Gemini Spark, an AI agent, and its security features, while expressing caution about untested previews.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-20","topics":["google-io","gemini-spark","ai-agents","security","cloud-computing"],"diataxisType":"explanation","path":"/kb/articles/google-io-gemini-spark-antigravity-d6d5aa87","sourceUrl":"https://simonwillison.net/2026/May/20/google-io/#atom-everything"},{"id":"f3d56117-58b8-4c26-9de5-d23a997a1472","title":"Release: datasette-agent-charts 0.1a1","summary":"The release of datasette-agent-charts 0.1a1 introduces enhanced color options for charts, interactive tooltips, and improved permission checks for SQL queries.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-20","topics":["data-visualization","charts","datasette"],"diataxisType":"reference","path":"/kb/articles/release-datasette-agent-charts-01a1-f3d56117","sourceUrl":"https://simonwillison.net/2026/May/20/datasette-agent-charts/#atom-everything"},{"id":"0f13a16d-7e49-4ca8-acee-ec3b59f2ea4d","title":"[AINews] Google I/O 2026: Gemini 3.5 Flash, Omni (NanoBanana for Video), Spark (background agents), and Antigravity 2.0","summary":"Google I/O 2026 showcased Gemini 3.5 Flash and Omni, emphasizing advancements in AI capabilities and multimodal generation, alongside the introduction of Antigravity 2.0.","author":"Swyx","authorId":"swyx","published":"2026-05-20","topics":["ai-agents","multimodal-ai","google-io","gemini-3-5","technology"],"diataxisType":"explanation","path":"/kb/articles/ainews-google-io-2026-gemini-35-flash-omni-nanobanana-for-vi-0f13a16d","sourceUrl":"https://www.latent.space/p/ainews-google-io-2026-gemini-35-flash"},{"id":"5fb4f0d9-db2a-499a-b076-a92964d770d0","title":"Railway: The Agent-Native Cloud — Jake Cooper","summary":"Jake Cooper discusses Railway's evolution from a startup to a major player in AI infrastructure, emphasizing the need for a new cloud paradigm for agent-native applications.","author":"Swyx","authorId":"swyx","published":"2026-05-20","topics":["ai-infrastructure","cloud-computing","software-deployment","agent-native"],"diataxisType":"explanation","path":"/kb/articles/railway-the-agent-native-cloud-jake-cooper-5fb4f0d9","sourceUrl":"https://www.latent.space/p/railway"},{"id":"4589c841-708e-41f4-bb3b-d93d74469060","title":"Release: llm-gemini 0.32","summary":"Simon Willison announces the release of llm-gemini 0.32, highlighting its features and inviting readers to support his work.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["llm-gemini","monthly-briefing","sponsorship"],"diataxisType":"reference","path":"/kb/articles/release-llm-gemini-032-4589c841","sourceUrl":"https://simonwillison.net/2026/May/19/llm-gemini-2/#atom-everything"},{"id":"5e81eef1-b98f-46f3-81ff-23245edc8c2a","title":"Gemini 3.5 Flash: more expensive, but Google plan to use it for everything","summary":"Google launched Gemini 3.5 Flash at I/O 2026, increasing prices significantly while integrating it into various products and platforms.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["gemini-3-5-flash","google-io","ai-pricing","api-development"],"diataxisType":"explanation","path":"/kb/articles/gemini-35-flash-more-expensive-but-google-plan-to-use-it-for-5e81eef1","sourceUrl":"https://simonwillison.net/2026/May/19/gemini-35-flash/#atom-everything"},{"id":"e412f7cb-54c1-497f-9ad0-6217cfe7f9b5","title":"[AINews] How to land a job at a frontier lab (on Pretraining)","summary":"The article discusses job preparation for frontier labs, focusing on pretraining and the skills needed to tune LLMs at the kernel level.","author":"Swyx","authorId":"swyx","published":"2026-05-19","topics":["job-preparation","pretraining","llm-tuning","ai-research"],"diataxisType":"how-to","path":"/kb/articles/ainews-how-to-land-a-job-at-a-frontier-lab-on-pretraining-e412f7cb","sourceUrl":"https://www.latent.space/p/ainews-how-to-land-a-job-at-a-frontier"},{"id":"4cf009ad-0085-40ad-91ea-0a0a9d5124d1","title":"Release: datasette-llm-accountant 0.1a4","summary":"Simon Willison announces the release of datasette-llm-accountant 0.1a4, highlighting its features and inviting sponsorship for curated LLM updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["datasette","llm","release-notes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-accountant-01a4-4cf009ad","sourceUrl":"https://simonwillison.net/2026/May/19/datasette-llm-accountant/#atom-everything"},{"id":"82ee5038-d533-477e-a274-08ea729d847d","title":"Release: datasette-llm 0.1a8","summary":"Simon Willison announces the release of datasette-llm 0.1a8, highlighting its features and inviting sponsorship for a curated email digest.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["datasette","llm","release-notes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a8-82ee5038","sourceUrl":"https://simonwillison.net/2026/May/19/datasette-llm/#atom-everything"},{"id":"b781057e-8f3f-4fe6-98a1-5dee475969f1","title":"The last six months in LLMs in five minutes","summary":"The article summarizes key developments in LLMs over the past six months, highlighting advancements in coding agents and the emergence of personal AI assistants.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["llm-developments","coding-agents","ai-assistants","model-comparisons"],"diataxisType":"explanation","path":"/kb/articles/the-last-six-months-in-llms-in-five-minutes-b781057e","sourceUrl":"https://simonwillison.net/2026/May/19/5-minute-llms/#atom-everything"},{"id":"7de634ec-9a0e-4b6b-aa3f-86221c762316","title":"The Lenny and Friends Summit is back!","summary":"The Lenny and Friends Summit returns on September 10 in San Francisco, aiming to connect product leaders through talks, workshops, and networking opportunities.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-19","topics":["product-leadership","networking-events","ai-in-product"],"diataxisType":"reference","path":"/kb/articles/the-lenny-and-friends-summit-is-back-7de634ec","sourceUrl":"https://www.lennysnewsletter.com/p/the-lenny-and-friends-summit-is-back"},{"id":"8560706c-04e6-400c-b61a-94e59829c055","title":"Release: llm-gemini 0.32a0","summary":"Simon Willison announces the release of llm-gemini 0.32a0, highlighting its features and inviting sponsorship for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-19","topics":["llm-gemini","monthly-briefing","sponsorship"],"diataxisType":"reference","path":"/kb/articles/release-llm-gemini-032a0-8560706c","sourceUrl":"https://simonwillison.net/2026/May/19/llm-gemini/#atom-everything"},{"id":"6869e085-648f-4cc3-bb6d-0d038cd96c92","title":"the aussie gov just screwed startup founders. my musings.","summary":"Geoffrey Huntley expresses frustration over new Australian government tax proposals that negatively impact startup founders and their financial planning.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-05-18","topics":["australian-startups","taxation","financial-planning"],"diataxisType":"explanation","path":"/kb/articles/the-aussie-gov-just-screwed-startup-founders-my-musings-6869e085","sourceUrl":"https://ghuntley.com/livid/"},{"id":"77db97f0-b530-45f8-9d5a-b4b37e088bb8","title":"The Next War Is Already Here. The West Isn't Ready. — Yaroslav Azhnyuk, The Fourth Law & Guest Host Noah Smith, Noahpinion","summary":"The article discusses the evolution of warfare in Ukraine, highlighting the technological advancements in drone warfare and the West's unpreparedness for modern conflicts.","author":"Swyx","authorId":"swyx","published":"2026-05-18","topics":["drone-warfare","geopolitics","military-technology","ukraine-war"],"diataxisType":"explanation","path":"/kb/articles/the-next-war-is-already-here-the-west-isnt-ready-yaroslav-az-77db97f0","sourceUrl":"https://www.latent.space/p/the-fourth-law"},{"id":"98efb619-8f7d-4305-b1e3-20eb81b89f78","title":"🎙️ How I AI: HTML is the new Markdown: How Anthropic engineers are building with Claude Code","summary":"Thariq Shihipar discusses how HTML is replacing Markdown for AI development, enhancing project planning and interaction with AI agents.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-18","topics":["html","ai-development","interactive-specs","design-systems"],"diataxisType":"how-to","path":"/kb/articles/-how-i-ai-html-is-the-new-markdown-how-anthropic-engineers-a-98efb619","sourceUrl":"https://www.lennysnewsletter.com/p/how-i-ai-html-is-the-new-markdown"},{"id":"b61426e3-16c3-4bbe-ab97-bfce3827a697","title":"HTML is the new Markdown: How Anthropic engineers are building with Claude Code | Thariq Shihipar","summary":"Thariq Shihipar from Anthropic discusses using HTML as a superior alternative to Markdown for AI workflows, enhancing engagement and product quality.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-18","topics":["html","markdown","ai-workflows","interactive-design","product-development"],"diataxisType":"how-to","path":"/kb/articles/html-is-the-new-markdown-how-anthropic-engineers-are-buildin-b61426e3","sourceUrl":"https://www.lennysnewsletter.com/p/html-is-the-new-markdown-how-anthropic"},{"id":"9c055617-9927-4533-a00d-e4258bc20cff","title":"Sighting: Glaucous-winged Gull, Brown Pelican, Snowy Egret, Canada Goose","summary":"Simon Willison shares his morning walk experience along the Los Angeles River, where he spotted a Glaucous-winged Gull, Brown Pelican, Snowy Egret, and Canada Goose.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-18","topics":["bird-watching","nature-photography","wildlife-sightings"],"diataxisType":"reference","path":"/kb/articles/sighting-glaucous-winged-gull-brown-pelican-snowy-egret-cana-9c055617","sourceUrl":"https://simonwillison.net/2026/May/18/sighting-362781627/#atom-everything"},{"id":"447f5269-62ea-49dd-b001-b236f3feb0ad","title":"GDS weighs in on the NHS’s decision to retreat from Open Source","summary":"The GDS responds to the NHS's decision to limit open source access, advocating for openness as the default approach in public sector projects.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-17","topics":["open-source","nhs","government-digital-service","vulnerability-management"],"diataxisType":"explanation","path":"/kb/articles/gds-weighs-in-on-the-nhss-decision-to-retreat-from-open-sour-447f5269","sourceUrl":"https://simonwillison.net/2026/May/17/gds-weighs-in/#atom-everything"},{"id":"8a8f7365-245c-4902-b924-1c0a4d5067b4","title":"Why we’re at the beginning of the AI hardware boom | Caitlin Kalinowski (ex–OpenAI, Meta, Apple)","summary":"Caitlin Kalinowski discusses her journey in AI hardware, the future of robotics, and insights from her experiences at OpenAI, Meta, and Apple.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-17","topics":["ai-hardware","robotics","consumer-electronics","vr","startups"],"diataxisType":"explanation","path":"/kb/articles/why-were-at-the-beginning-of-the-ai-hardware-boom-caitlin-ka-8a8f7365","sourceUrl":"https://www.lennysnewsletter.com/p/why-were-at-the-beginning-of-the"},{"id":"f3ac6ec8-1ef4-455e-a2a4-9d11c6ab8cd0","title":"🧠 Community Wisdom: Screening AI slop in hiring, Wispr Flow alternatives for voice transcription, multi-agent pipeline vs. MCP, and more","summary":"This week's Community Wisdom discusses AI screening in hiring, alternatives for voice transcription, and compares multi-agent pipelines with MCP.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-16","topics":["ai-hiring","voice-transcription","multi-agent-systems"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-screening-ai-slop-in-hiring-wispr-flow-alt-f3ac6ec8","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-screening-ai-slop"},{"id":"3020e741-73fc-43af-8387-a6ccbe4b5ed1","title":"A quote from Julia Evans","summary":"Julia Evans reflects on her journey of learning to appreciate CSS, emphasizing its complexities and the importance of understanding its challenges.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-16","topics":["css","web-development","frontend"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-julia-evans-3020e741","sourceUrl":"https://simonwillison.net/2026/May/16/julia-evans/#atom-everything"},{"id":"4b791eed-67b3-483c-b772-7bd8289ebced","title":"Warelay -> OpenClaw","summary":"Simon Willison discusses the evolution of the name OpenClaw, tracing its history through various iterations since its first commit in November 2025.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-16","topics":["open-source","whatsapp","ai-assistants","software-development"],"diataxisType":"explanation","path":"/kb/articles/warelay-openclaw-4b791eed","sourceUrl":"https://simonwillison.net/2026/May/16/openclaw-names/#atom-everything"},{"id":"56c76c34-6cb4-4672-864d-98b9fe2783b8","title":"[AINews] Cerebras' $60B IPO: Slowly, then All at Once","summary":"Cerebras' IPO at a $60B market cap marks a significant validation for the company, highlighting its role in serving large AI models amidst shifting market dynamics.","author":"Swyx","authorId":"swyx","published":"2026-05-16","topics":["cerebras-ipo","ai-hardware","inference-economics","compute-scarcity"],"diataxisType":"explanation","path":"/kb/articles/ainews-cerebras-60b-ipo-slowly-then-all-at-once-56c76c34","sourceUrl":"https://www.latent.space/p/ainews-cerebras-60b-ipo-slowly-then"},{"id":"2dd9b888-8d12-4ba8-94fe-07be867ebdfa","title":"Tool: QR code generator","summary":"Simon Willison introduces a customizable QR code generator that allows users to create QR codes from URLs or text with various design options.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-15","topics":["qr-codes","tool-development","customization"],"diataxisType":"how-to","path":"/kb/articles/tool-qr-code-generator-2dd9b888","sourceUrl":"https://simonwillison.net/2026/May/15/qr-code-generator/#atom-everything"},{"id":"a27bf774-6ebd-451d-b7a6-289d5f97d35c","title":"[AINews] Everything is Conductor","summary":"The article discusses the evolution of AI tools in coding and healthcare, highlighting GitHub's new app and the competitive landscape for agent-first UX.","author":"Swyx","authorId":"swyx","published":"2026-05-15","topics":["ai-agents","healthcare-ai","github","coding-tools"],"diataxisType":"explanation","path":"/kb/articles/ainews-everything-is-conductor-a27bf774","sourceUrl":"https://www.latent.space/p/ainews-everything-is-conductor"},{"id":"2bd8e7fe-ecbc-4ffc-853d-f370b4b2a6a3","title":"We Gave Every Employee an AI Agent. Here's What We're Doing Differently Now.","summary":"The article discusses the challenges and lessons learned from implementing AI agents in the workplace, emphasizing a shift from individual assistants to shared team resources.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-05-15","topics":["ai-agents","future-of-work","productivity","team-resources"],"diataxisType":"explanation","path":"/kb/articles/we-gave-every-employee-an-ai-agent-heres-what-were-doing-dif-2bd8e7fe","sourceUrl":"https://every.to/source-code/we-gave-every-employee-an-ai-agent-here-s-what-we-re-doing-differently-now"},{"id":"fde7dbeb-2eef-44be-9a1d-d77d37387074","title":"Release: datasette-llm-limits 0.1a0","summary":"The release of datasette-llm-limits 0.1a0 introduces a plugin for setting spending limits on LLM usage in Datasette, enhancing user control over costs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-15","topics":["datasette","llm-usage","plugin-development"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-limits-01a0-fde7dbeb","sourceUrl":"https://simonwillison.net/2026/May/15/datasette-llm-limits/#atom-everything"},{"id":"f0941b17-a1cc-415f-b066-f73803e32ac8","title":"Sighting: Western Gull, Rock Pigeon","summary":"Simon Willison shares his morning bird walk experience in Los Angeles, where he spotted a Western Gull and a Rock Pigeon.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-15","topics":["bird-watching","western-gull","rock-pigeon","los-angeles"],"diataxisType":"reference","path":"/kb/articles/sighting-western-gull-rock-pigeon-f0941b17","sourceUrl":"https://simonwillison.net/2026/May/15/sighting-361818285/#atom-everything"},{"id":"7918fe9d-53b7-49b5-8906-bbb2fac6d90f","title":"Release: inaturalist-clumper 0.1","summary":"Simon Willison announces the 0.1 release of inaturalist-clumper, a tool for publishing iNaturalist sightings on blogs, after several weeks of production use.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-15","topics":["inaturalist","blogging","software-release"],"diataxisType":"reference","path":"/kb/articles/release-inaturalist-clumper-01-7918fe9d","sourceUrl":"https://simonwillison.net/2026/May/15/inaturalist-clumper/#atom-everything"},{"id":"81737517-ed9e-4554-8475-611ee9214918","title":"Release: datasette-ip-rate-limit 0.1a0","summary":"Simon Willison introduces the datasette-ip-rate-limit plugin, designed to block IPs that overload the datasette.io site with requests.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-14","topics":["rate-limiting","datasette","web-development"],"diataxisType":"how-to","path":"/kb/articles/release-datasette-ip-rate-limit-01a0-81737517","sourceUrl":"https://simonwillison.net/2026/May/14/datasette-ip-rate-limit/#atom-everything"},{"id":"0c2d6e2f-f083-433d-9111-4356fa9e0b6f","title":"A quote from Mitchell Hashimoto","summary":"Mitchell Hashimoto discusses the increasing flexibility of programming languages, highlighting Bun's ability to switch languages quickly.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-14","topics":["programming-languages","software-development","rust"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-mitchell-hashimoto-0c2d6e2f","sourceUrl":"https://simonwillison.net/2026/May/14/mitchell-hashimoto/#atom-everything"},{"id":"db1db1df-3a41-4b77-96d3-bd3bff8add0b","title":"AI-Native Healthcare: 100M Doctor Visits, 10–20 Hours Saved, Prior Auth in Minutes — Janie Lee & Chai Asawa, Abridge","summary":"Abridge is revolutionizing healthcare documentation and workflows using AI, aiming to enhance patient-clinician interactions and streamline processes like prior authorization.","author":"Swyx","authorId":"swyx","published":"2026-05-14","topics":["ai-healthcare","clinical-documentation","patient-clinician-interaction","prior-authorization"],"diataxisType":"explanation","path":"/kb/articles/ai-native-healthcare-100m-doctor-visits-1020-hours-saved-pri-db1db1df","sourceUrl":"https://www.latent.space/p/abridge"},{"id":"983c46c3-90ea-49dd-83a8-a861d3d7e59c","title":"[AINews] Codex Rises, Claude Meters Programmatic Usage","summary":"The article discusses the rise of pro-Codex sentiment among AI engineers following the launch of GPT 5.5 and changes in Claude's pricing structure.","author":"Swyx","authorId":"swyx","published":"2026-05-14","topics":["ai-engineering","pricing-strategy","model-launch","programmatic-usage"],"diataxisType":"explanation","path":"/kb/articles/ainews-codex-rises-claude-meters-programmatic-usage-983c46c3","sourceUrl":"https://www.latent.space/p/ainews-codex-rises-claude-meters"},{"id":"e0353207-fae9-43b3-ab9c-3a9b4e74c47b","title":"Not so locked in any more","summary":"The article discusses the shift in programming language choices, highlighting a conversation about migrating apps to React Native and the decreasing lock-in effect of programming languages.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-14","topics":["react-native","programming-languages","app-development"],"diataxisType":"explanation","path":"/kb/articles/not-so-locked-in-any-more-e0353207","sourceUrl":"https://simonwillison.net/2026/May/14/not-so-locked-in/#atom-everything"},{"id":"2c2e1519-fd3f-4e99-8a20-62456f4a4eac","title":"[AINews] The End of Finetuning","summary":"The article discusses the implications of OpenAI's deprecation of finetuning APIs, highlighting shifts in the AI engineering landscape and the rise of alternative models.","author":"Swyx","authorId":"swyx","published":"2026-05-13","topics":["finetuning","ai-engineering","model-evaluation","inference-systems"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-end-of-finetuning-2c2e1519","sourceUrl":"https://www.latent.space/p/ainews-the-end-of-finetuning"},{"id":"6998d31c-b164-4a74-ab79-399e1eb47981","title":"Tool: CSP Allow-list Experiment","summary":"The article discusses an experiment demonstrating how to load an app in a CSP-protected sandboxed iframe and handle CSP errors effectively.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-13","topics":["csp","sandboxing","web-development","security"],"diataxisType":"explanation","path":"/kb/articles/tool-csp-allow-list-experiment-6998d31c","sourceUrl":"https://simonwillison.net/2026/May/13/csp-allow/#atom-everything"},{"id":"499c7d7f-a3ff-47ad-b300-4a85d106ae28","title":"Welcome to the Datasette blog","summary":"The Datasette blog has launched to share upcoming announcements and insights, built using OpenAI Codex. It aims to keep the community updated on developments.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-13","topics":["datasette","ai-assisted-programming","generative-ai"],"diataxisType":"explanation","path":"/kb/articles/welcome-to-the-datasette-blog-499c7d7f","sourceUrl":"https://simonwillison.net/2026/May/13/welcome-to-the-datasette-blog/#atom-everything"},{"id":"48724683-50f0-4091-957d-68ba77b49c62","title":"Quoting Boris Mann","summary":"Simon Willison reflects on the insights of Boris Mann, exploring his contributions to technology and community building.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-13","topics":["technology","community-building","insights"],"diataxisType":"explanation","path":"/kb/articles/quoting-boris-mann-48724683","sourceUrl":"https://simonwillison.net/2026/May/13/boris-mann/#atom-everything"},{"id":"8db1edb7-c2a3-416f-b302-69b7ab862265","title":"Release: llm 0.32a2","summary":"The release of llm 0.32a2 introduces a new endpoint for OpenAI models, enhancing reasoning capabilities and allowing users to see summarized reasoning tokens.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-12","topics":["openai","llms","generative-ai"],"diataxisType":"reference","path":"/kb/articles/release-llm-032a2-8db1edb7","sourceUrl":"https://simonwillison.net/2026/May/12/llm/#atom-everything"},{"id":"4735a4ca-24fb-4799-bafe-5b518c877668","title":"[AINews] Thinking Machines' Native Interaction Models - TML-Interaction-Small 276B-A12B - advances SOTA Realtime Voice and kills standard VAD","summary":"Thinking Machines introduces TML-Interaction-Small, a 276B parameter model that advances real-time voice interaction, emphasizing continuous interactivity and multimodal capabilities.","author":"Swyx","authorId":"swyx","published":"2026-05-12","topics":["real-time-voice","human-ai-collaboration","multimodal-interaction","ai-innovation"],"diataxisType":"explanation","path":"/kb/articles/ainews-thinking-machines-native-interaction-models-tml-inter-4735a4ca","sourceUrl":"https://www.latent.space/p/ainews-thinking-machines-native-interaction"},{"id":"dd6ce69b-357f-447d-8ee4-cdc2827381bc","title":"A quote from Mitchell Hashimoto","summary":"Mitchell Hashimoto discusses the motivations of Technical Decision Makers, emphasizing their focus on job security and adherence to industry trends.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-12","topics":["technical-decision-makers","job-security","industry-trends"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-mitchell-hashimoto-dd6ce69b","sourceUrl":"https://simonwillison.net/2026/May/12/mitchell-hashimoto/#atom-everything"},{"id":"416b6100-f46f-4a04-bb0d-1c95359add30","title":"A quote from Mo Bitar","summary":"Mo Bitar humorously suggests using automation buzzwords to impress CEOs and secure promotions, despite the lack of real substance behind the claims.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-12","topics":["automation","career-advice","ai-layoffs"],"diataxisType":"how-to","path":"/kb/articles/a-quote-from-mo-bitar-416b6100","sourceUrl":"https://simonwillison.net/2026/May/12/mo-bitar/#atom-everything"},{"id":"1e4ad5f8-1caa-4a35-80e9-1f815f2fdca3","title":"Release: datasette 1.0a29","summary":"The release of datasette 1.0a29 introduces new features and fixes, including improved visibility for table headers and a resolution for a critical segfault bug.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-12","topics":["open-source","data-exploration","bug-fixes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-10a29-1e4ad5f8","sourceUrl":"https://simonwillison.net/2026/May/12/datasette/#atom-everything"},{"id":"13c696f6-24e9-4ba9-8673-6331384a6fc1","title":"Your AI Use Is Breaking My Brain","summary":"The article discusses the overwhelming presence of AI in online writing, coining the term 'Zombie Internet' to describe the complex interactions between humans and AI.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-11","topics":["ai-writing","zombie-internet","digital-culture","content-creation"],"diataxisType":"explanation","path":"/kb/articles/your-ai-use-is-breaking-my-brain-13c696f6","sourceUrl":"https://simonwillison.net/2026/May/11/zombie-internet/#atom-everything"},{"id":"2d4372de-4958-4afe-8b2a-4d003f1371aa","title":"A quote from James Shore","summary":"James Shore emphasizes the importance of AI coding agents in reducing maintenance costs in proportion to increased coding speed, warning against unsustainable productivity gains.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-11","topics":["ai-coding","maintenance-costs","productivity","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-james-shore-2d4372de","sourceUrl":"https://simonwillison.net/2026/May/11/james-shore/#atom-everything"},{"id":"b1437402-6534-4d36-b333-ac2ab6e7f9cc","title":"Thoughts on GitLab’s workforce reduction","summary":"GitLab announces a workforce reduction and restructuring, aiming to flatten management and empower smaller teams, while shifting its values framework.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-11","topics":["gitlab","workforce-reduction","management-structure","software-development"],"diataxisType":"explanation","path":"/kb/articles/thoughts-on-gitlabs-workforce-reduction-b1437402","sourceUrl":"https://simonwillison.net/2026/May/11/gitlab-act-2/#atom-everything"},{"id":"4dc5313e-afc5-48e0-935b-b2166f675d92","title":"Spec-driven development: The AI engineering workflow at Notion | Ryan Nystrom","summary":"Ryan Nystrom discusses the AI engineering workflow at Notion, focusing on the spec-driven development process and the importance of fast CI in modern software development.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-11","topics":["ai-engineering","spec-driven-development","continuous-integration","notion-ai"],"diataxisType":"how-to","path":"/kb/articles/spec-driven-development-the-ai-engineering-workflow-at-notio-4dc5313e","sourceUrl":"https://www.lennysnewsletter.com/p/spec-driven-development-the-ai-engineering"},{"id":"8f613b8a-5085-4605-91cb-37ed410d2164","title":"TIL: Using LLM in the shebang line of a script","summary":"The article discusses using LLM in the shebang line of scripts, showcasing various examples of how to leverage LLM for generating outputs and executing functions.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-11","topics":["llm","scripting","programming","automation"],"diataxisType":"how-to","path":"/kb/articles/til-using-llm-in-the-shebang-line-of-a-script-8f613b8a","sourceUrl":"https://simonwillison.net/2026/May/11/llm-shebang/#atom-everything"},{"id":"93d7168f-6dfd-4edd-a308-e3a59d6003e8","title":"🎙️ How I AI: Quests, token leaderboards, and the elite AI adoption playbook & Notion’s spec-driven development ","summary":"The article discusses innovative strategies for AI adoption in organizations, emphasizing gamified internal tools and fostering a culture of curiosity and agency among employees.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-11","topics":["ai-adoption","internal-tools","gamification","team-culture"],"diataxisType":"how-to","path":"/kb/articles/-how-i-ai-quests-token-leaderboards-and-the-elite-ai-adoptio-93d7168f","sourceUrl":"https://www.lennysnewsletter.com/p/how-i-ai-quests-token-leaderboards"},{"id":"2153c9f6-4d0d-4ec1-a243-d473aded6cac","title":"Learning on the Shop floor","summary":"The article discusses Shopify's internal coding tool, River, which fosters a collaborative learning environment by making work visible and encouraging peer-to-peer learning.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-11","topics":["shopify","learning-environment","collaboration","coding-tools"],"diataxisType":"explanation","path":"/kb/articles/learning-on-the-shop-floor-2153c9f6","sourceUrl":"https://simonwillison.net/2026/May/11/learning-on-the-shop-floor/#atom-everything"},{"id":"9fc122d2-cb33-4a0a-9d44-d445466846bf","title":"How to build a company that withstands any era | Eric Ries, Lean Startup author","summary":"Eric Ries discusses how companies can protect their value and avoid failure in his new book, Incorruptible, highlighting governance structures and the risks of success.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-10","topics":["company-governance","startup-success","business-strategy"],"diataxisType":"how-to","path":"/kb/articles/how-to-build-a-company-that-withstands-any-era-eric-ries-lea-9fc122d2","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-build-a-company-that-withstands"},{"id":"741115e1-a6fa-4632-b8fc-4afa2d8b60b2","title":"A quote from New York Times Editors’ Note","summary":"The New York Times corrected a misattributed quote from Pierre Poilievre, revealing it was an A.I.-generated summary rather than an actual statement.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-10","topics":["ai-generated-content","media-accuracy","politics"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-new-york-times-editors-note-741115e1","sourceUrl":"https://simonwillison.net/2026/May/10/new-york-times-editors-note/#atom-everything"},{"id":"c5fc3ab0-02be-4746-8c5a-3b30dd23651e","title":"A quote from Andrew Quinn","summary":"Andrew Quinn discusses the importance of reinventing tools in programming to reach the true frontier of knowledge, rather than relying solely on existing solutions.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-10","topics":["programming","knowledge-reinvention","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-andrew-quinn-c5fc3ab0","sourceUrl":"https://simonwillison.net/2026/May/10/andrew-quinn/#atom-everything"},{"id":"57489e33-db9d-485c-b02b-a2fc4acaf825","title":"[AINews] Anthropic growing 10x/year while everyone else is laying off >10% of their workforce","summary":"Anthropic is experiencing rapid growth, being valued at $1-1.2T, while other companies face significant layoffs, highlighting a shift in the AI landscape.","author":"Swyx","authorId":"swyx","published":"2026-05-09","topics":["ai-growth","company-valuations","workforce-layoffs","ai-industry"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-growing-10xyear-while-everyone-else-is-layi-57489e33","sourceUrl":"https://www.latent.space/p/ainews-anthropic-growing-10xyear"},{"id":"f0275d37-dae7-4e5f-8c93-c11d71514d5d","title":"🧠 Community Wisdom: What to do when non-PMs start shipping directly to production, thoughts on Claude Code’s pricing A/B test, the use of gen AI in games, and more","summary":"This week's Community Wisdom discusses the implications of non-PMs shipping directly to production, Claude Code's pricing A/B test, and the role of generative AI in gaming.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-09","topics":["community-wisdom","product-management","generative-ai","pricing-strategy","gaming"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-what-to-do-when-non-pms-start-shipping-dir-f0275d37","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-what-to-do-when"},{"id":"887578b6-6da1-4280-9d66-9c0c06c6bb57","title":"A quote from Luke Curley","summary":"Luke Curley discusses the challenges of WebRTC in maintaining low latency during conference calls, emphasizing the trade-off between speed and audio quality.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-09","topics":["webrtc","audio-quality","latency-issues","voice-ai"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-luke-curley-887578b6","sourceUrl":"https://simonwillison.net/2026/May/9/luke-curley/#atom-everything"},{"id":"b2d0a22c-1721-42e9-98ef-1ddf76711172","title":"Using Claude Code: The Unreasonable Effectiveness of HTML","summary":"Thariq Shihipar advocates for using HTML over Markdown for outputs from Claude, highlighting its advantages for rich, interactive explanations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-08","topics":["html","markdown","ai-tools","interactive-explanations"],"diataxisType":"explanation","path":"/kb/articles/using-claude-code-the-unreasonable-effectiveness-of-html-b2d0a22c","sourceUrl":"https://simonwillison.net/2026/May/8/unreasonable-effectiveness-of-html/#atom-everything"},{"id":"db8af209-0c10-4e2e-b0fe-6b402112c266","title":"[AINews] GPT-Realtime-2, -Translate, and -Whisper: new SOTA realtime voice APIs","summary":"OpenAI has launched three new voice models, GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper, enhancing real-time voice interactions with improved reasoning and usability.","author":"Swyx","authorId":"swyx","published":"2026-05-08","topics":["voice-ai","gpt-realtime","openai","speech-technology"],"diataxisType":"explanation","path":"/kb/articles/ainews-gpt-realtime-2-translate-and-whisper-new-sota-realtim-db8af209","sourceUrl":"https://www.latent.space/p/ainews-gpt-realtime-2-translate-and"},{"id":"fba0b078-2b12-4084-b7b6-155e927fa779","title":"Notes on the xAI/Anthropic data center deal","summary":"Anthropic has struck a deal with SpaceX/xAI to use their Colossus data center, raising concerns about environmental impact and the future of xAI's Grok models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-07","topics":["data-centers","ai-ethics","environmental-impact"],"diataxisType":"explanation","path":"/kb/articles/notes-on-the-xaianthropic-data-center-deal-fba0b078","sourceUrl":"https://simonwillison.net/2026/May/7/xai-anthropic/#atom-everything"},{"id":"3c5f54b1-22af-4335-ae9f-939a0e2ba310","title":"Engineering","summary":"The article discusses recent updates and improvements in Claude Code quality, detailing various changes and enhancements made to the system.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-05-07","topics":["ai-agents","code-quality","software-engineering","tool-development"],"diataxisType":"explanation","path":"/kb/articles/engineering-3c5f54b1","sourceUrl":"https://www.anthropic.com/engineering"},{"id":"beb3e6a5-7634-4dbb-a4fc-350e1c88563b","title":"Release: llm-gemini 0.31","summary":"Simon Willison announces the release of llm-gemini 0.31, highlighting its significance in the evolving landscape of generative AI.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-07","topics":["generative-ai","llms","llm-release"],"diataxisType":"reference","path":"/kb/articles/release-llm-gemini-031-beb3e6a5","sourceUrl":"https://simonwillison.net/2026/May/7/llm-gemini/#atom-everything"},{"id":"128b0cca-f587-451e-b831-1c913f7eb245","title":"Behind the Scenes Hardening Firefox with Claude Mythos Preview","summary":"Mozilla leveraged the Claude Mythos preview to identify and fix hundreds of vulnerabilities in Firefox, significantly improving their security measures.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-07","topics":["firefox-security","ai-bug-reports","mozilla","vulnerability-fixing"],"diataxisType":"explanation","path":"/kb/articles/behind-the-scenes-hardening-firefox-with-claude-mythos-previ-128b0cca","sourceUrl":"https://simonwillison.net/2026/May/7/firefox-claude-mythos/#atom-everything"},{"id":"1a78b4a3-0871-488f-8f2b-24d0470ff35a","title":"Code with Claude: The 5 biggest updates explained","summary":"Claire discusses major updates from Anthropics Code with Claude event, focusing on features that enhance AI product development.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-07","topics":["ai-agents","automation","memory-systems","multi-agent-orchestration"],"diataxisType":"explanation","path":"/kb/articles/code-with-claude-the-5-biggest-updates-explained-1a78b4a3","sourceUrl":"https://www.lennysnewsletter.com/p/code-with-claude-the-5-biggest-updates"},{"id":"46e67c5e-c372-49aa-b695-7ba2d421d801","title":"Tool: GitHub Repo Stats","summary":"GitHub Repo Stats is a tool that provides detailed statistics about GitHub repositories, including commit counts and contributor information, by using the GitHub REST API.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-07","topics":["github","repository-stats","api-tools"],"diataxisType":"reference","path":"/kb/articles/tool-github-repo-stats-46e67c5e","sourceUrl":"https://simonwillison.net/2026/May/7/github-repo-stats/#atom-everything"},{"id":"f85e8e73-ad8e-48ac-98ca-03f406aab2ae","title":"Tool: Big Words","summary":"Simon Willison introduces a tool called Big Words for creating text slides in macOS presentations, which accepts URL query string arguments.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-07","topics":["macos-tools","presentations","web-development"],"diataxisType":"how-to","path":"/kb/articles/tool-big-words-f85e8e73","sourceUrl":"https://simonwillison.net/2026/May/7/big-words/#atom-everything"},{"id":"5f181e96-a59b-48e8-a08d-9f17dec394c6","title":"Newsroom","summary":"Anthropic Labs introduces Claude Design, a collaborative tool for creating visual work, alongside initiatives like Project Glasswing and insights from AI user feedback.","author":"Resources Anthropic","authorId":"resources-anthropic","published":"2026-05-07","topics":["ai-design","collaboration-tools","user-feedback","software-security"],"diataxisType":"explanation","path":"/kb/articles/newsroom-5f181e96","sourceUrl":"https://www.anthropic.com/news"},{"id":"6ede5258-2da7-4d7c-882d-b1aeaa63949d","title":"[AINews] Anthropic-SpaceXai's 300MW/$5B/yr deal for Colossus I, ARR growth is 8000% annualized","summary":"Anthropic announced a significant partnership with SpaceX to enhance Claude's capabilities, projecting an 8000% annualized ARR growth and introducing new features for Claude Managed Agents.","author":"Swyx","authorId":"swyx","published":"2026-05-07","topics":["ai-partnerships","cloud-computing","machine-learning","product-updates"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-spacexais-300mw5byr-deal-for-colossus-i-arr-6ede5258","sourceUrl":"https://www.latent.space/p/ainews-anthropic-spacexais-300mw5byr"},{"id":"e452a7c5-e91e-46fa-a16c-ab02e4de75f5","title":"[AINews] Silicon Valley gets Serious about Services","summary":"Silicon Valley is increasingly focusing on AI services, with companies like Anthropic and OpenAI launching initiatives to integrate AI into business processes.","author":"Swyx","authorId":"swyx","published":"2026-05-06","topics":["ai-services","business-integration","silicon-valley","ai-agents"],"diataxisType":"explanation","path":"/kb/articles/ainews-silicon-valley-gets-serious-about-services-e452a7c5","sourceUrl":"https://www.latent.space/p/ainews-silicon-valley-gets-serious"},{"id":"b31cc1cc-535b-4a48-88b3-e9ed34b9b5b5","title":"Vibe coding and agentic engineering are getting closer than I’d like","summary":"Simon Willison discusses the convergence of vibe coding and agentic engineering, reflecting on the implications for software quality and accountability in AI-assisted programming.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-06","topics":["ai-coding-tools","vibe-coding","agentic-engineering","software-quality"],"diataxisType":"explanation","path":"/kb/articles/vibe-coding-and-agentic-engineering-are-getting-closer-than--b31cc1cc","sourceUrl":"https://simonwillison.net/2026/May/6/vibe-coding-and-agentic-engineering/#atom-everything"},{"id":"bdab1d03-c099-482e-998e-e0f227a1b1f6","title":"Live blog: Code w/ Claude 2026","summary":"Simon Willison live blogs the Anthropic's Code w/ Claude event, highlighting keynotes on product updates and new features for Claude managed agents.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-06","topics":["ai-agents","product-updates","machine-learning","developer-tools"],"diataxisType":"explanation","path":"/kb/articles/live-blog-code-w-claude-2026-bdab1d03","sourceUrl":"https://simonwillison.net/2026/May/6/code-w-claude-2026/#atom-everything"},{"id":"d75dc58c-c314-4e23-bca8-17a28caa5430","title":"Quests, token leaderboards, and a skills marketplace: The elite AI adoption playbook | John Kim (Sendbird)","summary":"John Kim, CEO of Delight.ai, shares insights on transforming organizations into AI-native entities, highlighting innovative tools and strategies for AI adoption.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-06","topics":["ai-adoption","customer-experience","internal-tools","skills-marketplace","automation"],"diataxisType":"how-to","path":"/kb/articles/quests-token-leaderboards-and-a-skills-marketplace-the-elite-d75dc58c","sourceUrl":"https://www.lennysnewsletter.com/p/quests-token-leaderboards-and-a-skills"},{"id":"e14020ec-0782-4962-94ba-f3662175cacf","title":"Release: datasette-referrer-policy 0.1","summary":"The release of datasette-referrer-policy 0.1 addresses issues with OpenStreetMap tiles not displaying due to CAPTCHA and Referrer-Policy settings.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-05","topics":["datasette","openstreetmap","web-development"],"diataxisType":"how-to","path":"/kb/articles/release-datasette-referrer-policy-01-e14020ec","sourceUrl":"https://simonwillison.net/2026/May/5/datasette-referrer-policy/#atom-everything"},{"id":"62304b35-48c9-43ce-a821-1ddce9b63885","title":"Our AI started a cafe in Stockholm","summary":"The article discusses an AI-run cafe in Stockholm, highlighting humorous ordering mistakes and ethical concerns about AI's impact on human operations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-05","topics":["ai-experiments","ethical-ai","automation"],"diataxisType":"explanation","path":"/kb/articles/our-ai-started-a-cafe-in-stockholm-62304b35","sourceUrl":"https://simonwillison.net/2026/May/5/our-ai-started-a-cafe-in-stockholm/#atom-everything"},{"id":"f39f0e6c-50fe-4787-b8fc-8cb148fcfdb1","title":"🔬Doing Vibe Physics — Alex Lupsasca, OpenAI","summary":"Alex Lupsasca discusses the transformative impact of AI, particularly GPT-5, on theoretical physics, showcasing its ability to solve complex problems rapidly.","author":"Swyx","authorId":"swyx","published":"2026-05-05","topics":["ai-in-physics","gpt-5","theoretical-physics","black-holes","ai-research"],"diataxisType":"explanation","path":"/kb/articles/doing-vibe-physics-alex-lupsasca-openai-f39f0e6c","sourceUrl":"https://www.latent.space/p/lupsasca"},{"id":"d4cc68f5-9d46-44b6-8149-4549eb76dee3","title":"Release: datasette-llm 0.1a7","summary":"The release of datasette-llm 0.1a7 introduces a plugin for LLM integration, allowing configuration of default options for specific models in Datasette.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-05","topics":["llm-integration","datasette","plugin-development"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a7-d4cc68f5","sourceUrl":"https://simonwillison.net/2026/May/5/datasette-llm/#atom-everything"},{"id":"29d1c26c-a471-4858-bffb-00196082bdb5","title":"Why SaaS freemium playbooks don’t work in AI, and what to do instead","summary":"The article discusses the challenges of monetizing AI products using traditional SaaS freemium models and proposes a new framework for AI paywalls.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-05","topics":["saas","ai-monetization","freemium-model","product-strategy"],"diataxisType":"how-to","path":"/kb/articles/why-saas-freemium-playbooks-dont-work-in-ai-and-what-to-do-i-29d1c26c","sourceUrl":"https://www.lennysnewsletter.com/p/why-saas-freemium-playbooks-dont"},{"id":"159947cd-3307-45d3-96d3-2a3171877dde","title":"A quote from John Gruber","summary":"Simon Willison shares a quote from John Gruber about Y Combinator's stake in OpenAI, revealing it to be approximately 0.6% worth over $5 billion.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-05","topics":["y-combinator","openai","investment","ai"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-john-gruber-159947cd","sourceUrl":"https://simonwillison.net/2026/May/5/john-gruber/#atom-everything"},{"id":"5da60b1d-a8b6-484f-a89e-2c2aa393aa3f","title":"Release: llm-echo 0.5a0","summary":"Simon Willison announces the release of llm-echo 0.5a0, a debug plugin for LLM that provides a fake echo model for automated testing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-05","topics":["llm-echo","debugging","automated-testing","plugin-release"],"diataxisType":"reference","path":"/kb/articles/release-llm-echo-05a0-5da60b1d","sourceUrl":"https://simonwillison.net/2026/May/5/llm-echo/#atom-everything"},{"id":"bb59146f-f6cd-452c-90c6-8542e3a6fb3c","title":"April 2026 newsletter","summary":"The April 2026 newsletter by Simon Willison covers updates on Opus 4.7, GPT-5.5, and other highlights including LLM security research and model releases.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-04","topics":["gpt-5-5","llm-security","model-releases","newsletter"],"diataxisType":"reference","path":"/kb/articles/april-2026-newsletter-bb59146f","sourceUrl":"https://simonwillison.net/2026/May/4/april-newsletter/#atom-everything"},{"id":"0e25b69c-8611-42b3-a061-e9cfe2e57def","title":"🎙️ This week on How I AI: The internal AI tool that’s transforming how Stripe designs products","summary":"Owen Williams discusses Protodash, Stripe's internal AI prototyping tool that enhances collaboration between designers and PMs, enabling rapid, high-quality prototypes.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-04","topics":["ai-prototyping","design-systems","collaboration-tools","stripe"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-the-internal-ai-tool-thats-transformi-0e25b69c","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-the-internal"},{"id":"351c2fbe-d8ea-4408-8b16-24f9f819779c","title":"[AINews] The Other vs The Utility","summary":"The article discusses the evolving nature of AI, contrasting the utility-focused approach of models like GPT with the morally engaged philosophy of Claude, and the implications for AGI development.","author":"Swyx","authorId":"swyx","published":"2026-05-04","topics":["ai-culture","machine-intelligence","agile-development","ai-ethics"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-other-vs-the-utility-351c2fbe","sourceUrl":"https://www.latent.space/p/ainews-the-other-vs-the-utility"},{"id":"61d4e239-f2ba-4df1-8667-1d171f818eee","title":"Granite 4.1 3B SVG Pelican Gallery","summary":"Simon Willison discusses IBM's release of the Granite 4.1 family of LLMs and experiments with generating SVG images of a pelican riding a bicycle.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-04","topics":["granite-4-1","llms","svg-generation","model-experimentation"],"diataxisType":"explanation","path":"/kb/articles/granite-41-3b-svg-pelican-gallery-61d4e239","sourceUrl":"https://simonwillison.net/2026/May/4/granite-41-3b-svg-pelican-gallery/#atom-everything"},{"id":"f7af004a-099f-4152-9ced-f5d713fd93cd","title":"The internal AI tool that’s transforming how Stripe designs products | Owen Williams","summary":"Owen Williams discusses Protodash, an AI-powered prototyping tool at Stripe that enhances design processes for both designers and PMs.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-04","topics":["ai-tools","product-design","prototyping","stripe"],"diataxisType":"how-to","path":"/kb/articles/the-internal-ai-tool-thats-transforming-how-stripe-designs-p-f7af004a","sourceUrl":"https://www.lennysnewsletter.com/p/the-internal-ai-tool-thats-transforming"},{"id":"340c8ee8-37c7-4417-8cf9-00abe3aa3aa5","title":"Research: TRE Python binding — ReDoS robustness demo","summary":"This article discusses a Python ctypes binding to the TRE regex library, showcasing its robustness against ReDoS attacks compared to Python's built-in `re` module.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-04","topics":["regex-performance","redos","python","ctypes","benchmarking"],"diataxisType":"explanation","path":"/kb/articles/research-tre-python-binding-redos-robustness-demo-340c8ee8","sourceUrl":"https://simonwillison.net/2026/May/4/tre-python-binding/#atom-everything"},{"id":"67f548b8-7ca2-4f39-8db9-894e9390f81f","title":"A quote from Andy Masley","summary":"Andy Masley argues against the notion that data center construction is depleting farmland, citing significant agricultural productivity despite land sales.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-04","topics":["land-use","data-centers","agriculture"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-andy-masley-67f548b8","sourceUrl":"https://simonwillison.net/2026/May/4/andy-masley/#atom-everything"},{"id":"9171ed5d-074a-4fa5-9aac-ac4c73d1039b","title":"Tool: Redis Array Playground","summary":"Salvatore Sanfilippo introduced a new array data type to Redis, with various commands for manipulation, showcased in an interactive playground.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-04","topics":["redis","data-types","web-development","interactive-tools"],"diataxisType":"explanation","path":"/kb/articles/tool-redis-array-playground-9171ed5d","sourceUrl":"https://simonwillison.net/2026/May/4/redis-array/#atom-everything"},{"id":"351b1187-b2bc-4eb3-949e-6cf3094b833d","title":"A quote from Anthropic","summary":"The article discusses an automatic classifier's findings on sycophancy in conversations with Claude, highlighting specific domains where sycophantic behavior is more prevalent.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-03","topics":["ai-research","human-computer-interaction","social-behavior"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-anthropic-351b1187","sourceUrl":"https://simonwillison.net/2026/May/3/anthropic/#atom-everything"},{"id":"d939c0b5-d5bb-4a9c-88f4-d20fbda775c8","title":"Why cultivating agency matters more than cultivating skills in the AI era | Max Schoening (Head of Product, Notion)","summary":"Max Schoening discusses the importance of cultivating agency over skills in the AI era, emphasizing how this mindset can help individuals thrive in product development.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-03","topics":["ai-agents","product-development","design-thinking","software-quality"],"diataxisType":"explanation","path":"/kb/articles/why-cultivating-agency-matters-more-than-cultivating-skills--d939c0b5","sourceUrl":"https://www.lennysnewsletter.com/p/why-cultivating-agency-matters-more"},{"id":"29f03b41-77dc-4f99-9ad8-1d1302f74eff","title":"🧠 Community Wisdom: Claude Code tips for ADHD users, resources for managing up, going from corporate to startup, tiring of being your company’s AI evangelist, and more","summary":"This week's Community Wisdom shares insights on ADHD management, transitioning from corporate to startup, and the challenges of being an AI evangelist.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-05-02","topics":["adhd","corporate-to-startup","ai-evangelism","community-resources"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-claude-code-tips-for-adhd-users-resources--29f03b41","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-claude-code-tips"},{"id":"53d92955-d5d9-4b3d-8ad1-e2b6b508600c","title":"[AINews] AI Engineer World's Fair — Autoresearch, Memory, World Models, Tokenmaxxing, Agentic Commerce, and Vertical AI Call for Speakers","summary":"The AIE Worlds Fair is calling for speakers for its Wave 2, focusing on new tracks like Autoresearch and Agentic Commerce, aiming to showcase cutting-edge AI research.","author":"Swyx","authorId":"swyx","published":"2026-05-02","topics":["ai-engineering","autoresearch","agentic-commerce","vertical-ai","memory"],"diataxisType":"reference","path":"/kb/articles/ainews-ai-engineer-worlds-fair-autoresearch-memory-world-mod-53d92955","sourceUrl":"https://www.latent.space/p/ainews-ai-engineer-worlds-fair-autoresearch"},{"id":"ec23bdcc-3665-4ba6-ab19-5366b3969758","title":"Sightings","summary":"Simon Willison shares his experiences with a new camera and integrating wildlife photos into his blog using iNaturalist.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-02","topics":["wildlife-photography","blogging","iNaturalist","technology"],"diataxisType":"how-to","path":"/kb/articles/sightings-ec23bdcc","sourceUrl":"https://simonwillison.net/2026/May/2/sightings/#atom-everything"},{"id":"b132436e-7e02-40d2-9ee2-dc8376ec79bb","title":"Claude Code for Product Managers","summary":"Dan Shipper discusses how Claude Code has transformed product management by streamlining workflows and enhancing focus on critical tasks.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-05-01","topics":["product-management","ai-tools","workflow-optimization"],"diataxisType":"how-to","path":"/kb/articles/claude-code-for-product-managers-b132436e","sourceUrl":"https://every.to/source-code/claude-code-for-product-managers"},{"id":"48ff65d4-7e3e-4cdb-9565-871eebaa6057","title":"Tool: iNaturalist Sightings","summary":"Simon Willison created a tool to group his iNaturalist observations by time and location, built entirely on his phone using Python and hosted on GitHub.","author":"Simon Willison","authorId":"simon-willison","published":"2026-05-01","topics":["inaturalist","python","web-development","data-visualization"],"diataxisType":"how-to","path":"/kb/articles/tool-inaturalist-sightings-48ff65d4","sourceUrl":"https://simonwillison.net/2026/May/1/inat-sightings/#atom-everything"},{"id":"29860c96-fdc9-4d05-a916-bf73e5f5e978","title":"[AINews] Agents for Everything Else: Codex for Knowledge Work, Claude for Creative Work","summary":"The article discusses recent updates to Codex and Claude, highlighting Codex's expansion into knowledge work and Claude's new creative tools amidst evolving AI capabilities.","author":"Swyx","authorId":"swyx","published":"2026-05-01","topics":["ai-agents","knowledge-work","creative-tools","openai","codex"],"diataxisType":"explanation","path":"/kb/articles/ainews-agents-for-everything-else-codex-for-knowledge-work-c-29860c96","sourceUrl":"https://www.latent.space/p/ainews-agents-for-everything-else"},{"id":"9053483c-7db5-431d-a8b6-65da6dc54306","title":"The Zig project's rationale for their firm anti-AI contribution policy","summary":"The Zig project enforces a strict anti-LLM policy to prioritize contributor growth over LLM-assisted contributions, emphasizing the value of human involvement in open source.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-30","topics":["open-source","ai-policy","contributor-development"],"diataxisType":"explanation","path":"/kb/articles/the-zig-projects-rationale-for-their-firm-anti-ai-contributi-9053483c","sourceUrl":"https://simonwillison.net/2026/Apr/30/zig-anti-ai/#atom-everything"},{"id":"1dff5c0f-43cf-494c-a77c-8dc334510c60","title":"Our evaluation of OpenAI’s GPT-5.5 cyber capabilities","summary":"Simon Willison evaluates the cyber capabilities of OpenAI's GPT-5.5, discussing its implications for AI security and development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-30","topics":["ai-security","generative-ai","gpt"],"diataxisType":"explanation","path":"/kb/articles/our-evaluation-of-openais-gpt-55-cyber-capabilities-1dff5c0f","sourceUrl":"https://simonwillison.net/2026/Apr/30/gpt-55-cyber-capabilities/#atom-everything"},{"id":"cc001aff-1ae1-49f8-bf37-8ddfec61ee64","title":"[AINews] The Inference Inflection","summary":"The article discusses the rising demand for CPU compute in the AI inference landscape, highlighting key industry insights and trends.","author":"Swyx","authorId":"swyx","published":"2026-04-30","topics":["cpu-compute","ai-inference","tech-industry","gpu-workloads"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-inference-inflection-cc001aff","sourceUrl":"https://www.latent.space/p/ainews-the-inference-inflection"},{"id":"457a26e1-5f0e-4355-9ba0-2280ee672b16","title":"A quote from Andrew Kelley","summary":"The article discusses the distinction between human mistakes and LLM hallucinations in coding, using a metaphor about smokers to illustrate the differences in digital behavior.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-30","topics":["llm","coding","digital-behavior"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-andrew-kelley-457a26e1","sourceUrl":"https://simonwillison.net/2026/Apr/30/andrew-kelley/#atom-everything"},{"id":"ae87c268-22fe-4545-a51a-17f3f23f3919","title":"We need RSS for sharing abundant vibe-coded apps","summary":"Simon Willison discusses the need for RSS feeds to share vibe-coded apps, emphasizing the personal and frequent nature of app development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-30","topics":["rss-feeds","app-development","vibe-coding"],"diataxisType":"explanation","path":"/kb/articles/we-need-rss-for-sharing-abundant-vibe-coded-apps-ae87c268","sourceUrl":"https://simonwillison.net/2026/Apr/30/rss-vibe-coded-apps/#atom-everything"},{"id":"ebf2cfc0-ff3c-4bdf-99ea-cf77e6c97fee","title":"Codex CLI 0.128.0 adds /goal","summary":"Codex CLI 0.128.0 introduces a new /goal feature that allows users to set objectives for the coding agent, enhancing its looping capabilities until the goal is met or the token budget is exhausted.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-30","topics":["codex-cli","generative-ai","prompt-engineering"],"diataxisType":"explanation","path":"/kb/articles/codex-cli-01280-adds-goal-ebf2cfc0","sourceUrl":"https://simonwillison.net/2026/Apr/30/codex-goals/#atom-everything"},{"id":"a72f0190-21a1-4d6f-9603-8c9224b855d4","title":"[AINews] not much happened today","summary":"The article discusses the latest developments in AI, including model releases and trends, while acknowledging a quieter news day for AINews.","author":"Swyx","authorId":"swyx","published":"2026-04-29","topics":["ai-models","inference-systems","multiagents","open-source-ai"],"diataxisType":"reference","path":"/kb/articles/ainews-not-much-happened-today-a72f0190","sourceUrl":"https://www.latent.space/p/ainews-not-much-happened-today"},{"id":"9c28ec29-f4a2-4c8b-96c0-a77e5db3ca3a","title":"Release: llm 0.32a0","summary":"Simon Willison announces the release of llm 0.32a0, highlighting its features and inviting readers to support his work through sponsorship.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-29","topics":["llm-release","sponsorship","monthly-briefing"],"diataxisType":"reference","path":"/kb/articles/release-llm-032a0-9c28ec29","sourceUrl":"https://simonwillison.net/2026/Apr/29/llm-2/#atom-everything"},{"id":"36c09739-2021-4eb4-a264-b47842d055a4","title":"LLM 0.32a0 is a major backwards-compatible refactor","summary":"LLM 0.32a0 introduces significant changes to the LLM Python library, enhancing its ability to handle diverse input and output types through a new messaging system and streaming capabilities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-29","topics":["python-library","llm","software-development","ai-models","streaming-data"],"diataxisType":"reference","path":"/kb/articles/llm-032a0-is-a-major-backwards-compatible-refactor-36c09739","sourceUrl":"https://simonwillison.net/2026/Apr/29/llm/#atom-everything"},{"id":"5ea51482-bf7b-4cd4-ac40-46d4288321ff","title":"Release: llm 0.32a1","summary":"Simon Willison announces the release of llm 0.32a1, highlighting its features and inviting readers to support his work through sponsorship.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-29","topics":["llm-release","monthly-briefing","sponsorship"],"diataxisType":"reference","path":"/kb/articles/release-llm-032a1-5ea51482","sourceUrl":"https://simonwillison.net/2026/Apr/29/llm-3/#atom-everything"},{"id":"da5529d8-b79e-4de0-a07a-cf8a24e0507c","title":"Introducing talkie: a 13B vintage language model from 1930","summary":"The article introduces talkie, a 13B vintage language model trained on pre-1931 English text, exploring its capabilities and research potential.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-28","topics":["language-models","historical-ai","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/introducing-talkie-a-13b-vintage-language-model-from-1930-da5529d8","sourceUrl":"https://simonwillison.net/2026/Apr/28/talkie/#atom-everything"},{"id":"08cc6a67-309e-442a-b51f-47fc54aa3036","title":"[AINews] ImageGen is on the Path to AGI","summary":"The article discusses the advancements in multimodal AI models like GPT-Image-2 and their implications for achieving AGI, emphasizing the necessity of creative applications in AI development.","author":"Swyx","authorId":"swyx","published":"2026-04-28","topics":["agi","multimodal-ai","gpt-image-2","ai-development"],"diataxisType":"explanation","path":"/kb/articles/ainews-imagegen-is-on-the-path-to-agi-08cc6a67","sourceUrl":"https://www.latent.space/p/ainews-imagegen-is-on-the-path-to"},{"id":"8217338b-4cae-4687-a742-ff80e39479b4","title":"Your Couch-to-5K for AI","summary":"Lenny Rachitsky introduces a Couch-to-5K style program for AI learning, emphasizing small, manageable steps to build expertise and overcome procrastination.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-28","topics":["ai-learning","productivity","behavior-change","personal-development"],"diataxisType":"how-to","path":"/kb/articles/your-couch-to-5k-for-ai-8217338b","sourceUrl":"https://www.lennysnewsletter.com/p/your-couch-to-5k-for-ai"},{"id":"a4b24cc6-86a2-4deb-8004-5e4fd1135e79","title":"A quote from OpenAI Codex base_instructions","summary":"The article discusses a guideline from OpenAI Codex regarding the relevance of discussing certain creatures in user queries.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-28","topics":["openai-codex","ai-guidelines","gpt-5-5"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-openai-codex-base_instructions-a4b24cc6","sourceUrl":"https://simonwillison.net/2026/Apr/28/openai-codex/#atom-everything"},{"id":"a2f380d9-1d46-4d7e-96a5-07e2d5ad5bb4","title":"What’s new in pip 26.1—lockfiles and dependency cooldowns!","summary":"The article discusses the new features in pip 26.1, including lockfiles and dependency cooldowns, while dropping support for Python 3.9.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-28","topics":["python","pip","dependency-management","software-updates"],"diataxisType":"explanation","path":"/kb/articles/whats-new-in-pip-261lockfiles-and-dependency-cooldowns-a2f380d9","sourceUrl":"https://simonwillison.net/2026/Apr/28/pip-261/#atom-everything"},{"id":"56cabc95-53ce-4ae1-b38f-d3d2c24e8984","title":"A quote from Matthew Yglesias","summary":"Simon Willison shares a quote from Matthew Yglesias expressing a preference for AI-assisted software development over vibecoding.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-28","topics":["ai-assisted-programming","vibe-coding","software-development"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-matthew-yglesias-56cabc95","sourceUrl":"https://simonwillison.net/2026/Apr/28/matthew-yglesias/#atom-everything"},{"id":"685ec087-e6fe-4983-96dc-4ca21b7a2679","title":"🎙️ This week on How I AI: GPT 5.5, Claude Design, and GPT Images 2.0 hands-on reviews—plus an inside look at Memelord","summary":"This article reviews GPT 5.5, Claude Design, and GPT Images 2.0, highlighting their capabilities and limitations in real-world applications.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-27","topics":["gpt-5-5","claude-design","ai-images","autonomous-agents"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-gpt-55-claude-design-and-gpt-images-2-685ec087","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-gpt-55-claude"},{"id":"c0183879-9d13-47be-a97e-0001a7fbe5c9","title":"From a $6.90 newsletter to $3M API: How a non-coder built Memelord | Jason Levin","summary":"Jason Levin transformed Memelord from a $6.90 newsletter into a $3M API, showcasing innovative approaches to meme creation and marketing without coding.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-27","topics":["ai-meme-creation","no-code-development","startup-growth","marketing-strategies"],"diataxisType":"how-to","path":"/kb/articles/from-a-690-newsletter-to-3m-api-how-a-non-coder-built-memelo-c0183879","sourceUrl":"https://www.lennysnewsletter.com/p/from-a-690-newsletter-to-3m-api-how"},{"id":"6ebdc648-3ea5-4f9a-8bb8-516d1065ea08","title":"microsoft/VibeVoice","summary":"Microsoft's VibeVoice is a Whisper-style audio model for speech-to-text, featuring speaker diarization and released on January 21, 2026.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-27","topics":["speech-to-text","audio-processing","machine-learning"],"diataxisType":"how-to","path":"/kb/articles/microsoftvibevoice-6ebdc648","sourceUrl":"https://simonwillison.net/2026/Apr/27/vibevoice/#atom-everything"},{"id":"0d49796a-1a6d-4cc9-83f7-d3fb4f4a6b4b","title":"Speech translation in Google Meet is now rolling out to mobile devices","summary":"Google Meet's speech translation feature is now available on mobile devices, allowing real-time translation between multiple languages, though still in alpha testing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-27","topics":["speech-translation","google-meet","mobile-devices","language-technology"],"diataxisType":"explanation","path":"/kb/articles/speech-translation-in-google-meet-is-now-rolling-out-to-mobi-0d49796a","sourceUrl":"https://simonwillison.net/2026/Apr/27/speech-translation-in-google-meet-is-now-rolling-out-to-mobile-d/#atom-everything"},{"id":"a6afd69d-95eb-4246-bb32-260fe2f0a611","title":"Physical AI that Moves the World — Qasar Younis & Peter Ludwig, Applied Intuition","summary":"Qasar Younis and Peter Ludwig discuss the evolution of physical AI at Applied Intuition, emphasizing the importance of deployment and reliability in safety-critical systems.","author":"Swyx","authorId":"swyx","published":"2026-04-27","topics":["physical-ai","autonomy","safety-critical-systems","robotics","ai-deployment"],"diataxisType":"explanation","path":"/kb/articles/physical-ai-that-moves-the-world-qasar-younis-peter-ludwig-a-a6afd69d","sourceUrl":"https://www.latent.space/p/appliedintuition"},{"id":"f28d353c-a265-43fc-8439-883e8d8060d2","title":"Tracking the history of the now-deceased OpenAI Microsoft AGI clause","summary":"The article discusses the evolution and eventual termination of the AGI clause in the Microsoft-OpenAI partnership, highlighting key developments and implications.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-27","topics":["openai","microsoft","agi-clause","partnership","artificial-intelligence"],"diataxisType":"explanation","path":"/kb/articles/tracking-the-history-of-the-now-deceased-openai-microsoft-ag-f28d353c","sourceUrl":"https://simonwillison.net/2026/Apr/27/now-deceased-agi-clause/#atom-everything"},{"id":"5aa6c61a-89be-44a9-87dd-da7ff995d871","title":"Snapchat CEO: Why distribution has become the most important moat | Evan Spiegel","summary":"Evan Spiegel discusses the importance of distribution in consumer tech, Snap's innovative design process, and the impact of AI on the industry.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-26","topics":["consumer-tech","ai-innovation","design-process","competitive-advantage"],"diataxisType":"explanation","path":"/kb/articles/snapchat-ceo-why-distribution-has-become-the-most-important--5aa6c61a","sourceUrl":"https://www.lennysnewsletter.com/p/snapchat-ceo-why-distribution-is"},{"id":"ff232954-6f3a-4672-9601-914b8df5736c","title":"A quote from Romain Huet","summary":"Romain Huet confirms that OpenAI has unified Codex and the main model, with GPT-5.5 enhancing coding capabilities further.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-25","topics":["ai-models","coding","openai"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-romain-huet-ff232954","sourceUrl":"https://simonwillison.net/2026/Apr/25/romain-huet/#atom-everything"},{"id":"c7b51e5b-8f40-4f44-9351-042d78371e5b","title":"🧠 Community Wisdom: What great product leaders do differently, Claude vs. ChatGPT, receiving hard feedback two weeks into the job, tips for joining a pre-PMF startup, and more","summary":"This week's Community Wisdom discusses insights from product leaders, the differences between Claude and ChatGPT, and tips for joining a pre-PMF startup.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-25","topics":["product-leadership","feedback","startup-tips","ai-comparison"],"diataxisType":"how-to","path":"/kb/articles/-community-wisdom-what-great-product-leaders-do-differently--c7b51e5b","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-what-great-product"},{"id":"b2672669-d8ea-4ea0-8752-b88742d914a2","title":"GPT-5.5 prompting guide","summary":"The GPT-5.5 prompting guide offers tips for effectively using the new model, emphasizing the need for fresh prompts and user-visible updates during multi-step tasks.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-25","topics":["gpt-5-5","prompt-engineering","openai","api-guides"],"diataxisType":"how-to","path":"/kb/articles/gpt-55-prompting-guide-b2672669","sourceUrl":"https://simonwillison.net/2026/Apr/25/gpt-5-5-prompting-guide/#atom-everything"},{"id":"8d6f339c-6152-428c-92fb-31c481ad0862","title":"[AINews] DeepSeek V4 Pro (1.6T-A49B) and Flash (284B-A13B), Base and Instruct — runnable on Huawei Ascend chips","summary":"DeepSeek has launched its V4 models, featuring significant advancements in long-context capabilities and hybrid reasoning modes, marking a major update since V3.","author":"Swyx","authorId":"swyx","published":"2026-04-25","topics":["deep-learning","model-architecture","ai-research","huawei","open-source"],"diataxisType":"reference","path":"/kb/articles/ainews-deepseek-v4-pro-16t-a49b-and-flash-284b-a13b-base-and-8d6f339c","sourceUrl":"https://www.latent.space/p/ainews-deepseek-v4-pro-16t-a49b-and"},{"id":"a82092c1-22d6-4a88-b930-d2bd0154d472","title":"WHY ARE YOU LIKE THIS","summary":"Simon Willison discusses a chaotic image generated by ChatGPT Images 2.0, showcasing a horse riding an astronaut on a pelican riding a bicycle.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-25","topics":["generative-ai","chatgpt","text-to-image"],"diataxisType":"explanation","path":"/kb/articles/why-are-you-like-this-a82092c1","sourceUrl":"https://simonwillison.net/2026/Apr/25/why-are-you-like-this/#atom-everything"},{"id":"231bc475-efb9-4d95-94a6-ea8fbda7140c","title":"DeepSeek V4—almost on the frontier, a fraction of the price","summary":"DeepSeek has launched its V4 series with two models, Pro and Flash, offering competitive pricing and efficiency compared to other frontier models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["ai-models","deep-learning","cost-efficiency"],"diataxisType":"explanation","path":"/kb/articles/deepseek-v4almost-on-the-frontier-a-fraction-of-the-price-231bc475","sourceUrl":"https://simonwillison.net/2026/Apr/24/deepseek-v4/#atom-everything"},{"id":"25f465b1-d468-472e-8324-6012967270e8","title":"An update on recent Claude Code quality reports","summary":"The article discusses recent issues with Claude Code's quality, highlighting bugs in its harness that affected user experience and model performance.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["ai-models","software-quality","bug-fixes"],"diataxisType":"explanation","path":"/kb/articles/an-update-on-recent-claude-code-quality-reports-25f465b1","sourceUrl":"https://simonwillison.net/2026/Apr/24/recent-claude-code-quality-reports/#atom-everything"},{"id":"151a4dd1-5a7d-4286-bbd6-1d1b92984868","title":"russellromney/honker","summary":"The article discusses 'honker', a Rust SQLite extension that implements Postgres NOTIFY/LISTEN semantics, enabling efficient queue and stream management in SQLite.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["sqlite","rust","database-extensions","asynchronous-programming"],"diataxisType":"explanation","path":"/kb/articles/russellromneyhonker-151a4dd1","sourceUrl":"https://simonwillison.net/2026/Apr/24/honker/#atom-everything"},{"id":"bf3be779-eb10-495e-97ff-ac5ab7057c6b","title":"Tool: Millisecond Converter","summary":"Simon Willison introduces a tool for converting milliseconds, highlighting its utility in LLM developments and offering a subscription for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["tooling","llm-developments","sponsorship"],"diataxisType":"reference","path":"/kb/articles/tool-millisecond-converter-bf3be779","sourceUrl":"https://simonwillison.net/2026/Apr/24/milliseconds/#atom-everything"},{"id":"aba0b3d3-01fe-4a82-9b8a-40d119f50eb9","title":"Serving the For You feed","summary":"The article discusses Bluesky's For You feed, a customizable recommendation system run by spacecowboy, detailing its architecture and operational costs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["bluesky","recommendation-systems","custom-feeds","data-architecture"],"diataxisType":"explanation","path":"/kb/articles/serving-the-for-you-feed-aba0b3d3","sourceUrl":"https://simonwillison.net/2026/Apr/24/serving-the-for-you-feed/#atom-everything"},{"id":"c962ca58-ec8a-4bac-8602-b92b28c2a75a","title":"The people do not yearn for automation","summary":"The article discusses the disconnect between the rise of AI and the general public's aversion to automation, emphasizing that not everything can be reduced to data.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["ai-automation","public-perception","software-brain","human-experience"],"diataxisType":"explanation","path":"/kb/articles/the-people-do-not-yearn-for-automation-c962ca58","sourceUrl":"https://simonwillison.net/2026/Apr/24/the-people-do-not-yearn-for-automation/#atom-everything"},{"id":"7ef4ae15-c278-4ae1-8250-bcef26eaa22a","title":"[AINews] GPT 5.5 and OpenAI Codex Superapp ","summary":"OpenAI launched GPT-5.5, emphasizing improved coding capabilities and token efficiency, while also integrating Codex into its superapp strategy.","author":"Swyx","authorId":"swyx","published":"2026-04-24","topics":["gpt-5-5","openai-codex","ai-agents","token-efficiency"],"diataxisType":"explanation","path":"/kb/articles/ainews-gpt-55-and-openai-codex-superapp--7ef4ae15","sourceUrl":"https://www.latent.space/p/ainews-gpt-55-and-openai-codex-superapp"},{"id":"51ea461a-8ecb-47f3-a68a-d20d8ee210bd","title":"Release: llm 0.31","summary":"The release of llm 0.31 introduces new features for accessing large language models, including options for verbosity and image detail levels.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["large-language-models","openai","command-line-tools"],"diataxisType":"reference","path":"/kb/articles/release-llm-031-51ea461a","sourceUrl":"https://simonwillison.net/2026/Apr/24/llm/#atom-everything"},{"id":"2ff4cd00-a797-40ff-ae7a-3e6820527730","title":"It's a big one","summary":"This week's newsletter includes whimsical content featuring animals and links to blog posts, along with a new chapter of the Agentic Engineering Patterns guide.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-24","topics":["newsletter","agentic-engineering","blog-posts","humor"],"diataxisType":"reference","path":"/kb/articles/its-a-big-one-2ff4cd00","sourceUrl":"https://simonwillison.net/2026/Apr/24/weekly/#atom-everything"},{"id":"70aaa5e0-e5a3-4c32-ba30-0f4d77ed956e","title":"An update on recent Claude Code quality reports","summary":"Over the past month, we’ve been looking into reports that Claude’s responses have worsened for some users. We’ve traced these reports to three separate changes that affected Claude Code, the Claude Agent SDK, and Claude Cowork. The API was not impacted.\nAll three issues have now been resolved as of ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-04-24","topics":["AI","Technology"],"diataxisType":"explanation","path":"/kb/articles/an-update-on-recent-claude-code-quality-reports-70aaa5e0","sourceUrl":"https://www.anthropic.com/engineering/april-23-postmortem"},{"id":"81ffa40b-0e58-49e5-b8f5-d486fc7b1bda","title":"[AINews] Tasteful Tokenmaxxing","summary":"The article discusses the latest advancements in AI technology announced at Cloud Next, focusing on Tokenmaxxing and new models from major players like Google and Alibaba.","author":"Swyx","authorId":"swyx","published":"2026-04-23","topics":["ai-technology","tokenmaxxing","cloud-computing","open-models","enterprise-ai"],"diataxisType":"explanation","path":"/kb/articles/ainews-tasteful-tokenmaxxing-81ffa40b","sourceUrl":"https://www.latent.space/p/ainews-tasteful-tokenmaxxing"},{"id":"180a68ef-63ff-4bd4-ac16-f0c4242cf091","title":"Release: llm-openai-via-codex 0.1a0","summary":"Simon Willison announces the release of llm-openai-via-codex 0.1a0, highlighting its significance in LLM developments.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-23","topics":["openai","llm","codex","development","sponsorship"],"diataxisType":"reference","path":"/kb/articles/release-llm-openai-via-codex-01a0-180a68ef","sourceUrl":"https://simonwillison.net/2026/Apr/23/llm-openai-via-codex/#atom-everything"},{"id":"ea8b9d42-2265-4e5c-82f6-1df9cfdfc596","title":"A quote from Maggie Appleton","summary":"Simon Willison shares a quote from Maggie Appleton about the benefits of learning in public, highlighting how it can lead to unexpected opportunities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-23","topics":["digital-gardening","podcasting","streaming","competence","networking"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-maggie-appleton-ea8b9d42","sourceUrl":"https://simonwillison.net/2026/Apr/23/maggie-appleton/#atom-everything"},{"id":"0eab1fd7-dd2e-49a0-9235-13495aee5ea8","title":"AIE Europe Debrief + Agent Labs Thesis: Unsupervised Learning x Latent Space Crossover Special (2026)","summary":"The article discusses the evolution of AI since the last Unsupervised Learning x Latent Space Crossover special, covering topics like AI infrastructure, coding wars, and the future of AI applications.","author":"Swyx","authorId":"swyx","published":"2026-04-23","topics":["ai-infrastructure","coding-ai","unsupervised-learning","ai-startups","latent-space"],"diataxisType":"explanation","path":"/kb/articles/aie-europe-debrief-agent-labs-thesis-unsupervised-learning-x-0eab1fd7","sourceUrl":"https://www.latent.space/p/unsupervised-learning-2026"},{"id":"42578d88-49ba-4651-9555-18f861e8aad6","title":"GPT 5.5 just did what no other model could","summary":"Lenny Rachitsky reviews OpenAI's GPT 5.5, highlighting its advanced capabilities in various tasks and its efficiency compared to previous models.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-23","topics":["gpt-5-5","ai-testing","autonomous-agents","tech-debt"],"diataxisType":"explanation","path":"/kb/articles/gpt-55-just-did-what-no-other-model-could-42578d88","sourceUrl":"https://www.lennysnewsletter.com/p/gpt-55-just-did-what-no-other-model"},{"id":"31978327-676a-4436-a9a8-3e58d96f55b3","title":"Extract PDF text in your browser with LiteParse for the web","summary":"LiteParse is an open-source tool for extracting text from PDFs, now available for use directly in the browser, enhancing accessibility and usability.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-23","topics":["pdf-parsing","web-development","open-source","ocr","liteparse"],"diataxisType":"how-to","path":"/kb/articles/extract-pdf-text-in-your-browser-with-liteparse-for-the-web-31978327","sourceUrl":"https://simonwillison.net/2026/Apr/23/liteparse-for-the-web/#atom-everything"},{"id":"03422dc7-21fa-44ab-a5d3-657cba4960ca","title":"How Anthropic’s product team moves faster than anyone else | Cat Wu (Head of Product, Claude Code)","summary":"Cat Wu, Head of Product at Anthropic, discusses the rapid evolution of AI product development and the essential skills for product managers in the AI space.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-23","topics":["ai-product-development","product-management","anthropic"],"diataxisType":"how-to","path":"/kb/articles/how-anthropics-product-team-moves-faster-than-anyone-else-ca-03422dc7","sourceUrl":"https://www.lennysnewsletter.com/p/how-anthropics-product-team-moves"},{"id":"fa552a8d-377d-4a3d-82f8-5ad9dd14acfa","title":"A pelican for GPT-5.5 via the semi-official Codex backdoor API","summary":"The article discusses the release of GPT-5.5, its integration with OpenAI Codex, and the implications for API access and pricing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-23","topics":["gpt-5-5","openai-codex","api-access","ai-pricing"],"diataxisType":"explanation","path":"/kb/articles/a-pelican-for-gpt-55-via-the-semi-official-codex-backdoor-ap-fa552a8d","sourceUrl":"https://simonwillison.net/2026/Apr/23/gpt-5-5/#atom-everything"},{"id":"747b7034-3fc5-45fc-abaa-39b6d4fa8819","title":"Changes to GitHub Copilot Individual plans","summary":"GitHub has announced changes to its Copilot Individual plans, including usage limits and a pause on signups, amid rising compute demands from agentic workflows.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-22","topics":["github-copilot","pricing-changes","ai-agents"],"diataxisType":"explanation","path":"/kb/articles/changes-to-github-copilot-individual-plans-747b7034","sourceUrl":"https://simonwillison.net/2026/Apr/22/changes-to-github-copilot/#atom-everything"},{"id":"76bce63b-966b-40e1-a5e1-480f2334b200","title":"A quote from Bobby Holley","summary":"Bobby Holley reflects on the collaboration with Anthropic and the successful release of Firefox 150, emphasizing the importance of focus in overcoming challenges.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-22","topics":["firefox","vulnerabilities","collaboration","technology"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-bobby-holley-76bce63b","sourceUrl":"https://simonwillison.net/2026/Apr/22/bobby-holley/#atom-everything"},{"id":"4588b496-63d6-46a7-9f87-8c5471517911","title":"[AINews] OpenAI launches GPT-Image-2","summary":"OpenAI has launched GPT-Image-2, a powerful image generation model that enhances text rendering and usability for various applications.","author":"Swyx","authorId":"swyx","published":"2026-04-22","topics":["image-generation","openai","gpt-image-2","ai-tools"],"diataxisType":"explanation","path":"/kb/articles/ainews-openai-launches-gpt-image-2-4588b496","sourceUrl":"https://www.latent.space/p/ainews-openai-launches-gpt-image"},{"id":"63e5c795-1263-4fcb-8b09-abed20352ac1","title":"What Claude Design is actually good for (and why Figma isn’t dead yet)","summary":"Lenny Rachitsky explores the capabilities of Claude Design and its potential to complement or replace Figma, highlighting key use cases and limitations of AI design tools.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-22","topics":["ai-design","claude-design","figma","design-systems"],"diataxisType":"how-to","path":"/kb/articles/what-claude-design-is-actually-good-for-and-why-figma-isnt-d-63e5c795","sourceUrl":"https://www.lennysnewsletter.com/p/what-claude-design-is-actually-good"},{"id":"5a19a631-0d56-4acb-8eeb-e4e855a49bb5","title":"Is Claude Code going to cost $100/month? Probably not—it’s all very confusing","summary":"The article discusses the confusion surrounding Claude Code's pricing changes, which may increase to $100/month, and the implications for users and trust in Anthropic.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-22","topics":["pricing-strategy","ai-agents","transparency","user-trust"],"diataxisType":"explanation","path":"/kb/articles/is-claude-code-going-to-cost-100month-probably-notits-all-ve-5a19a631","sourceUrl":"https://simonwillison.net/2026/Apr/22/claude-code-confusion/#atom-everything"},{"id":"18ec5cad-317b-4b94-af9a-f8904c82c4a2","title":"Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model","summary":"Qwen3.6-27B is a new 27B dense model that outperforms its predecessor in coding benchmarks while being significantly smaller in size.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-22","topics":["ai-coding","model-comparison","hugging-face"],"diataxisType":"explanation","path":"/kb/articles/qwen36-27b-flagship-level-coding-in-a-27b-dense-model-18ec5cad","sourceUrl":"https://simonwillison.net/2026/Apr/22/qwen36-27b/#atom-everything"},{"id":"7d12fd00-fdcd-4726-b3a6-046990c0e06b","title":"Shopify’s AI Phase Transition: 2026 Usage Explosion, Unlimited Opus-4.6 Token Budget, Tangle, Tangent, SimGym — with Mikhail Parakhin, Shopify CTO","summary":"Mikhail Parakhin, Shopify's CTO, discusses the company's AI initiatives, including Tangle, Tangent, and SimGym, and the evolution of AI coding and deployment challenges.","author":"Swyx","authorId":"swyx","published":"2026-04-22","topics":["ai-adoption","ml-experimentation","customer-simulation","token-budgets"],"diataxisType":"explanation","path":"/kb/articles/shopifys-ai-phase-transition-2026-usage-explosion-unlimited--7d12fd00","sourceUrl":"https://www.latent.space/p/shopify"},{"id":"5a0fe41f-e391-4939-92c2-8207d03e0dc7","title":"scosman/pelicans_riding_bicycles","summary":"Simon Willison shares insights on generative AI and its developments in a link post dated April 21, 2026.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-21","topics":["generative-ai","llms","training-data","pelican-riding-a-bicycle"],"diataxisType":"explanation","path":"/kb/articles/scosmanpelicans_riding_bicycles-5a0fe41f","sourceUrl":"https://simonwillison.net/2026/Apr/21/scosman/#atom-everything"},{"id":"42bbcba9-df84-47a5-9837-7acd14151d37","title":"New: A free year of Cursor, Google AI Pro, Notion, Supabase, Gumloop, and Fin for subscribers","summary":"Lenny Rachitsky announces a free year of premium products for subscribers of his newsletter, enhancing their access to AI tools and resources.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-21","topics":["productivity-tools","ai-tools","subscription-offers","newsletter"],"diataxisType":"reference","path":"/kb/articles/new-a-free-year-of-cursor-google-ai-pro-notion-supabase-guml-42bbcba9","sourceUrl":"https://www.lennysnewsletter.com/p/new-a-free-year-of-cursor-google"},{"id":"d9235379-5a0e-4a20-8619-16f1ce6893d3","title":"[AINews] Moonshot Kimi K2.6: the world's leading Open Model refreshes to catch up to Opus 4.6 (ahead of DeepSeek v4?)","summary":"Moonshot Kimi K2.6 refreshes its leading open model, showcasing significant advancements and competition against other models like Gemini 3.1 and Qwen3.6-Max-Preview.","author":"Swyx","authorId":"swyx","published":"2026-04-21","topics":["open-models","ai-agents","machine-learning","model-comparison"],"diataxisType":"explanation","path":"/kb/articles/ainews-moonshot-kimi-k26-the-worlds-leading-open-model-refre-d9235379","sourceUrl":"https://www.latent.space/p/ainews-moonshot-kimi-k26-the-worlds"},{"id":"641f8362-80d0-4636-86e7-5e30e46d1f05","title":"Where’s the raccoon with the ham radio? (ChatGPT Images 2.0)","summary":"OpenAI's ChatGPT Images 2.0 was released, showcasing significant improvements in image generation, particularly in complex illustrations like 'Where's Waldo' style images.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-21","topics":["image-generation","openai","ai-models","ham-radio"],"diataxisType":"explanation","path":"/kb/articles/wheres-the-raccoon-with-the-ham-radio-chatgpt-images-20-641f8362","sourceUrl":"https://simonwillison.net/2026/Apr/21/gpt-image-2/#atom-everything"},{"id":"d45bbc3d-390a-4d4e-ad96-b09427adc87a","title":"A quote from Andreas Påhlsson-Notini","summary":"Andreas Påhlsson-Notini critiques AI agents for their human-like flaws, emphasizing their lack of focus and tendency to negotiate with reality.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-21","topics":["ai-agents","human-factors","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-andreas-phlsson-notini-d45bbc3d","sourceUrl":"https://simonwillison.net/2026/Apr/21/andreas-pahlsson-notini/#atom-everything"},{"id":"64d554bc-50e0-442b-b90e-85407763fb81","title":"🎙️ This week on How I AI: How Intercom 2x’d their engineering velocity with Claude Code","summary":"Brian Scanlan from Intercom discusses how the company doubled its engineering velocity in nine months using Claude Code, emphasizing the importance of measuring and improving workflows.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-20","topics":["engineering-velocity","ai-integration","productivity","automation","technical-leadership"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-how-intercom-2xd-their-engineering-ve-64d554bc","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-how-intercom"},{"id":"caf70c66-0de4-40b7-8a6b-e5736c722bee","title":"TIL: SQL functions in Google Sheets to fetch data from Datasette","summary":"The article discusses methods to fetch data from Datasette into Google Sheets using SQL functions, including examples of different approaches.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-20","topics":["google-sheets","datasette","sql-functions"],"diataxisType":"how-to","path":"/kb/articles/til-sql-functions-in-google-sheets-to-fetch-data-from-datase-caf70c66","sourceUrl":"https://simonwillison.net/2026/Apr/20/datasette-sql/#atom-everything"},{"id":"2a8428b0-1e66-4b1f-9223-ab5b710f5476","title":"🔬 Training Transformers to solve 95% failure rate of Cancer Trials — Ron Alfa & Daniel Bear, Noetik","summary":"The article discusses how AI models like Noetik's TARIO-2 can potentially improve cancer treatment success rates by better matching patients to therapies, addressing the 95% failure rate in trials.","author":"Swyx","authorId":"swyx","published":"2026-04-20","topics":["ai-in-healthcare","cancer-research","biotech-innovation"],"diataxisType":"explanation","path":"/kb/articles/-training-transformers-to-solve-95-failure-rate-of-cancer-tr-2a8428b0","sourceUrl":"https://www.latent.space/p/noetik"},{"id":"91528908-4d5b-4dcc-bb13-c694981208a1","title":"Claude Token Counter, now with model comparisons","summary":"Simon Willison updates the Claude Token Counter tool to allow model comparisons, highlighting changes in tokenization between Claude Opus 4.7 and 4.6.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-20","topics":["tokenization","model-comparison","ai-tools","claude"],"diataxisType":"reference","path":"/kb/articles/claude-token-counter-now-with-model-comparisons-91528908","sourceUrl":"https://simonwillison.net/2026/Apr/20/claude-token-counts/#atom-everything"},{"id":"24a555a8-55d2-401b-9d26-107bdbb75f18","title":"How Intercom 2x’d their engineering velocity in 9 months with Claude Code | Brian Scanlan","summary":"Brian Scanlan discusses how Intercom doubled their engineering velocity in nine months using Claude Code, focusing on AI-first engineering practices.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-20","topics":["ai-engineering","engineering-velocity","product-development","telemetry-infrastructure"],"diataxisType":"how-to","path":"/kb/articles/how-intercom-2xd-their-engineering-velocity-in-9-months-with-24a555a8","sourceUrl":"https://www.lennysnewsletter.com/p/how-intercom-2xd-their-engineering"},{"id":"d3598d5f-42eb-40a0-8b9e-035f94681118","title":"Release: llm-openrouter 0.6","summary":"The article announces the release of llm-openrouter 0.6, highlighting a new command for refreshing available models and the inclusion of an HTML UI for Kimi 2.6.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-20","topics":["llm-openrouter","model-refresh","html-ui"],"diataxisType":"reference","path":"/kb/articles/release-llm-openrouter-06-d3598d5f","sourceUrl":"https://simonwillison.net/2026/Apr/20/llm-openrouter/#atom-everything"},{"id":"a91e2371-3ada-44d1-ab03-1fa66885825e","title":"Headless everything for personal AI","summary":"The article discusses the rise of headless services for personal AIs, emphasizing their efficiency and potential impact on SaaS pricing models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-19","topics":["headless-services","personal-ai","api-economy","saas-pricing"],"diataxisType":"explanation","path":"/kb/articles/headless-everything-for-personal-ai-a91e2371","sourceUrl":"https://simonwillison.net/2026/Apr/19/headless-everything/#atom-everything"},{"id":"1ac8b259-66cc-4992-aa30-49199b28dba2","title":"Why half of product managers are in trouble | Nikhyl Singhal (Meta, Google)","summary":"Nikhyl Singhal discusses the challenges facing product managers, including job risks and the impact of AI on the industry.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-19","topics":["product-management","ai-impact","career-development"],"diataxisType":"explanation","path":"/kb/articles/why-half-of-product-managers-are-in-trouble-nikhyl-singhal-m-1ac8b259","sourceUrl":"https://www.lennysnewsletter.com/p/why-half-of-product-managers-are-in-trouble"},{"id":"80e61cb4-831a-453f-ac7d-445e998e40b6","title":"[AINews] The Two Sides of OpenClaw","summary":"The article discusses the contrasting narratives of OpenClaw presented at TED and AIE talks, highlighting security incidents and scaling challenges in open source.","author":"Swyx","authorId":"swyx","published":"2026-04-18","topics":["open-source","ai-security","scaling-issues","technology-conferences"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-two-sides-of-openclaw-80e61cb4","sourceUrl":"https://www.latent.space/p/ainews-the-two-sides-of-openclaw"},{"id":"5d5b3c91-a4f3-485e-8658-b31d8d69ccd5","title":"Changes in the system prompt between Claude Opus 4.6 and 4.7","summary":"The article discusses the updates in the system prompt of Claude Opus 4.7 compared to 4.6, highlighting changes in functionality and user interaction.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-18","topics":["ai-systems","prompt-engineering","user-interaction"],"diataxisType":"reference","path":"/kb/articles/changes-in-the-system-prompt-between-claude-opus-46-and-47-5d5b3c91","sourceUrl":"https://simonwillison.net/2026/Apr/18/opus-system-prompt/#atom-everything"},{"id":"927b858a-d171-4854-9eb7-7a89d3172bf4","title":"Research: Claude system prompts as a git timeline","summary":"The article discusses transforming Anthropic's Claude system prompt history into a git-based exploration tool for better tracking and comparison of prompt evolution.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-18","topics":["git","prompt-engineering","ai-research"],"diataxisType":"how-to","path":"/kb/articles/research-claude-system-prompts-as-a-git-timeline-927b858a","sourceUrl":"https://simonwillison.net/2026/Apr/18/extract-system-prompts/#atom-everything"},{"id":"8a6a1c68-c55c-4e90-860a-96b7903b093b","title":"Adding a new content type to my blog-to-newsletter tool - Agentic Engineering Patterns","summary":"The article discusses the addition of a new content type called 'beats' to a blog-to-newsletter tool, detailing the process and coding involved.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-18","topics":["blogging","newsletter","coding","substack","content-management"],"diataxisType":"how-to","path":"/kb/articles/adding-a-new-content-type-to-my-blog-to-newsletter-tool-agen-8a6a1c68","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/adding-a-new-content-type/#atom-everything"},{"id":"5dfa661a-f4be-4c06-9a93-a52cf8d75d7e","title":"🧠 Community Wisdom: Sharing Claude Code skills across the org, Why did lennysjobs.com shut down, opening an LLC for side work, AI for SMB sales, and more","summary":"This week's Community Wisdom discusses sharing Claude Code skills, the shutdown of lennysjobs.com, opening an LLC for side work, and AI applications for SMB sales.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-18","topics":["community-wisdom","ai-for-smbs","entrepreneurship","lennysjobs","claude-code"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-sharing-claude-code-skills-across-the-org--5dfa661a","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-sharing-claude-code"},{"id":"9c07f538-6215-48cf-af62-bd2ad9af4eb2","title":"Join us at PyCon US 2026 in Long Beach—we have new AI and security tracks this year","summary":"PyCon US 2026 will be held in Long Beach from May 13-19, featuring new AI and Security tracks, alongside community-focused events and talks.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-17","topics":["ai-track","security-track","python-community","conference-experience"],"diataxisType":"explanation","path":"/kb/articles/join-us-at-pycon-us-2026-in-long-beachwe-have-new-ai-and-sec-9c07f538","sourceUrl":"https://simonwillison.net/2026/Apr/17/pycon-us-2026/#atom-everything"},{"id":"310b0451-1073-4074-9cb7-f45b5d9528e2","title":"[AINews] Anthropic Claude Opus 4.7 - literally one step better than 4.6 in every dimension","summary":"Anthropic launched Claude Opus 4.7, showcasing improvements in efficiency, vision capabilities, and overall performance compared to 4.6, while maintaining pricing.","author":"Swyx","authorId":"swyx","published":"2026-04-17","topics":["ai-models","multimodal-ai","benchmarking","tokenization"],"diataxisType":"reference","path":"/kb/articles/ainews-anthropic-claude-opus-47-literally-one-step-better-th-310b0451","sourceUrl":"https://www.latent.space/p/ainews-anthropic-claude-opus-47-literally"},{"id":"9401b4b8-a02e-4938-a447-dfa1fb6943ce","title":"Release: datasette 1.0a28","summary":"The release of datasette 1.0a28 addresses compatibility issues from 1.0a27 and introduces new methods for managing database connections and resources.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-17","topics":["open-source","data-exploration","software-release"],"diataxisType":"reference","path":"/kb/articles/release-datasette-10a28-9401b4b8","sourceUrl":"https://simonwillison.net/2026/Apr/17/datasette/#atom-everything"},{"id":"ed74eac1-2519-469c-8b5b-af137baa3c65","title":"[AINews] RIP Pull Requests (2005-2026)","summary":"The article discusses the decline of Pull Requests in software development, highlighting the impact of Generative AI and the shift towards alternative collaboration methods.","author":"Swyx","authorId":"swyx","published":"2026-04-16","topics":["pull-requests","generative-ai","software-development","git","ai-agents"],"diataxisType":"explanation","path":"/kb/articles/ainews-rip-pull-requests-2005-2026-ed74eac1","sourceUrl":"https://www.latent.space/p/ainews-rip-pull-requests-2005-2026"},{"id":"4119b526-e282-456d-bfb2-f5720dbc11f8","title":"Tool: datasette.io news preview","summary":"Simon Willison discusses the creation of a custom preview UI for the datasette.io news section, aimed at simplifying error checking in YAML formatting.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-16","topics":["datasette","yaml","ui-design","github"],"diataxisType":"how-to","path":"/kb/articles/tool-datasetteio-news-preview-4119b526","sourceUrl":"https://simonwillison.net/2026/Apr/16/datasette-io-preview/#atom-everything"},{"id":"133c5e00-833a-4baf-95f7-ece7d9533835","title":"Release: llm-anthropic 0.25","summary":"The release of llm-anthropic 0.25 introduces new features including the claude-opus-4.7 model and enhanced output options for improved user experience.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-16","topics":["llm","anthropic","model-release","ai-development"],"diataxisType":"reference","path":"/kb/articles/release-llm-anthropic-025-133c5e00","sourceUrl":"https://simonwillison.net/2026/Apr/16/llm-anthropic/#atom-everything"},{"id":"5e43bab8-2c6d-4c37-867f-3e1daf7d0ea0","title":"Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7","summary":"Simon Willison compares the Qwen3.6-35B-A3B and Claude Opus 4.7 models using a humorous benchmark of generating pelicans, favoring Qwen for its superior output.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-16","topics":["ai-models","image-generation","benchmarking"],"diataxisType":"explanation","path":"/kb/articles/qwen36-35b-a3b-on-my-laptop-drew-me-a-better-pelican-than-cl-5e43bab8","sourceUrl":"https://simonwillison.net/2026/Apr/16/qwen-beats-opus/#atom-everything"},{"id":"e0fc849e-680b-4430-b928-519a3ed51a77","title":"Release: datasette 1.0a27","summary":"The release of datasette 1.0a27 introduces significant changes, including the removal of Django-style CSRF tokens and the addition of a RenameTableEvent for renamed tables.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["open-source","data-exploration","sqlite"],"diataxisType":"reference","path":"/kb/articles/release-datasette-10a27-e0fc849e","sourceUrl":"https://simonwillison.net/2026/Apr/15/datasette/#atom-everything"},{"id":"532e0dfd-af45-4e68-8b9f-7357794e3222","title":"Notion’s Token Town: 5 Rebuilds, 100+ Tools, MCP vs CLIs and the Software Factory Future — Simon Last & Sarah Sachs of Notion","summary":"Swyx interviews Sarah Sachs and Simon Last from Notion about their journey in developing Custom Agents, exploring the challenges and innovations in AI tooling for productivity.","author":"Swyx","authorId":"swyx","published":"2026-04-15","topics":["ai-agents","productivity-tools","software-engineering","notion"],"diataxisType":"explanation","path":"/kb/articles/notions-token-town-5-rebuilds-100-tools-mcp-vs-clis-and-the--532e0dfd","sourceUrl":"https://www.latent.space/p/notion"},{"id":"a1afa0cb-46f3-4248-9ad5-2b31256d3e53","title":"Zig 0.16.0 release notes: “Juicy Main”","summary":"The Zig 0.16.0 release introduces 'Juicy Main', a dependency injection feature for the main() function, enhancing program structure and access to useful properties.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["zig","release-notes","programming","dependency-injection"],"diataxisType":"reference","path":"/kb/articles/zig-0160-release-notes-juicy-main-a1afa0cb","sourceUrl":"https://simonwillison.net/2026/Apr/15/juicy-main/#atom-everything"},{"id":"1f3b8622-d806-4fb6-af2e-c8676176d3c5","title":"A quote from John Gruber","summary":"John Gruber discusses the diminishing quality of third-party apps on Apple platforms, attributing it to a lack of motivation among developers.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["apple-platforms","app-development","software-quality"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-john-gruber-1f3b8622","sourceUrl":"https://simonwillison.net/2026/Apr/15/john-gruber/#atom-everything"},{"id":"ad22c932-02b3-4f66-8152-bbc3014bb4d6","title":"Release: datasette-ports 0.3","summary":"The release of datasette-ports 0.3 introduces features to identify running Datasette instances, showing their working directories and database paths.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["datasette","software-release","database-management"],"diataxisType":"reference","path":"/kb/articles/release-datasette-ports-03-ad22c932","sourceUrl":"https://simonwillison.net/2026/Apr/15/datasette-ports/#atom-everything"},{"id":"69e1867f-41b9-430a-8d59-40008efeccbb","title":"Tool: Gemini 3.1 Flash TTS","summary":"Gemini 3.1 Flash TTS is a tool for converting text to natural-sounding speech, featuring customizable voice options and dynamic delivery tags.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["text-to-speech","ai-tools","voice-synthesis"],"diataxisType":"how-to","path":"/kb/articles/tool-gemini-31-flash-tts-69e1867f","sourceUrl":"https://simonwillison.net/2026/Apr/15/gemini-flash-tts/#atom-everything"},{"id":"d5d2f0e4-d067-4885-bc34-7e2b7309c091","title":"Mini-Vibe Check: Claude Managed Agents Handle the Infrastructure Work","summary":"The article discusses the implications of Claude Managed Agents in AI infrastructure, highlighting their potential to revolutionize agent development and usage.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-04-15","topics":["ai-agents","infrastructure","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/mini-vibe-check-claude-managed-agents-handle-the-infrastruct-d5d2f0e4","sourceUrl":"https://every.to/source-code/mini-vibe-check-claude-managed-agents-handle-the-infrastructure-work"},{"id":"c72610e3-29c8-4166-bc20-8a3005f42736","title":"Listen: Not all AI agents are created equal","summary":"The episode discusses the complexities of prioritizing AI agent initiatives and categorizes agents into three architectural types, providing insights on success metrics and course correction.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-15","topics":["ai-agents","architecture","success-metrics"],"diataxisType":"how-to","path":"/kb/articles/listen-not-all-ai-agents-are-created-equal-c72610e3","sourceUrl":"https://www.lennysnewsletter.com/p/listen-not-all-ai-agents-are-created"},{"id":"d32e21db-bf31-4e1e-a54f-f76e8b3fb788","title":"Gemini 3.1 Flash TTS","summary":"Google released Gemini 3.1 Flash TTS, a new text-to-speech model that uses prompts to generate audio, showcasing a vibrant and energetic delivery style.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["text-to-speech","gemini-api","audio-production","ai-models"],"diataxisType":"explanation","path":"/kb/articles/gemini-31-flash-tts-d32e21db","sourceUrl":"https://simonwillison.net/2026/Apr/15/gemini-31-flash-tts/#atom-everything"},{"id":"fec8efc9-6cb1-4d2b-959b-5bb1b9b6d172","title":"[AINews] Humanity's Last Gasp","summary":"The article discusses the paradox of increased workload among knowledge workers despite advancements in AI, exploring the implications for productivity and future challenges.","author":"Swyx","authorId":"swyx","published":"2026-04-15","topics":["ai-workload","productivity","knowledge-workers","agent-technology"],"diataxisType":"explanation","path":"/kb/articles/ainews-humanitys-last-gasp-fec8efc9","sourceUrl":"https://www.latent.space/p/ainews-humanitys-last-gasp"},{"id":"b7122ac5-0b65-4055-a73a-4f72294d10d4","title":"Release: datasette-export-database 0.3a1","summary":"Simon Willison announces the release of datasette-export-database 0.3a1, highlighting its features and inviting support for his work.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["datasette","database","llm-developments"],"diataxisType":"reference","path":"/kb/articles/release-datasette-export-database-03a1-b7122ac5","sourceUrl":"https://simonwillison.net/2026/Apr/15/datasette-export-database/#atom-everything"},{"id":"fe0a7bea-fb5e-4f86-bd1e-c4b8633d122f","title":"A quote from Kyle Kingsbury","summary":"Kyle Kingsbury discusses the emerging roles of accountability in machine learning systems, highlighting the potential for various forms of oversight and responsibility.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-15","topics":["machine-learning","accountability","data-protection"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-kyle-kingsbury-fe0a7bea","sourceUrl":"https://simonwillison.net/2026/Apr/15/kyle-kingsbury/#atom-everything"},{"id":"ab91a6fb-90cd-4894-8142-666ac5b78d6d","title":"Trusted access for the next era of cyber defense","summary":"The article discusses OpenAI's new model GPT-5.4-Cyber and its Trusted Access program aimed at enhancing cybersecurity through user identity verification.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-14","topics":["cybersecurity","openai","gpt-5","trusted-access"],"diataxisType":"explanation","path":"/kb/articles/trusted-access-for-the-next-era-of-cyber-defense-ab91a6fb","sourceUrl":"https://simonwillison.net/2026/Apr/14/trusted-access-openai/#atom-everything"},{"id":"e090fefd-9ea9-4f64-a1cc-8ec18e9d4189","title":"Not all AI agents are created equal","summary":"The article discusses the importance of categorizing AI agents to prioritize development effectively, highlighting three architectural types and their implications for planning.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-14","topics":["ai-agents","product-development","prioritization","automation"],"diataxisType":"how-to","path":"/kb/articles/not-all-ai-agents-are-created-equal-e090fefd","sourceUrl":"https://www.lennysnewsletter.com/p/not-all-ai-agents-are-created-equal"},{"id":"febc892b-7f9b-48ac-a03c-fe170acf6ec4","title":"datasette PR #2689: Replace token-based CSRF with Sec-Fetch-Site header protection","summary":"The article discusses the replacement of token-based CSRF protection in Datasette with Sec-Fetch-Site header protection, streamlining the process and enhancing security.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-14","topics":["csrf-protection","datasette","security","software-development"],"diataxisType":"explanation","path":"/kb/articles/datasette-pr-2689-replace-token-based-csrf-with-sec-fetch-si-febc892b","sourceUrl":"https://simonwillison.net/2026/Apr/14/replace-token-based-csrf/#atom-everything"},{"id":"a7faa917-6816-4517-910b-35c998215a33","title":"Cybersecurity Looks Like Proof of Work Now","summary":"The article discusses the economic implications of cybersecurity practices, emphasizing the need for increased investment in security measures to effectively identify vulnerabilities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-14","topics":["cybersecurity","ai-safety","open-source","economic-incentives"],"diataxisType":"explanation","path":"/kb/articles/cybersecurity-looks-like-proof-of-work-now-a7faa917","sourceUrl":"https://simonwillison.net/2026/Apr/14/cybersecurity-proof-of-work/#atom-everything"},{"id":"2e52dbf3-01aa-48ae-8934-280a40585b36","title":"[AINews] Top Local Models List - April 2026","summary":"The article reviews the top local models recommended by the community, highlighting their strengths and use cases for various deployments.","author":"Swyx","authorId":"swyx","published":"2026-04-14","topics":["local-models","ai-recommendations","community-consensus"],"diataxisType":"reference","path":"/kb/articles/ainews-top-local-models-list-april-2026-2e52dbf3","sourceUrl":"https://www.latent.space/p/ainews-top-local-models-list-april"},{"id":"569cbf4f-58e6-419e-9659-40577dce4c0c","title":"Research: Exploring the new `servo` crate","summary":"The article explores the new `servo` crate, an embeddable browser engine for Rust, highlighting its features and capabilities, including a CLI tool for screenshots.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-13","topics":["rust","browser-engine","embeddable-libraries"],"diataxisType":"explanation","path":"/kb/articles/research-exploring-the-new-servo-crate-569cbf4f","sourceUrl":"https://simonwillison.net/2026/Apr/13/servo-crate-exploration/#atom-everything"},{"id":"ad81f1a9-7dd6-4f00-afb0-1742b7a98a84","title":"Steve Yegge","summary":"Steve Yegge discusses Google's AI adoption and internal culture, highlighting a divide among employees regarding technology use and a stagnant hiring environment.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-13","topics":["ai-adoption","google-engineering","tech-culture"],"diataxisType":"explanation","path":"/kb/articles/steve-yegge-ad81f1a9","sourceUrl":"https://simonwillison.net/2026/Apr/13/steve-yegge/#atom-everything"},{"id":"efdaeae9-6957-415d-ab2a-e82fff02fcf9","title":"Claude Cowork 101: How to automate your workday without touching code | JJ Englert (Tenex)","summary":"JJ Englert provides a comprehensive tutorial on using Claude Cowork, a tool for automating workflows without coding, covering project creation and AI integration.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-13","topics":["workflow-automation","ai-tools","productivity","no-code"],"diataxisType":"tutorial","path":"/kb/articles/claude-cowork-101-how-to-automate-your-workday-without-touch-efdaeae9","sourceUrl":"https://www.lennysnewsletter.com/p/claude-cowork-101-how-to-automate"},{"id":"e92548ee-527a-45a1-be75-7ef75530686e","title":"A quote from Bryan Cantrill","summary":"Bryan Cantrill discusses the inherent flaws of LLMs, emphasizing the importance of human laziness in optimizing systems and avoiding unnecessary complexity.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-13","topics":["llms","human-factors","system-design"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-bryan-cantrill-e92548ee","sourceUrl":"https://simonwillison.net/2026/Apr/13/bryan-cantrill/#atom-everything"},{"id":"c7ba8580-bd11-4055-a219-8d56b841ec85","title":"The Folder Is the Agent ","summary":"Dan Shipper discusses the concept of using project folders as agents in AI workflows, emphasizing the importance of context in enhancing AI performance.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-04-13","topics":["ai-agents","compound-engineering","workflow-automation"],"diataxisType":"how-to","path":"/kb/articles/the-folder-is-the-agent--c7ba8580","sourceUrl":"https://every.to/source-code/the-folder-is-the-agent"},{"id":"e738f292-ef24-47c1-bb3d-2e80df14d6fd","title":"🎙️ This week on How I AI: Claude Cowork tutorial for non-engineers + Build your own Slack inbox (for $0)","summary":"This article discusses a tutorial on building a custom AI-powered Slack inbox and automating workflows using Claude Cowork, aimed at non-engineers.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-13","topics":["ai-powered-tools","slack-integration","automation","productivity","non-technical-users"],"diataxisType":"tutorial","path":"/kb/articles/-this-week-on-how-i-ai-claude-cowork-tutorial-for-non-engine-e738f292","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-claude-cowork"},{"id":"bdef2945-3b13-4b19-bfb1-aed930ca5aea","title":"Gemma 4 audio with MLX","summary":"The article provides a recipe for transcribing audio files on macOS using the Gemma 4 E2B model with MLX, showcasing its capabilities with a practical example.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-12","topics":["audio-transcription","machine-learning","gemma-4","macos"],"diataxisType":"how-to","path":"/kb/articles/gemma-4-audio-with-mlx-bdef2945","sourceUrl":"https://simonwillison.net/2026/Apr/12/mlx-audio/#atom-everything"},{"id":"8c3047ad-421d-43cc-9830-fb9553fe36eb","title":"Hard truths about building in the AI era | Keith Rabois (Khosla Ventures)","summary":"Keith Rabois discusses hiring frameworks, the importance of customer interaction, and the evolving role of PMs in the AI era during his conversation with Lenny Rachitsky.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-12","topics":["ai-era","hiring-frameworks","product-management","venture-capital"],"diataxisType":"explanation","path":"/kb/articles/hard-truths-about-building-in-the-ai-era-keith-rabois-khosla-8c3047ad","sourceUrl":"https://www.lennysnewsletter.com/p/hard-truths-about-building-in-the-ai-era"},{"id":"16ada60d-0e30-4c5c-8486-1f52d9208e8f","title":"🧠 Community Wisdom: Vibe coding as a team sport, teaching interviewers to surface real signal, what’s missing from leadership offsites, and more","summary":"This week's Community Wisdom discusses team coding dynamics, effective interviewing techniques, and insights on leadership offsites.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-11","topics":["team-coding","interviewing-techniques","leadership-offsites"],"diataxisType":"explanation","path":"/kb/articles/-community-wisdom-vibe-coding-as-a-team-sport-teaching-inter-16ada60d","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-vibe-coding-as-a"},{"id":"13f0f387-505f-4500-babc-6469df34bfd7","title":"SQLite 3.53.0","summary":"SQLite 3.53.0 introduces significant improvements including enhanced ALTER TABLE functionality and new JSON functions, along with CLI mode enhancements.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-11","topics":["sqlite","database","cli-improvements","json-functions"],"diataxisType":"reference","path":"/kb/articles/sqlite-3530-13f0f387","sourceUrl":"https://simonwillison.net/2026/Apr/11/sqlite/#atom-everything"},{"id":"56f579fb-2cdd-4106-ab83-d0993d610f36","title":"Tool: SQLite Query Result Formatter Demo","summary":"Simon Willison introduces a demo tool for formatting SQLite query results, highlighting its features and benefits for developers.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-11","topics":["sqlite","query-formatter","developer-tools"],"diataxisType":"tutorial","path":"/kb/articles/tool-sqlite-query-result-formatter-demo-56f579fb","sourceUrl":"https://simonwillison.net/2026/Apr/11/sqlite-qrf/#atom-everything"},{"id":"e9f08cfc-a397-4625-9486-2118be895cae","title":"Kākāpō parrots","summary":"Simon Willison discusses Kākāpō parrots in a monthly briefing, highlighting important developments in the field of LLMs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-10","topics":["kakapo","llm-developments","monthly-briefing"],"diataxisType":"reference","path":"/kb/articles/kkp-parrots-e9f08cfc","sourceUrl":"https://simonwillison.net/2026/Apr/10/kakapo/#atom-everything"},{"id":"a18b69ac-7ab1-4d28-bf22-3c04cecb6669","title":"[AINews] AI Engineer Europe 2026","summary":"The article covers the highlights from AIE Day 1 in Europe, discussing advancements in AI models, orchestration patterns, and the evolving landscape of agent frameworks.","author":"Swyx","authorId":"swyx","published":"2026-04-10","topics":["ai-models","agent-frameworks","orchestration-patterns","open-source-ai"],"diataxisType":"explanation","path":"/kb/articles/ainews-ai-engineer-europe-2026-a18b69ac","sourceUrl":"https://www.latent.space/p/ainews-ai-engineer-europe-2026"},{"id":"b6226695-7644-4d51-8d29-e17bf0887acb","title":"ChatGPT voice mode is a weaker model","summary":"The OpenAI voice mode operates on an older, less capable model than expected, highlighting disparities in AI performance across different applications.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-10","topics":["ai-capabilities","voice-assistants","reinforcement-learning"],"diataxisType":"explanation","path":"/kb/articles/chatgpt-voice-mode-is-a-weaker-model-b6226695","sourceUrl":"https://simonwillison.net/2026/Apr/10/voice-mode-is-weaker/#atom-everything"},{"id":"77453e3a-7060-4825-86d8-b24554311abf","title":"Scaling Managed Agents: Decoupling the brain from the hands","summary":"Get started with Claude Managed Agents by following our\ndocs\n.\nA running topic on the Engineering Blog is how to\nbuild effective agents\nand\ndesign harnesses\nfor\nlong-running work\n. A common thread across this work is that harnesses encode assumptions about what Claude can’t do on its own. However, t","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-04-09","topics":["AI","Technology"],"diataxisType":"explanation","path":"/kb/articles/scaling-managed-agents-decoupling-the-brain-from-the-hands-77453e3a","sourceUrl":"https://www.anthropic.com/engineering/managed-agents"},{"id":"0bdc0534-2a14-48eb-8d5b-ebca26280547","title":"Tool: GitHub Repo Size","summary":"Simon Willison discusses the GitHub Repo Size tool, highlighting its features and the importance of monitoring repository sizes for developers.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-09","topics":["github","repo-size","developer-tools"],"diataxisType":"reference","path":"/kb/articles/tool-github-repo-size-0bdc0534","sourceUrl":"https://simonwillison.net/2026/Apr/9/github-repo-size/#atom-everything"},{"id":"4274d184-0130-45ef-981f-b4dd52fac4db","title":"How We Run a 25-person Company on Four AI Agents","summary":"The article discusses how Every, a 25-person company, utilizes four AI agents to streamline task management and improve efficiency in their operations.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-04-09","topics":["ai-agents","task-management","notion","automation"],"diataxisType":"how-to","path":"/kb/articles/how-we-run-a-25-person-company-on-four-ai-agents-4274d184","sourceUrl":"https://every.to/source-code/how-we-run-a-25-person-company-on-four-ai-agents"},{"id":"6e68bb8b-00d3-4f5e-b28f-3093a5e41b44","title":"Release: asgi-gzip 0.3","summary":"The release of asgi-gzip 0.3 addresses issues with gzip middleware for ASGI applications, ensuring proper handling of SSE responses.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-09","topics":["asgi","gzip","sse","datasette"],"diataxisType":"reference","path":"/kb/articles/release-asgi-gzip-03-6e68bb8b","sourceUrl":"https://simonwillison.net/2026/Apr/9/asgi-gzip/#atom-everything"},{"id":"103c10e6-ec3c-49c1-a769-f7c5de8f49b6","title":"Meta’s new model is Muse Spark, and meta.ai chat has some interesting tools","summary":"Meta has launched Muse Spark, a new AI model with competitive benchmarks and various tools for users, including image generation and code execution.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-08","topics":["ai-models","meta-ai","image-generation","api-tools"],"diataxisType":"explanation","path":"/kb/articles/metas-new-model-is-muse-spark-and-metaai-chat-has-some-inter-103c10e6","sourceUrl":"https://simonwillison.net/2026/Apr/8/muse-spark/#atom-everything"},{"id":"076a1a4a-0c06-4f72-9928-a131d998126f","title":"A quote from Giles Turnbull","summary":"Simon Willison shares a quote from Giles Turnbull reflecting on the mixed feelings about AI tools in professional settings.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-08","topics":["ai-tools","ai-ethics","human-voice"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-giles-turnbull-076a1a4a","sourceUrl":"https://simonwillison.net/2026/Apr/8/giles-turnbull/#atom-everything"},{"id":"72d57342-c2e5-4649-a2d7-b1da3dadd2e3","title":"I built a custom Slack inbox. It was easier than you’d think. | Yash Tekriwal (Clay)","summary":"Yash Tekriwal discusses building a custom Slack inbox to manage daily notifications and shares insights on productivity tools and AI applications.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-08","topics":["productivity","slack-integration","ai-tools","automation"],"diataxisType":"how-to","path":"/kb/articles/i-built-a-custom-slack-inbox-it-was-easier-than-youd-think-y-72d57342","sourceUrl":"https://www.lennysnewsletter.com/p/i-built-a-custom-slack-inbox-it-was"},{"id":"ebdf1ec1-2449-4f64-9369-dea8ff243f7c","title":"[AINews] Anthropic @ $30B ARR, Project GlassWing and Claude Mythos Preview — first model too dangerous to release since GPT-2","summary":"Anthropic's revenue surged to $30B ARR, unveiling Claude Mythos, a powerful model deemed too dangerous for public release, alongside Project Glasswing for cyber defense.","author":"Swyx","authorId":"swyx","published":"2026-04-08","topics":["ai-revenue","model-capabilities","cybersecurity","project-glasswing"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-30b-arr-project-glasswing-and-claude-mythos-ebdf1ec1","sourceUrl":"https://www.latent.space/p/ainews-anthropic-30b-arr-project"},{"id":"14e970f9-6c87-4e26-a1d1-3a7cf9d0e736","title":"[AINews] Meta Superintelligence Labs announces Muse Spark, first frontier model on their completely new stack","summary":"Meta Superintelligence Labs has launched Muse Spark, their first frontier model on a new infrastructure stack, with bigger models in development.","author":"Swyx","authorId":"swyx","published":"2026-04-08","topics":["meta","superintelligence","muse-spark","ai-models"],"diataxisType":"explanation","path":"/kb/articles/ainews-meta-superintelligence-labs-announces-muse-spark-firs-14e970f9","sourceUrl":"https://www.latent.space/p/ainews-meta-superintelligence-labs"},{"id":"04e05b37-3d49-4f8f-b9df-daf17fe24925","title":"Listen: A visual guide to getting out of a creative slump","summary":"Lenny Rachitsky discusses creative slumps with his wife, Michelle Rial, who shares 12 steps to overcome creative blocks and reignite inspiration.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-08","topics":["creativity","self-improvement","mental-health"],"diataxisType":"how-to","path":"/kb/articles/listen-a-visual-guide-to-getting-out-of-a-creative-slump-04e05b37","sourceUrl":"https://www.lennysnewsletter.com/p/listen-a-visual-guide-to-getting"},{"id":"3c1fb58c-c51b-478b-bd49-e494b3bec196","title":"GLM-5.1: Towards Long-Horizon Tasks","summary":"GLM-5.1 is a new AI model from Z.ai that showcases impressive capabilities in generating SVG and HTML content, including animations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-07","topics":["ai-models","svg-animation","css-animations"],"diataxisType":"explanation","path":"/kb/articles/glm-51-towards-long-horizon-tasks-3c1fb58c","sourceUrl":"https://simonwillison.net/2026/Apr/7/glm-51/#atom-everything"},{"id":"05cd022f-13f5-48a4-b2ab-be7495d58901","title":"Extreme Harness Engineering for Token Billionaires: 1M LOC, 1B toks/day, 0% human code, 0% human review — Ryan Lopopolo, OpenAI Frontier & Symphony","summary":"Ryan Lopopolo discusses Harness Engineering, a new approach to software development using AI agents with zero human-written code, showcased in OpenAI's internal projects.","author":"Swyx","authorId":"swyx","published":"2026-04-07","topics":["harness-engineering","ai-agents","software-development","openai"],"diataxisType":"explanation","path":"/kb/articles/extreme-harness-engineering-for-token-billionaires-1m-loc-1b-05cd022f","sourceUrl":"https://www.latent.space/p/harness-eng"},{"id":"aae99147-6dd8-4948-93b2-7f6a35eb1c1e","title":"[AINews] Gemma 4 crosses 2 million downloads","summary":"Gemma 4 has achieved 2 million downloads in its first week, showcasing rapid local adoption and strong user enthusiasm for its practical usability.","author":"Swyx","authorId":"swyx","published":"2026-04-07","topics":["gemma-4","ai-adoption","local-inference","hermes-agents","open-source"],"diataxisType":"reference","path":"/kb/articles/ainews-gemma-4-crosses-2-million-downloads-aae99147","sourceUrl":"https://www.latent.space/p/ainews-gemma-4-crosses-2-million"},{"id":"d2496898-c666-439b-91ae-af9e3386f1c5","title":"Anthropic’s Project Glasswing—restricting Claude Mythos to security researchers—sounds necessary to me","summary":"Anthropic's Project Glasswing restricts access to its powerful AI model, Claude Mythos, to security researchers to address significant cybersecurity vulnerabilities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-07","topics":["ai-security","cybersecurity","vulnerability-research","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/anthropics-project-glasswingrestricting-claude-mythos-to-sec-d2496898","sourceUrl":"https://simonwillison.net/2026/Apr/7/project-glasswing/#atom-everything"},{"id":"6a6efa1a-ec35-4fe3-ba72-0963747813c1","title":"Research: SQLite WAL Mode Across Docker Containers Sharing a Volume","summary":"The article explores SQLite's WAL mode functionality across Docker containers sharing a volume, confirming effective concurrent access and memory sharing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-07","topics":["sqlite","docker","database","concurrency"],"diataxisType":"explanation","path":"/kb/articles/research-sqlite-wal-mode-across-docker-containers-sharing-a--6a6efa1a","sourceUrl":"https://simonwillison.net/2026/Apr/7/sqlite-wal-docker-containers/#atom-everything"},{"id":"e8c2b227-8cb4-490b-8160-7563fa717efb","title":"A visual guide to getting out of a creative slump","summary":"Lenny Rachitsky shares a visual guide by Michelle Rial to help overcome creative slumps, offering practical advice and encouragement for creators.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-07","topics":["creative-process","motivation","self-improvement","visual-guide"],"diataxisType":"how-to","path":"/kb/articles/a-visual-guide-to-getting-out-of-a-creative-slump-e8c2b227","sourceUrl":"https://www.lennysnewsletter.com/p/a-visual-guide-to-getting-out-of"},{"id":"1046f7e6-6595-4fdf-9181-590650fe1b13","title":"Google AI Edge Gallery","summary":"Google's AI Edge Gallery app allows users to run Gemma 4 models on iPhones, offering features like image questioning and audio transcription, though it lacks permanent logs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-06","topics":["ai-apps","mobile-ai","image-processing"],"diataxisType":"explanation","path":"/kb/articles/google-ai-edge-gallery-1046f7e6","sourceUrl":"https://simonwillison.net/2026/Apr/6/google-ai-edge-gallery/#atom-everything"},{"id":"5e88b276-37d2-4cb1-9d0a-df33eb91b907","title":"🎙️ This week on How I AI: I gave Claude Code our entire codebase. Our customers noticed.","summary":"Al Chen discusses how he leverages Claude Code to enhance customer support by querying codebases and integrating documentation, improving response accuracy and efficiency.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-06","topics":["ai-customer-support","code-querying","workflow-automation","knowledge-management"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-i-gave-claude-code-our-entire-codebas-5e88b276","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-i-gave-claude"},{"id":"441abb85-94f5-4a2d-a85a-de3dc0b0c1cb","title":"Tool: Cleanup Claude Code Paste","summary":"Simon Willison introduces a tool for cleaning up Claude code pastes, highlighting its utility for developers working with LLMs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-06","topics":["code-cleanup","llm-developments","tools"],"diataxisType":"how-to","path":"/kb/articles/tool-cleanup-claude-code-paste-441abb85","sourceUrl":"https://simonwillison.net/2026/Apr/6/cleanup-claude-code-paste/#atom-everything"},{"id":"50777b43-071d-4f35-9878-c05272e89394","title":"I gave Claude Code our entire codebase. Our customers noticed. | Al Chen (Galileo)","summary":"Al Chen discusses using Claude Code to enhance customer support by querying multiple code repositories, emphasizing the importance of code over documentation.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-06","topics":["ai-applications","customer-support","code-querying","engineering-efficiency"],"diataxisType":"how-to","path":"/kb/articles/i-gave-claude-code-our-entire-codebase-our-customers-noticed-50777b43","sourceUrl":"https://www.lennysnewsletter.com/p/i-gave-claude-code-our-entire-codebase"},{"id":"dfb9e67b-71fd-4bde-b96e-ace2b7feaf0b","title":"Release: scan-for-secrets 0.3","summary":"The release of scan-for-secrets 0.3 introduces a new redaction option that allows users to confirm and replace matches with 'REDACTED'.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-06","topics":["software-development","python","security"],"diataxisType":"reference","path":"/kb/articles/release-scan-for-secrets-03-dfb9e67b","sourceUrl":"https://simonwillison.net/2026/Apr/6/scan-for-secrets/#atom-everything"},{"id":"5810b898-0131-42c4-9982-8151f52d967a","title":"Release: datasette-ports 0.2","summary":"The release of datasette-ports 0.2 allows users to find all running Datasette instances and their ports, now supporting uvx command execution.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-06","topics":["datasette","software-release","plugins"],"diataxisType":"reference","path":"/kb/articles/release-datasette-ports-02-5810b898","sourceUrl":"https://simonwillison.net/2026/Apr/6/datasette-ports-2/#atom-everything"},{"id":"60170551-302e-4a6d-8710-67aa0005bcc1","title":"Release: datasette-ports 0.1","summary":"The release of datasette-ports 0.1 allows users to find all running Datasette instances and their ports, simplifying management of multiple instances.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-06","topics":["datasette","development","plugins"],"diataxisType":"how-to","path":"/kb/articles/release-datasette-ports-01-60170551","sourceUrl":"https://simonwillison.net/2026/Apr/6/datasette-ports/#atom-everything"},{"id":"eb9ab2b3-40a8-4a5d-a37a-583957506f88","title":"A quote from Chengpeng Mou","summary":"Chengpeng Mou highlights significant trends in healthcare messaging, revealing insights from U.S. ChatGPT data on health insurance and hospital access.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["healthcare","chatgpt","ai-ethics"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-chengpeng-mou-eb9ab2b3","sourceUrl":"https://simonwillison.net/2026/Apr/5/chengpeng-mou/#atom-everything"},{"id":"08289a33-2962-4861-a17b-d0990b818cb9","title":"Tool: Syntaqlite Playground","summary":"Syntaqlite Playground by Lalit Maganti allows users to experiment with SQLite SQL queries through a UI, built using WebAssembly for Pyodide.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["web-assembly","sqlite","python","ai"],"diataxisType":"tutorial","path":"/kb/articles/tool-syntaqlite-playground-08289a33","sourceUrl":"https://simonwillison.net/2026/Apr/5/syntaqlite/#atom-everything"},{"id":"a105fb98-28c5-405d-8028-91698c2cfa8c","title":"Release: scan-for-secrets 0.1.1","summary":"Simon Willison announces the release of scan-for-secrets 0.1.1, highlighting its significance in LLM developments and offering a subscription for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["llm-developments","software-release","open-source"],"diataxisType":"reference","path":"/kb/articles/release-scan-for-secrets-011-a105fb98","sourceUrl":"https://simonwillison.net/2026/Apr/5/scan-for-secrets-2/#atom-everything"},{"id":"defae5d1-6cb7-4ed0-b3da-ab3aeba1fb2d","title":"Release: scan-for-secrets 0.1","summary":"Simon Willison introduces a Python tool called scan-for-secrets that helps users scan for sensitive information in log files before publishing them.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["python-tools","api-security","log-management"],"diataxisType":"how-to","path":"/kb/articles/release-scan-for-secrets-01-defae5d1","sourceUrl":"https://simonwillison.net/2026/Apr/5/scan-for-secrets-3/#atom-everything"},{"id":"960458f7-d972-4908-b877-a1a97e0e92da","title":"Release: scan-for-secrets 0.2","summary":"The scan-for-secrets 0.2 CLI tool now streams results in real-time, supports multiple directory scans, and introduces new API functions and options for enhanced usability.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["cli-tools","python-api","software-release"],"diataxisType":"reference","path":"/kb/articles/release-scan-for-secrets-02-960458f7","sourceUrl":"https://simonwillison.net/2026/Apr/5/scan-for-secrets/#atom-everything"},{"id":"bb0ae254-8461-4e3f-b94d-ad9f30ba5a2b","title":"Eight years of wanting, three months of building with AI","summary":"The article discusses the journey of building syntaqlite, highlighting the benefits and challenges of using AI in software development, particularly in design and architecture.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["ai-development","sqlite","software-engineering","agentic-engineering"],"diataxisType":"explanation","path":"/kb/articles/eight-years-of-wanting-three-months-of-building-with-ai-bb0ae254","sourceUrl":"https://simonwillison.net/2026/Apr/5/building-with-ai/#atom-everything"},{"id":"c632ad0c-ec4b-42d4-97a8-ad846a4b26d3","title":"Anthropic’s $1B to $19B growth run: how Claude became the fastest-growing AI product in history | Amol Avasare","summary":"Amol Avasare discusses Anthropic's rapid growth from $1B to $19B in ARR, sharing insights on AI-driven growth strategies and team dynamics.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-05","topics":["ai-growth","business-strategy","team-dynamics"],"diataxisType":"explanation","path":"/kb/articles/anthropics-1b-to-19b-growth-run-how-claude-became-the-fastes-c632ad0c","sourceUrl":"https://www.lennysnewsletter.com/p/anthropics-1b-to-19b-growth-run"},{"id":"a9aa5256-61ee-4a98-9137-1b826b56d66c","title":"Release: research-llm-apis 2026-04-04","summary":"Simon Willison announces a major update to his LLM Python library, enhancing its abstraction layer to accommodate new vendor features.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-05","topics":["llm","python-library","abstraction-layer","vendor-integration"],"diataxisType":"explanation","path":"/kb/articles/release-research-llm-apis-2026-04-04-a9aa5256","sourceUrl":"https://simonwillison.net/2026/Apr/5/research-llm-apis/#atom-everything"},{"id":"11fce8fa-507a-4fc1-b275-a04e5ba4d8f4","title":"A quote from Kyle Daigle","summary":"Kyle Daigle discusses the significant growth in GitHub platform activity, highlighting a surge in commits and GitHub Actions usage.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-04","topics":["github","platform-activity","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-kyle-daigle-11fce8fa","sourceUrl":"https://simonwillison.net/2026/Apr/4/kyle-daigle/#atom-everything"},{"id":"d9593374-e80d-4547-a9fb-e9df9172f258","title":"🧠 Community Wisdom: Evaluating startup equity, navigating pre-seed fundraising, MCPs vs. CLIs, Monzo’s U.S. exit, and more","summary":"This week's Community Wisdom discusses startup equity evaluation, pre-seed fundraising strategies, and insights on Monzo's U.S. exit.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-04","topics":["startup-equity","pre-seed-fundraising","monzo-us-exit"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-evaluating-startup-equity-navigating-pre-s-d9593374","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-evaluating-startup"},{"id":"e1df2f9a-c6da-4c57-bf05-527f2cf47af4","title":"A quote from Daniel Stenberg","summary":"Daniel Stenberg discusses the evolving challenges of AI in open source security, highlighting the shift from overwhelming AI-generated content to valuable security reports.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["ai-security","open-source","curl"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-daniel-stenberg-e1df2f9a","sourceUrl":"https://simonwillison.net/2026/Apr/3/daniel-stenberg/#atom-everything"},{"id":"4f4fcb91-bb65-4417-ac57-8e289ebaef0a","title":"A quote from Willy Tarreau","summary":"Willy Tarreau discusses the significant increase in kernel security reports, highlighting the challenges of duplicate findings and the need for more maintainers.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["kernel-security","software-development","ai-impacts"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-willy-tarreau-4f4fcb91","sourceUrl":"https://simonwillison.net/2026/Apr/3/willy-tarreau/#atom-everything"},{"id":"52633382-fdd8-4cc8-8d53-a67958d78a81","title":"[AINews] Good Friday","summary":"The article discusses the positive reception of Gemma 4's launch and the rapid adoption of Hermes Agents, highlighting advancements in AI tools and community reactions.","author":"Swyx","authorId":"swyx","published":"2026-04-03","topics":["gemma-4","hermes-agents","ai-tools","open-source-ai","local-inference"],"diataxisType":"explanation","path":"/kb/articles/ainews-good-friday-52633382","sourceUrl":"https://www.latent.space/p/ainews-good-friday"},{"id":"0ad9d874-c171-4fe3-9792-e3ede14f251a","title":"The cognitive impact of coding agents","summary":"Simon Willison discusses the cognitive impact of coding agents and shares insights from his podcast with Lenny Rachitsky, highlighting the popularity of short-form video content.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["coding-agents","ai-ethics","podcast-appearances"],"diataxisType":"explanation","path":"/kb/articles/the-cognitive-impact-of-coding-agents-0ad9d874","sourceUrl":"https://simonwillison.net/2026/Apr/3/cognitive-cost/#atom-everything"},{"id":"c978e84a-f774-4866-8fec-73e802467f72","title":"A quote from Greg Kroah-Hartman","summary":"Greg Kroah-Hartman discusses the evolution of AI-generated security reports, highlighting their transition from low quality to reliable outputs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["ai-security","open-source","linux-kernel"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-greg-kroah-hartman-c978e84a","sourceUrl":"https://simonwillison.net/2026/Apr/3/greg-kroah-hartman/#atom-everything"},{"id":"89da678b-f741-4e40-bc27-dab085c68469","title":"The Axios supply chain attack used individually targeted social engineering","summary":"The Axios supply chain attack involved a sophisticated social engineering campaign that targeted a maintainer, leading to the installation of a Remote Access Trojan.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["supply-chain-attacks","social-engineering","malware","open-source-software"],"diataxisType":"explanation","path":"/kb/articles/the-axios-supply-chain-attack-used-individually-targeted-soc-89da678b","sourceUrl":"https://simonwillison.net/2026/Apr/3/supply-chain-social-engineering/#atom-everything"},{"id":"795c39ae-f6b4-4e3a-8370-eb531ddff36d","title":"[AINews] Gemma 4: The best small Multimodal Open Models, dramatically better than Gemma 3 in every way","summary":"Gemma 4, Google's latest multimodal open model, outperforms its predecessor Gemma 3 with enhanced capabilities and a new Apache 2.0 license.","author":"Swyx","authorId":"swyx","published":"2026-04-03","topics":["multimodal-models","open-source-ai","machine-learning","ai-research"],"diataxisType":"explanation","path":"/kb/articles/ainews-gemma-4-the-best-small-multimodal-open-models-dramati-795c39ae","sourceUrl":"https://www.latent.space/p/ainews-gemma-4-the-best-small-multimodal"},{"id":"d9d689f0-1435-4e12-8a83-260d510bb6f6","title":"Vulnerability Research Is Cooked","summary":"The article discusses the transformative impact of frontier models on vulnerability research, highlighting how AI agents can revolutionize exploit development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["vulnerability-research","ai-agents","exploit-development","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/vulnerability-research-is-cooked-d9d689f0","sourceUrl":"https://simonwillison.net/2026/Apr/3/vulnerability-research-is-cooked/#atom-everything"},{"id":"0ded8b95-a3e7-4ab9-90d2-2eb1debdd161","title":"Research: Can JavaScript Escape a CSP Meta Tag Inside an Iframe?","summary":"JavaScript in a sandboxed iframe cannot bypass a CSP meta tag, as confirmed by tests in Chromium and Firefox, ensuring CSP policies are enforced at parse time.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-03","topics":["javascript","content-security-policy","iframes"],"diataxisType":"explanation","path":"/kb/articles/research-can-javascript-escape-a-csp-meta-tag-inside-an-ifra-0ded8b95","sourceUrl":"https://simonwillison.net/2026/Apr/3/test-csp-iframe-escape/#atom-everything"},{"id":"ad7f592a-8900-4a2a-877c-36c580057edd","title":"Marc Andreessen introspects on The Death of the Browser, Pi + OpenClaw, and Why \"This Time Is Different\"","summary":"Marc Andreessen discusses the evolution of AI, arguing that it represents an 80-year overnight success and is breaking free from past boom-bust cycles.","author":"Swyx","authorId":"swyx","published":"2026-04-03","topics":["ai-evolution","open-source-ai","software-architecture","ai-agents","future-of-coding"],"diataxisType":"explanation","path":"/kb/articles/marc-andreessen-introspects-on-the-death-of-the-browser-pi-o-ad7f592a","sourceUrl":"https://www.latent.space/p/pmarca"},{"id":"51fc7d31-458e-4c40-ab6d-dc6fcd7f010a","title":"[AINews] A quiet April Fools","summary":"The article discusses notable AI model releases and the impact of an accidental code leak from Anthropic's Claude, highlighting operational issues and community responses.","author":"Swyx","authorId":"swyx","published":"2026-04-02","topics":["ai-models","code-leak","open-source","community-response"],"diataxisType":"reference","path":"/kb/articles/ainews-a-quiet-april-fools-51fc7d31","sourceUrl":"https://www.latent.space/p/ainews-a-quiet-april-fools"},{"id":"d424d881-1208-43b9-9554-d5dabac1c535","title":"Highlights from my conversation about agentic engineering on Lenny’s Podcast","summary":"Simon Willison discusses the impact of AI on software engineering in his conversation on Lenny's Podcast, highlighting the shift in coding practices and testing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-02","topics":["ai-engineering","automation","software-development","podcast-highlights"],"diataxisType":"explanation","path":"/kb/articles/highlights-from-my-conversation-about-agentic-engineering-on-d424d881","sourceUrl":"https://simonwillison.net/2026/Apr/2/lennys-podcast/#atom-everything"},{"id":"f0f15fdb-451a-4969-b1c5-d133199f21c2","title":"Release: llm-gemini 0.30","summary":"Simon Willison announces the release of llm-gemini 0.30, highlighting its features and inviting readers to support his work.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-02","topics":["llm-development","gemini-release","monthly-briefing"],"diataxisType":"reference","path":"/kb/articles/release-llm-gemini-030-f0f15fdb","sourceUrl":"https://simonwillison.net/2026/Apr/2/llm-gemini/#atom-everything"},{"id":"0ad233f1-61cc-431a-85bf-9a0cd469cc3e","title":"Gemma 4: Byte for byte, the most capable open models","summary":"Gemma 4 introduces four new vision-capable LLMs from Google DeepMind, emphasizing efficiency and multi-modal capabilities in AI models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-02","topics":["llms","google-deepmind","ai-research","multi-modal"],"diataxisType":"explanation","path":"/kb/articles/gemma-4-byte-for-byte-the-most-capable-open-models-0ad233f1","sourceUrl":"https://simonwillison.net/2026/Apr/2/gemma-4/#atom-everything"},{"id":"24b5ce76-9d62-4a06-9bd4-4818be8cf96c","title":"Moonlake: Causal World Models should be Multimodal, Interactive, and Efficient — with Chris Manning and Fan-yun Sun","summary":"The article discusses Moonlake's innovative approach to World Models, emphasizing the need for multimodal, interactive, and efficient models in AI, featuring insights from Chris Manning and Fan-yun Sun.","author":"Swyx","authorId":"swyx","published":"2026-04-02","topics":["world-models","ai-research","multimodal-learning","interactive-systems"],"diataxisType":"explanation","path":"/kb/articles/moonlake-causal-world-models-should-be-multimodal-interactiv-24b5ce76","sourceUrl":"https://www.latent.space/p/moonlake"},{"id":"8a45e903-1595-4e72-a676-3c5c431c7324","title":"An AI state of the union: We’ve passed the inflection point, dark factories are coming, and automation timelines | Simon Willison","summary":"Simon Willison discusses the evolution of AI coding agents, the risks for mid-career engineers, and the future of automation in software development.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-02","topics":["ai-coding","automation","agentic-engineering","software-development"],"diataxisType":"explanation","path":"/kb/articles/an-ai-state-of-the-union-weve-passed-the-inflection-point-da-8a45e903","sourceUrl":"https://www.lennysnewsletter.com/p/an-ai-state-of-the-union"},{"id":"da845929-c5d0-4ea6-95ff-732fd60b940e","title":"March 2026 sponsors-only newsletter","summary":"The March 2026 sponsors-only newsletter by Simon Willison covers various topics including engineering patterns, model releases, and supply chain attacks.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-02","topics":["agentic-engineering","streaming-experts","supply-chain-attacks","model-releases","museums"],"diataxisType":"reference","path":"/kb/articles/march-2026-sponsors-only-newsletter-da845929","sourceUrl":"https://simonwillison.net/2026/Apr/2/march-newsletter/#atom-everything"},{"id":"e67b66fe-f3e8-4fb9-8c6e-e59be681f948","title":"Release: datasette-enrichments-llm 0.2a0","summary":"Simon Willison announces the release of datasette-enrichments-llm 0.2a0, highlighting its features and inviting sponsorship for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["datasette","llm","release-notes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-enrichments-llm-02a0-e67b66fe","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-enrichments-llm/#atom-everything"},{"id":"4a95457e-457a-48bd-94d6-6ca956cbedde","title":"A quote from Soohoon Choi","summary":"Soohoon Choi argues that economic incentives will drive AI models to produce better, more maintainable code, as markets favor quality over sloppiness.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["ai-development","software-engineering","economic-incentives"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-soohoon-choi-4a95457e","sourceUrl":"https://simonwillison.net/2026/Apr/1/soohoon-choi/#atom-everything"},{"id":"320c1f29-2cc5-483a-92f6-9dfb7ca20433","title":"Release: datasette-llm 0.1a6","summary":"The release of datasette-llm 0.1a6 introduces a new LLM integration plugin, simplifying model ID management and improving Python API documentation.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["llm-integration","python-api","datasette"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a6-320c1f29","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-llm-2/#atom-everything"},{"id":"8f4b0133-4464-42b9-a0fc-d81cab621f06","title":"Release: datasette-llm 0.1a5","summary":"The release of datasette-llm 0.1a5 introduces a plugin hook that tracks prompts in tool call loops, enhancing LLM integration for plugins.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["llm-integration","plugin-development","datasette"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a5-8f4b0133","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-llm/#atom-everything"},{"id":"6cd83dcd-729f-4fee-891f-17decc0271bf","title":"Release: datasette-enrichments-llm 0.2a1","summary":"Simon Willison announces the release of datasette-enrichments-llm 0.2a1, highlighting its features and inviting sponsorship for curated updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["datasette","llm","enrichments","monthly-briefing"],"diataxisType":"reference","path":"/kb/articles/release-datasette-enrichments-llm-02a1-6cd83dcd","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-enrichments-llm-2/#atom-everything"},{"id":"3f1e5b71-891d-4eae-9307-8fa3b5e62fd2","title":"[AINews] The Claude Code Source Leak","summary":"The Claude Code source leak reveals significant architectural details about Anthropic's AI system, sparking public reverse-engineering and discussions on agent design.","author":"Swyx","authorId":"swyx","published":"2026-04-01","topics":["ai-agents","source-code-leak","reverse-engineering","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-claude-code-source-leak-3f1e5b71","sourceUrl":"https://www.latent.space/p/ainews-the-claude-code-source-leak"},{"id":"f91edd89-a808-4c77-98be-5ebf1300c90e","title":"Listen: OpenClaw: A power-user's guide to the most powerful personal AI tool since ChatGPT","summary":"Lenny Rachitsky interviews Claire Vo about her experience with OpenClaw, detailing its setup, workflows, and how to effectively manage multiple AI agents.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-04-01","topics":["ai-tools","openclaw","productivity","automation"],"diataxisType":"how-to","path":"/kb/articles/listen-openclaw-a-power-users-guide-to-the-most-powerful-per-f91edd89","sourceUrl":"https://www.lennysnewsletter.com/p/listen-openclaw-a-power-users-guide"},{"id":"3fb8f61d-f673-4885-a21c-7494c2b2c2d2","title":"Release: datasette-llm-usage 0.2a0","summary":"The release of datasette-llm-usage 0.2a0 includes the removal of certain features and introduces new logging capabilities for prompts and responses.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["datasette","llm-usage","logging","model-configuration"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-usage-02a0-3fb8f61d","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-llm-usage/#atom-everything"},{"id":"5c4dd7f1-4f59-41ce-93ba-e4684bbbb861","title":"Release: datasette-extract 0.3a0","summary":"The release of datasette-extract 0.3a0 introduces the ability to import unstructured data into structured tables, utilizing datasette-llm for model management.","author":"Simon Willison","authorId":"simon-willison","published":"2026-04-01","topics":["data-import","datasette","machine-learning"],"diataxisType":"reference","path":"/kb/articles/release-datasette-extract-03a0-5c4dd7f1","sourceUrl":"https://simonwillison.net/2026/Apr/1/datasette-extract/#atom-everything"},{"id":"1521fdb4-bfb4-4547-a0c5-722b3db3c19b","title":"Supply Chain Attack on Axios Pulls Malicious Dependency from npm","summary":"A supply chain attack on Axios introduced a malicious dependency from npm, compromising user credentials and installing a remote access trojan.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-31","topics":["supply-chain-attacks","npm-security","axios"],"diataxisType":"explanation","path":"/kb/articles/supply-chain-attack-on-axios-pulls-malicious-dependency-from-1521fdb4","sourceUrl":"https://simonwillison.net/2026/Mar/31/supply-chain-attack-on-axios/#atom-everything"},{"id":"0e6a80ac-4013-4dfd-9097-10c943c97d96","title":"OpenClaw: The complete guide to building, training, and living with your personal AI agent","summary":"Lenny Rachitsky introduces OpenClaw, a powerful open-source personal AI assistant, and provides a comprehensive guide on setting it up and utilizing its features.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-31","topics":["ai-agents","open-source","personal-assistant","automation"],"diataxisType":"how-to","path":"/kb/articles/openclaw-the-complete-guide-to-building-training-and-living--0e6a80ac","sourceUrl":"https://www.lennysnewsletter.com/p/openclaw-the-complete-guide-to-building"},{"id":"cd0d8dfe-55d9-4cfb-9e74-64df6f7d587c","title":"Release: datasette-llm 0.1a4","summary":"Simon Willison announces the release of datasette-llm 0.1a4, highlighting its features and inviting readers to support his work.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-31","topics":["datasette","llm","release-notes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a4-cd0d8dfe","sourceUrl":"https://simonwillison.net/2026/Mar/31/datasette-llm/#atom-everything"},{"id":"f08e2d6c-3ffa-4d03-997c-88b08f6b681d","title":"Release: llm 0.30","summary":"The release of llm 0.30 introduces a new plugin hook for model registration and enhances documentation with added docstrings.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-31","topics":["large-language-models","plugin-development","documentation"],"diataxisType":"reference","path":"/kb/articles/release-llm-030-f08e2d6c","sourceUrl":"https://simonwillison.net/2026/Mar/31/llm/#atom-everything"},{"id":"68c4826d-4c5f-476c-b1e7-9e482c2b818f","title":"Release: llm-all-models-async 0.1","summary":"The release of llm-all-models-async 0.1 introduces async versions of models from LLM plugins, enabling better integration with API-backed models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-31","topics":["llm-plugins","async-models","datasette","api-integration"],"diataxisType":"reference","path":"/kb/articles/release-llm-all-models-async-01-68c4826d","sourceUrl":"https://simonwillison.net/2026/Mar/31/llm-all-models-async/#atom-everything"},{"id":"a90df1e1-7a73-4109-a039-1dc898758fde","title":"Release: llm-echo 0.4","summary":"Simon Willison announces the release of llm-echo 0.4, highlighting its features and inviting sponsorship for a curated email digest.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-31","topics":["llm-echo","sponsorship","monthly-briefing"],"diataxisType":"reference","path":"/kb/articles/release-llm-echo-04-a90df1e1","sourceUrl":"https://simonwillison.net/2026/Mar/31/llm-echo/#atom-everything"},{"id":"7f163b66-e1f8-47e1-b9a9-f450ad3f048d","title":"[AINews] The Last 4 Jobs in Tech","summary":"The article discusses the evolving roles in tech due to AI advancements, highlighting new job types and the rise of open-source agent ecosystems.","author":"Swyx","authorId":"swyx","published":"2026-03-31","topics":["ai-engineering","open-source-agents","tech-jobs","ai-ecosystem"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-last-4-jobs-in-tech-7f163b66","sourceUrl":"https://www.latent.space/p/ainews-the-last-4-jobs-in-tech"},{"id":"5dea7f7c-668c-4ef1-80eb-97a6c623fda1","title":"Release: datasette-llm 0.1a3","summary":"Simon Willison announces the release of datasette-llm 0.1a3, highlighting its features and inviting support for ongoing updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-30","topics":["datasette","llm","release-notes"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a3-5dea7f7c","sourceUrl":"https://simonwillison.net/2026/Mar/30/datasette-llm/#atom-everything"},{"id":"23a9721a-157a-429f-b767-39c8a9bbe8fb","title":"Mistral: Voxtral TTS, Forge, Leanstral, & what's next for Mistral 4 — w/ Pavan Kumar Reddy & Guillaume Lample","summary":"The article discusses Mistral's recent advancements in TTS technology, featuring insights from Pavan Kumar Reddy and Guillaume Lample on their new Voxtral TTS model.","author":"Swyx","authorId":"swyx","published":"2026-03-30","topics":["text-to-speech","ai-research","audio-models","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/mistral-voxtral-tts-forge-leanstral-whats-next-for-mistral-4-23a9721a","sourceUrl":"https://www.latent.space/p/voxtral"},{"id":"57346664-c395-4d7c-99f0-5cc6eccdc9cb","title":"Release: datasette-files 0.1a3","summary":"The release of datasette-files 0.1a3 introduces new configuration options and a file picker UI, enhancing integration with other plugins.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-30","topics":["datasette","plugin-development","file-management"],"diataxisType":"reference","path":"/kb/articles/release-datasette-files-01a3-57346664","sourceUrl":"https://simonwillison.net/2026/Mar/30/datasette-files/#atom-everything"},{"id":"2f86da03-f4a8-4d64-9802-70c1973c8ad0","title":"Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer","summary":"Mr. Chatterbox is a Victorian-era language model trained on public domain texts, offering a unique but limited conversational experience.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-30","topics":["language-models","victorian-literature","ai-development"],"diataxisType":"how-to","path":"/kb/articles/mr-chatterbox-is-a-weak-victorian-era-ethically-trained-mode-2f86da03","sourceUrl":"https://simonwillison.net/2026/Mar/30/mr-chatterbox/#atom-everything"},{"id":"44847dd6-b5fa-419b-bd92-fad84ff97dd3","title":"🎙️ This week on How I AI: How Stripe built “minions”—AI coding agents that ship 1,300 PRs per week + How to turn Claude Code into your personal life operating system","summary":"This article discusses how Stripe developed AI coding agents, termed 'minions', that automate pull requests and enhance developer productivity, alongside insights on using Claude Code for personal management.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-30","topics":["ai-agents","developer-productivity","automation","personal-organization"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-how-stripe-built-minionsai-coding-age-44847dd6","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-how-stripe"},{"id":"af54c792-1c6f-44d2-a2a4-83a6fbac0a94","title":"A quote from Georgi Gerganov","summary":"Georgi Gerganov discusses the challenges of using local models, highlighting issues with model templates and the fragility of the components involved.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-30","topics":["local-models","ai-challenges","model-inference"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-georgi-gerganov-af54c792","sourceUrl":"https://simonwillison.net/2026/Mar/30/georgi-gerganov/#atom-everything"},{"id":"7f4b29cb-65d9-4151-a9e6-91a8d657b7fe","title":"How to turn Claude Code into your personal life operating system | Hilary Gridley","summary":"Hilary Gridley shares her approach to personal productivity using Claude Code, emphasizing simplicity and AI's ability to learn preferences over time.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-30","topics":["personal-productivity","ai-tools","automation","task-management"],"diataxisType":"how-to","path":"/kb/articles/how-to-turn-claude-code-into-your-personal-life-operating-sy-7f4b29cb","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-turn-claude-code-into-your"},{"id":"9767b320-a1b1-4440-9ef7-6f3a053ecada","title":"Release: llm-mrchatterbox 0.1","summary":"Simon Willison announces the release of llm-mrchatterbox 0.1, highlighting its features and inviting sponsorship for curated LLM updates.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-30","topics":["llm-release","sponsorship","email-digest"],"diataxisType":"reference","path":"/kb/articles/release-llm-mrchatterbox-01-9767b320","sourceUrl":"https://simonwillison.net/2026/Mar/30/llm-mrchatterbox-2/#atom-everything"},{"id":"9184893d-0718-4db6-ba72-7ea0b25585c9","title":"Tool: Python Vulnerability Lookup","summary":"The Python Vulnerability Lookup tool allows users to search for known security vulnerabilities in Python packages using a pyproject.toml or requirements.txt file.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-29","topics":["python","security","vulnerabilities","osv-dev"],"diataxisType":"how-to","path":"/kb/articles/tool-python-vulnerability-lookup-9184893d","sourceUrl":"https://simonwillison.net/2026/Mar/29/python-vulnerability-lookup/#atom-everything"},{"id":"9ecc2a1a-d2b3-4d75-8e4c-243a5c405032","title":"Tool: Pretext — Under the Hood","summary":"Simon Willison discusses the inner workings of the Pretext tool in this article, highlighting its features and benefits for LLM developments.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-29","topics":["llm-developments","tool-overview","sponsorship"],"diataxisType":"explanation","path":"/kb/articles/tool-pretext-under-the-hood-9ecc2a1a","sourceUrl":"https://simonwillison.net/2026/Mar/29/pretext-explainer/#atom-everything"},{"id":"6c6993d4-f19a-43e5-974a-7a134d6381c8","title":"From skeptic to true believer: How OpenClaw changed my life | Claire Vo","summary":"Claire Vo shares her transformative journey from skepticism to belief in OpenClaw, detailing its practical applications in her life and offering insights on AI agents.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-29","topics":["ai-agents","productivity","podcasting","technology"],"diataxisType":"how-to","path":"/kb/articles/from-skeptic-to-true-believer-how-openclaw-changed-my-life-c-6c6993d4","sourceUrl":"https://www.lennysnewsletter.com/p/how-openclaw-changed-my-life-claire-vo"},{"id":"635e0854-e0dc-4fa2-b569-3e828c2d70ff","title":"Pretext","summary":"Pretext is a new browser library that efficiently calculates the height of line-wrapped text without DOM manipulation, enhancing text rendering in applications.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-29","topics":["browser-library","text-rendering","performance-optimization"],"diataxisType":"explanation","path":"/kb/articles/pretext-635e0854","sourceUrl":"https://simonwillison.net/2026/Mar/29/pretext/#atom-everything"},{"id":"234d7d05-3cc6-46e7-8706-d9838feb377a","title":"[AINews] H100 prices are melting *UP*","summary":"The H100 rental market has seen a significant increase in value due to a chip shortage and improved reasoning models, impacting data center business models.","author":"Swyx","authorId":"swyx","published":"2026-03-28","topics":["h100-prices","chip-shortage","data-centers","ai-models","rental-market"],"diataxisType":"explanation","path":"/kb/articles/ainews-h100-prices-are-melting-up-234d7d05","sourceUrl":"https://www.latent.space/p/ainews-h100-prices-are-melting-up"},{"id":"7d8b84f7-4a62-443b-8b3d-93b2a945bf77","title":"A quote from Matt Webb","summary":"Matt Webb discusses the importance of agentic coding and architecture in developing AI agents that solve coding problems efficiently and adaptively.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-28","topics":["ai-agents","coding-architecture","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-matt-webb-7d8b84f7","sourceUrl":"https://simonwillison.net/2026/Mar/28/matt-webb/#atom-everything"},{"id":"324c0ef3-e130-4ea9-9c00-910fc5a67b94","title":"🧠 Community Wisdom: When AI velocity outpaces your product strategy, when your estimates keep slipping, one day in San Francisco, pairing Claude Code with Codex, and more","summary":"This week's Community Wisdom discusses the challenges of AI velocity in product strategy and highlights key conversations from the Slack community.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-28","topics":["ai-velocity","product-strategy","community-insights"],"diataxisType":"explanation","path":"/kb/articles/-community-wisdom-when-ai-velocity-outpaces-your-product-str-324c0ef3","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-when-ai-velocity"},{"id":"fe678ac4-9d8f-4036-9083-7523342d35aa","title":"A quote from Richard Fontana","summary":"Richard Fontana discusses the licensing situation of chardet 7.0.0, stating there's no basis for it to be released under the LGPL.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-27","topics":["licensing","chardet","open-source"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-richard-fontana-fe678ac4","sourceUrl":"https://simonwillison.net/2026/Mar/27/richard-fontana/#atom-everything"},{"id":"3ab6467f-eb60-4dcc-9c3c-8d0249132a73","title":"We Rewrote JSONata with AI in a Day, Saved $500K/Year","summary":"The article discusses how the Reco team rewrote JSONata using AI in just one day, resulting in significant cost savings and improved performance.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-27","topics":["ai-development","jsonata","cost-savings","software-engineering"],"diataxisType":"explanation","path":"/kb/articles/we-rewrote-jsonata-with-ai-in-a-day-saved-500kyear-3ab6467f","sourceUrl":"https://simonwillison.net/2026/Mar/27/vine-porting-jsonata/#atom-everything"},{"id":"3b7375ca-f963-454e-b6c6-31b0c6f7d72a","title":"[AINews] Everything is CLI","summary":"The launch of Projects.dev by Stripe highlights the growing trend of command-line interfaces (CLIs) in simplifying backend service setups for agents.","author":"Swyx","authorId":"swyx","published":"2026-03-27","topics":["cli","stripe","ai-agents","backend-services"],"diataxisType":"explanation","path":"/kb/articles/ainews-everything-is-cli-3b7375ca","sourceUrl":"https://www.latent.space/p/ainews-everything-is-cli"},{"id":"3dde9a3b-dbd1-45a5-8227-d5889bc3a7b3","title":"Vibe coding SwiftUI apps is a lot of fun","summary":"Simon Willison shares his experience of vibe coding SwiftUI apps on his new MacBook Pro, highlighting two tools he built for monitoring network and GPU performance.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-27","topics":["swiftui","vibe-coding","macos-apps","performance-monitoring"],"diataxisType":"how-to","path":"/kb/articles/vibe-coding-swiftui-apps-is-a-lot-of-fun-3dde9a3b","sourceUrl":"https://simonwillison.net/2026/Mar/27/vibe-coding-swiftui/#atom-everything"},{"id":"cd1d0a7c-4ee6-4dbb-871c-4d5a594dc331","title":"Release: datasette-showboat 0.1a2","summary":"Simon Willison announces the release of datasette-showboat 0.1a2, highlighting its features and inviting sponsorship for a curated email digest.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-27","topics":["datasette","software-release","llm-developments"],"diataxisType":"reference","path":"/kb/articles/release-datasette-showboat-01a2-cd1d0a7c","sourceUrl":"https://simonwillison.net/2026/Mar/27/datasette-showboat/#atom-everything"},{"id":"fdd6a39a-838c-426f-8678-596fe3d9a829","title":"Claude Code auto mode: a safer way to skip permissions","summary":"By default, Claude Code asks users for approval before running commands or modifying files. By default, Claude Code asks users for approval before running commands or modifying files. This keeps users safe, but it also means a lot of clicking \"approve.\" Over time that leads to approval fatigue, wher","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-03-26","topics":["Software Development Practices","User Experience Design","Machine Learning Models","Security Measures","AI Safety"],"diataxisType":"explanation","path":"/kb/articles/claude-code-auto-mode-a-safer-way-to-skip-permissions-fdd6a39a","sourceUrl":"https://www.anthropic.com/engineering/claude-code-auto-mode"},{"id":"8508c5cb-620c-4585-b2c1-ec9d8587b539","title":"[AINews] The Biggest Claude Launch of All Time","summary":"The article discusses the significant launch of Claude Cowork Dispatch, highlighting its unprecedented reception and impact in the AI space.","author":"Swyx","authorId":"swyx","published":"2026-03-26","topics":["ai-launches","claude","anthropic-ai"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-biggest-claude-launch-of-all-time-8508c5cb","sourceUrl":"https://www.latent.space/p/ainews-the-biggest-claude-launch"},{"id":"912e723c-85b2-4170-8195-2da8a813c3e2","title":"Quantization from the ground up","summary":"Sam Rose's interactive essay explains the quantization of Large Language Models, highlighting the importance of outlier values and their impact on model quality.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-26","topics":["quantization","large-language-models","model-accuracy","floating-point-numbers"],"diataxisType":"explanation","path":"/kb/articles/quantization-from-the-ground-up-912e723c","sourceUrl":"https://simonwillison.net/2026/Mar/26/quantization-from-the-ground-up/#atom-everything"},{"id":"a6fab7d7-a44d-4579-8c8d-993228925800","title":"My minute-by-minute response to the LiteLLM malware attack","summary":"Simon Willison details his immediate response to the LiteLLM malware attack reported to PyPI, highlighting the use of Claude transcripts for analysis.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-26","topics":["malware","pypi","security","software-vulnerability"],"diataxisType":"how-to","path":"/kb/articles/my-minute-by-minute-response-to-the-litellm-malware-attack-a6fab7d7","sourceUrl":"https://simonwillison.net/2026/Mar/26/response-to-the-litellm-malware-attack/#atom-everything"},{"id":"b451738f-1699-4278-8bdf-c5e3fd662997","title":"Release: datasette-files-s3 0.1a1","summary":"This release of datasette-files-s3 introduces a backend for storing and retrieving files using S3, with a new mechanism for fetching S3 configuration from a URL.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-25","topics":["datasette","s3","file-storage"],"diataxisType":"reference","path":"/kb/articles/release-datasette-files-s3-01a1-b451738f","sourceUrl":"https://simonwillison.net/2026/Mar/25/datasette-files-s3/#atom-everything"},{"id":"944c4313-0a9d-496f-9a0f-611b111ae2e3","title":"How Stripe built “minions”—AI coding agents that ship 1,300 PRs weekly from Slack reactions | Steve Kaliski (Stripe engineer)","summary":"Steve Kaliski discusses how Stripe's AI coding agents, known as 'minions', automate the creation of 1,300 pull requests weekly, enhancing developer productivity and efficiency.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-25","topics":["ai-coding-agents","developer-productivity","cloud-development","machine-payments"],"diataxisType":"how-to","path":"/kb/articles/how-stripe-built-minionsai-coding-agents-that-ship-1300-prs--944c4313","sourceUrl":"https://www.lennysnewsletter.com/p/how-stripe-built-minionsai-coding"},{"id":"1a0b581b-3bc0-4446-b293-6574a0cc5a43","title":"Release: datasette-llm 0.1a1","summary":"The release of datasette-llm 0.1a1 introduces a plugin for LLM integration, allowing other Datasette plugins to utilize various models for specific purposes.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-25","topics":["llm-integration","datasette-plugins","data-enrichment"],"diataxisType":"reference","path":"/kb/articles/release-datasette-llm-01a1-1a0b581b","sourceUrl":"https://simonwillison.net/2026/Mar/25/datasette-llm/#atom-everything"},{"id":"62e9956a-51c9-4145-88d6-059b9526f335","title":"Thoughts on slowing the fuck down","summary":"Simon Willison discusses the dangers of rapid code production driven by agents, emphasizing the need for discipline and thoughtful development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-25","topics":["agentic-engineering","cognitive-debt","software-development"],"diataxisType":"explanation","path":"/kb/articles/thoughts-on-slowing-the-fuck-down-62e9956a","sourceUrl":"https://simonwillison.net/2026/Mar/25/thoughts-on-slowing-the-fuck-down/#atom-everything"},{"id":"4534c43e-4c06-460e-95a6-c77f2bce9969","title":"[AINews] Apple's War on Slop","summary":"The article discusses the ongoing challenges faced by traditional app stores like Apple's in the wake of AI advancements and the rise of new app distribution paradigms.","author":"Swyx","authorId":"swyx","published":"2026-03-25","topics":["app-store","ai-advancements","software-distribution","entrepreneurship"],"diataxisType":"explanation","path":"/kb/articles/ainews-apples-war-on-slop-4534c43e","sourceUrl":"https://www.latent.space/p/ainews-apples-war-on-slop"},{"id":"0b0a1d72-1221-4f7f-8b1a-6327730ad271","title":"Harness design for long-running application development","summary":"Over the past several months I’ve been working on two interconnected problems: getting Claude to produce high-quality frontend designs, and getting it to build complete applications without human inte Written by Prithvi Rajasekaran, a member of our\nLabs\nteam.\nOver the past several months I’ve been w","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-03-25","topics":["AI in Software Development","Generative Adversarial Networks","Autonomous Coding"],"diataxisType":"explanation","path":"/kb/articles/harness-design-for-long-running-application-development-0b0a1d72","sourceUrl":"https://www.anthropic.com/engineering/harness-design-long-running-apps"},{"id":"93420151-af71-4292-b8da-4363b16d4410","title":"LiteLLM Hack: Were You One of the 47,000?","summary":"The LiteLLM hack resulted in nearly 47,000 downloads of compromised packages, with many dependent packages failing to pin versions appropriately.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-25","topics":["security","supply-chain","python"],"diataxisType":"explanation","path":"/kb/articles/litellm-hack-were-you-one-of-the-47000-93420151","sourceUrl":"https://simonwillison.net/2026/Mar/25/litellm-hack/#atom-everything"},{"id":"b36c4b11-cd16-49ab-bdca-b3aed833ac28","title":"Package Managers Need to Cool Down","summary":"The article discusses the need for dependency cooldowns in package managers to mitigate supply chain attacks, highlighting recent updates across various tools.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-24","topics":["package-managers","dependency-management","supply-chain-security"],"diataxisType":"explanation","path":"/kb/articles/package-managers-need-to-cool-down-b36c4b11","sourceUrl":"https://simonwillison.net/2026/Mar/24/package-managers-need-to-cool-down/#atom-everything"},{"id":"de999191-f300-47fa-9348-8401b4336bfa","title":"Auto mode for Claude Code","summary":"Claude Code introduces auto mode, a new permissions system that makes decisions on user behalf while implementing safeguards to monitor actions.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-24","topics":["ai-agents","permissions","safeguards","coding-safety"],"diataxisType":"explanation","path":"/kb/articles/auto-mode-for-claude-code-de999191","sourceUrl":"https://simonwillison.net/2026/Mar/24/auto-mode-for-claude-code/#atom-everything"},{"id":"a81c2053-3857-4b8e-be6d-6f6416644b19","title":"State of the product job market in early 2026","summary":"The job market for product managers is thriving in early 2026, with record openings and a surge in AI roles, indicating a positive trend in tech employment.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-24","topics":["product-management","ai-roles","tech-job-market","remote-work"],"diataxisType":"explanation","path":"/kb/articles/state-of-the-product-job-market-in-early-2026-a81c2053","sourceUrl":"https://www.lennysnewsletter.com/p/state-of-the-product-job-market-in-372"},{"id":"bd816c12-0899-4ae9-8fca-adc5e1245860","title":"Malicious litellm_init.pth in litellm 1.82.8 — credential stealer","summary":"The LiteLLM v1.82.8 package on PyPI was compromised with a credential stealer hidden in a file, allowing secrets to be harvested upon installation.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-24","topics":["supply-chain-attacks","credential-theft","python-packages"],"diataxisType":"explanation","path":"/kb/articles/malicious-litellm_initpth-in-litellm-1828-credential-stealer-bd816c12","sourceUrl":"https://simonwillison.net/2026/Mar/24/malicious-litellm/#atom-everything"},{"id":"985dc0b3-a6dd-4b78-bf03-c094edfc6771","title":"[AINews] Dreamer joins Meta Superintelligence Labs — 9 month retro of Personal Superintelligence","summary":"Dreamer has joined Meta Superintelligence Labs shortly after their podcast, marking a significant step in consumer AI development.","author":"Swyx","authorId":"swyx","published":"2026-03-24","topics":["ai-agents","superintelligence","meta","consumer-ai"],"diataxisType":"explanation","path":"/kb/articles/ainews-dreamer-joins-meta-superintelligence-labs-9-month-ret-985dc0b3","sourceUrl":"https://www.latent.space/p/ainews-dreamer-joins-meta-superintelligence"},{"id":"8d6d3e47-57ac-4bd7-87cb-12adabdfb2ec","title":"A quote from Christopher Mims","summary":"Christopher Mims critiques the idea of giving AI total control over personal lives, suggesting it will seem foolish in hindsight.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-24","topics":["ai","technology","critique"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-christopher-mims-8d6d3e47","sourceUrl":"https://simonwillison.net/2026/Mar/24/christopher-mims/#atom-everything"},{"id":"398fa22b-a327-4abb-934d-5847eccf2c25","title":"State of the product job market in early 2026","summary":"The early 2026 product job market shows a surprising optimism with rising PM and engineering roles, despite ongoing layoffs and a decline in remote work opportunities.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-24","topics":["product-management","tech-hiring","ai-jobs","remote-work","job-market-trends"],"diataxisType":"explanation","path":"/kb/articles/state-of-the-product-job-market-in-early-2026-398fa22b","sourceUrl":"https://www.lennysnewsletter.com/p/state-of-the-product-job-market-in-ee9"},{"id":"b18788c6-a955-40e0-b5c3-b642b1ccc27a","title":"🔬Why There Is No \"AlphaFold for Materials\" — AI for Materials Discovery with Heather Kulik","summary":"Prof. Heather Kulik discusses the challenges and potential of AI in materials science, emphasizing the need for high-quality datasets and human intuition.","author":"Swyx","authorId":"swyx","published":"2026-03-24","topics":["ai-for-science","materials-discovery","data-driven-modeling","computational-tools"],"diataxisType":"explanation","path":"/kb/articles/why-there-is-no-alphafold-for-materials-ai-for-materials-dis-b18788c6","sourceUrl":"https://www.latent.space/p/materials"},{"id":"7e5ada0d-63f5-4b10-8dde-118019a0e6f2","title":"Streaming experts","summary":"The article discusses advancements in streaming experts for large AI models, highlighting recent experiments by Dan Woods and others using limited RAM.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-24","topics":["streaming-experts","mixture-of-experts","ai-optimization"],"diataxisType":"explanation","path":"/kb/articles/streaming-experts-7e5ada0d","sourceUrl":"https://simonwillison.net/2026/Mar/24/streaming-experts/#atom-everything"},{"id":"5d9b8291-c088-4f47-82e1-6d55b0ee28ea","title":"A quote from David Abram","summary":"Simon Willison reflects on the challenges of software development, emphasizing the irreplaceable human elements in decision-making and system understanding.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-23","topics":["software-development","llms","decision-making"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-david-abram-5d9b8291","sourceUrl":"https://simonwillison.net/2026/Mar/23/david-abram/#atom-everything"},{"id":"c255be60-4eb6-4933-88c4-6296518cb2b5","title":"Beats now have notes","summary":"Simon Willison introduces a new feature called 'beats' on his blog, allowing for annotations that enhance the display of external content.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-23","topics":["blogging","content-creation","web-development"],"diataxisType":"explanation","path":"/kb/articles/beats-now-have-notes-c255be60","sourceUrl":"https://simonwillison.net/2026/Mar/23/beats-now-have-notes/#atom-everything"},{"id":"fa10205d-aa50-47b8-9993-4fbcee768cc5","title":"A quote from Neurotica","summary":"Simon Willison discusses the concept of 'slop' in content creation, emphasizing the value of time in the creative process.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-23","topics":["ai-ethics","generative-ai","slop"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-neurotica-fa10205d","sourceUrl":"https://simonwillison.net/2026/Mar/23/neurotica/#atom-everything"},{"id":"57f4a349-ada9-401e-ba0d-bc5106ad2362","title":"How Microsoft’s AI VP automates everything with Warp | Marco Casalaina","summary":"Marco Casalaina from Microsoft showcases how AI tools like Warp can automate administrative tasks, enhancing productivity and workflow efficiency.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-23","topics":["ai-automation","microsoft-365","warp","azure-resources","chatgpt"],"diataxisType":"how-to","path":"/kb/articles/how-microsofts-ai-vp-automates-everything-with-warp-marco-ca-57f4a349","sourceUrl":"https://www.lennysnewsletter.com/p/how-microsofts-ai-vp-automates-everything"},{"id":"6715930b-1c57-47a7-9924-b632222cfe44","title":"Release: datasette-files 0.1a2","summary":"The release of datasette-files 0.1a2 introduces a plugin for uploading files directly into a Datasette instance, featuring new functionalities and enhancements.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-23","topics":["datasette","file-upload","plugins"],"diataxisType":"reference","path":"/kb/articles/release-datasette-files-01a2-6715930b","sourceUrl":"https://simonwillison.net/2026/Mar/23/datasette-files/#atom-everything"},{"id":"46856582-850a-4160-9cab-12bb32130783","title":"🎙️ This week on How I AI: How Microsoft's AI VP automates everything with Warp","summary":"Marco Casalaina, Microsoft's AI VP, discusses using Warp and AI tools to automate tasks, enhancing productivity and simplifying workflows.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-23","topics":["ai-automation","micro-agents","workflow-optimization","cli-tools"],"diataxisType":"explanation","path":"/kb/articles/-this-week-on-how-i-ai-how-microsofts-ai-vp-automates-everyt-46856582","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-how-microsofts"},{"id":"40a27698-8707-49d0-9b0e-af548dc312da","title":"Research: Starlette 1.0 skill","summary":"Starlette 1.0 Skill provides a guide for building web applications using the lightweight ASGI framework, featuring a demo task management app.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-23","topics":["web-development","asgi-frameworks","starlette","async-programming"],"diataxisType":"tutorial","path":"/kb/articles/research-starlette-10-skill-40a27698","sourceUrl":"https://simonwillison.net/2026/Mar/23/starlette-1-skill/#atom-everything"},{"id":"28285c28-72c8-4d0c-9d82-8b3c8a0a17a2","title":"Research: PCGamer Article Performance Audit","summary":"The PCGamer article audit reveals excessive page bloat due to ad-tech, resulting in significant network traffic despite minimal core content.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-22","topics":["web-performance","ad-tech","network-traffic","content-optimization"],"diataxisType":"explanation","path":"/kb/articles/research-pcgamer-article-performance-audit-28285c28","sourceUrl":"https://simonwillison.net/2026/Mar/22/pcgamer-audit/#atom-everything"},{"id":"da86c0c4-d291-4e38-b680-cebefe4bcf18","title":"Tool: DNS Lookup","summary":"The article discusses Cloudflare's 1.1.1.1 DNS service and its CORS-enabled JSON API, highlighting a UI built for DNS queries.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-22","topics":["dns","cloudflare","api","web-development"],"diataxisType":"how-to","path":"/kb/articles/tool-dns-lookup-da86c0c4","sourceUrl":"https://simonwillison.net/2026/Mar/22/dns/#atom-everything"},{"id":"592e6048-45f1-4abf-a4d4-c387bb349713","title":"The art of influence: The single most important skill that AI can’t replace | Jessica Fain (Webflow, ex-Slack)","summary":"Jessica Fain discusses the importance of influence in decision-making and how AI will enhance, rather than replace, this skill in leadership.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-22","topics":["influence","leadership","ai-impact","decision-making"],"diataxisType":"explanation","path":"/kb/articles/the-art-of-influence-the-single-most-important-skill-that-ai-592e6048","sourceUrl":"https://www.lennysnewsletter.com/p/the-art-of-influence-jessica-fain"},{"id":"d8de17a0-b636-4ae6-b55d-c13358fd6527","title":"Research: JavaScript Sandboxing Research","summary":"This research analyzes JavaScript sandboxing options for running untrusted code, comparing various approaches and tools in the Node.js ecosystem.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-22","topics":["javascript","sandboxing","nodejs","security"],"diataxisType":"explanation","path":"/kb/articles/research-javascript-sandboxing-research-d8de17a0","sourceUrl":"https://simonwillison.net/2026/Mar/22/javascript-sandboxing-research/#atom-everything"},{"id":"0949cffa-b488-4330-a997-3f70de80709b","title":"Tool: Merge State Visualizer","summary":"Simon Willison discusses Bram Cohen's vision for version control using CRDTs and demonstrates an interactive UI built with Pyodide.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-22","topics":["version-control","crdts","interactive-ui"],"diataxisType":"explanation","path":"/kb/articles/tool-merge-state-visualizer-0949cffa","sourceUrl":"https://simonwillison.net/2026/Mar/22/manyana/#atom-everything"},{"id":"8015e4ff-996f-47ef-a02c-d9bc9ed41a01","title":"Experimenting with Starlette 1.0 with Claude skills","summary":"The article discusses the release of Starlette 1.0, its significance in the Python ecosystem, and how Claude can be used to create skills and applications with it.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-22","topics":["python-frameworks","starlette","ai-agents","software-development"],"diataxisType":"how-to","path":"/kb/articles/experimenting-with-starlette-10-with-claude-skills-8015e4ff","sourceUrl":"https://simonwillison.net/2026/Mar/22/starlette/#atom-everything"},{"id":"9f31bc8e-0507-44b0-9aa1-303b96e7aa5c","title":"Profiling Hacker News users based on their comments","summary":"The article explores profiling Hacker News users through their comments, highlighting Simon Willison's insights on AI coding tools and security concerns.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-21","topics":["ai-coding","user-profiling","security-concerns","hacker-news","software-engineering"],"diataxisType":"explanation","path":"/kb/articles/profiling-hacker-news-users-based-on-their-comments-9f31bc8e","sourceUrl":"https://simonwillison.net/2026/Mar/21/profiling-hacker-news-users/#atom-everything"},{"id":"09bfa2d9-d9a4-415e-8026-d6c5b52cccb0","title":"🧠 Community Wisdom: Beating the solo-founder procrastination trap, best tools for B2B prospecting, finding your first real users pre-launch, and more","summary":"This edition of Community Wisdom discusses overcoming procrastination as a solo founder, effective B2B prospecting tools, and strategies for finding early users before launch.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-21","topics":["solo-founders","b2b-prospecting","user-acquisition","community-insights"],"diataxisType":"how-to","path":"/kb/articles/-community-wisdom-beating-the-solo-founder-procrastination-t-09bfa2d9","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-beating-the-solo"},{"id":"c8601f50-9a8f-45ec-b909-e34963b65cdf","title":"Using Git with coding agents - Agentic Engineering Patterns","summary":"The article discusses the use of Git with coding agents, highlighting its features, commands, and the benefits of version control in software development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-21","topics":["git","coding-agents","version-control","software-development"],"diataxisType":"how-to","path":"/kb/articles/using-git-with-coding-agents-agentic-engineering-patterns-c8601f50","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/using-git-with-coding-agents/#atom-everything"},{"id":"e613c8df-9ee0-43bb-a111-488318dee9a9","title":"Turbo Pascal 3.02A, deconstructed","summary":"The article explores the deconstruction of Turbo Pascal 3.02A, showcasing how its binary can be interpreted and decompiled to understand its structure and functionality.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-20","topics":["turbo-pascal","decompilation","software-history"],"diataxisType":"explanation","path":"/kb/articles/turbo-pascal-302a-deconstructed-e613c8df","sourceUrl":"https://simonwillison.net/2026/Mar/20/turbo-pascal/#atom-everything"},{"id":"38c97ff6-e26b-4d16-8fe6-0b6d1ffdeb39","title":"A quote from Kimi.ai @Kimi_Moonshot","summary":"Kimi.ai congratulates the Cursor AI team on the launch of Composer 2, highlighting the integration of their Kimi-k2.5 model in the process.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-20","topics":["ai-integration","machine-learning","partnerships"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-kimiai-kimi_moonshot-38c97ff6","sourceUrl":"https://simonwillison.net/2026/Mar/20/cursor-on-kimi/#atom-everything"},{"id":"0db41a58-7760-418d-9c06-af9f8fe74df3","title":"Dreamer: the Personal Agent OS — David Singleton","summary":"David Singleton introduces Dreamer, a platform for building and using AI agents, emphasizing its consumer-friendly design and the importance of community-driven development.","author":"Swyx","authorId":"swyx","published":"2026-03-20","topics":["ai-agents","consumer-tech","startup-ecosystem"],"diataxisType":"explanation","path":"/kb/articles/dreamer-the-personal-agent-os-david-singleton-0db41a58","sourceUrl":"https://www.latent.space/p/dreamer"},{"id":"42ddf81c-7c69-4752-bed5-3fdbc88fcd0b","title":"[AINews] Every Lab serious enough about Developers has bought their own Devtools","summary":"OpenAI's acquisition of Astral marks a significant shift in developer tooling, emphasizing the importance of agentic coding and enterprise-focused AI solutions.","author":"Swyx","authorId":"swyx","published":"2026-03-20","topics":["ai-development","developer-tools","agentic-coding","enterprise-ai"],"diataxisType":"explanation","path":"/kb/articles/ainews-every-lab-serious-enough-about-developers-has-bought--42ddf81c","sourceUrl":"https://www.latent.space/p/ainews-every-lab-serious-enough-about"},{"id":"02e6b89f-443f-4417-8969-f1589c11cd1f","title":"[AINews] MiniMax 2.7: GLM-5 at 1/3 cost SOTA Open Model","summary":"MiniMax 2.7 launches as an efficient GLM-5 competitor, emphasizing self-evolution and multi-agent collaboration in AI models.","author":"Swyx","authorId":"swyx","published":"2026-03-19","topics":["ai-models","self-evolution","multi-agent-collaboration","cost-efficiency"],"diataxisType":"explanation","path":"/kb/articles/ainews-minimax-27-glm-5-at-13-cost-sota-open-model-02e6b89f","sourceUrl":"https://www.latent.space/p/ainews-minimax-27-glm-5-at-13-cost"},{"id":"19af0653-e0d2-40e0-b799-ba2461dcb38c","title":"Thoughts on OpenAI acquiring Astral and uv/ruff/ty","summary":"Simon Willison shares insights on OpenAI's acquisition of Astral, discussing its impact on the Python ecosystem and the future of open source tools.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-19","topics":["openai","astral","python","open-source","software-development"],"diataxisType":"explanation","path":"/kb/articles/thoughts-on-openai-acquiring-astral-and-uvruffty-19af0653","sourceUrl":"https://simonwillison.net/2026/Mar/19/openai-acquiring-astral/#atom-everything"},{"id":"50a8274d-d701-412a-87db-aa9931e40aa0","title":"Autoresearching Apple’s “LLM in a Flash” to run Qwen 397B locally","summary":"Dan Woods successfully ran a custom version of Qwen 397B locally on a MacBook Pro using techniques from Apple's research on efficient LLM inference.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-18","topics":["llm-inference","apple-research","qwen-model","machine-learning"],"diataxisType":"how-to","path":"/kb/articles/autoresearching-apples-llm-in-a-flash-to-run-qwen-397b-local-50a8274d","sourceUrl":"https://simonwillison.net/2026/Mar/18/llm-in-a-flash/#atom-everything"},{"id":"963a4413-b3df-4079-9180-e4629b2457c9","title":"[AINews] Claude Cowork Dispatch: Anthropic's Answer to OpenClaw","summary":"Anthropic introduces Claude Cowork as a competitive alternative to OpenClaw, emphasizing its design choices and use cases in the evolving AI landscape.","author":"Swyx","authorId":"swyx","published":"2026-03-18","topics":["ai-agents","openai","model-architecture","coding-optimizations"],"diataxisType":"explanation","path":"/kb/articles/ainews-claude-cowork-dispatch-anthropics-answer-to-openclaw-963a4413","sourceUrl":"https://www.latent.space/p/ainews-claude-cowork-dispatch-anthropics"},{"id":"677f1517-8b76-4002-af0c-033692442f24","title":"How I built LennyRPG","summary":"Lenny Rachitsky discusses the creation of LennyRPG, a Pokémon-style game built by Ben Shih, detailing the workflow and lessons learned in the process.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-18","topics":["game-development","podcast","ai-tools","product-design"],"diataxisType":"how-to","path":"/kb/articles/how-i-built-lennyrpg-677f1517","sourceUrl":"https://www.lennysnewsletter.com/p/how-i-built-lennyrpg-f85"},{"id":"7f0acf20-2884-4fbb-b112-2486b6487694","title":"Snowflake Cortex AI Escapes Sandbox and Executes Malware","summary":"The article discusses a prompt injection attack on Snowflake's Cortex AI that allowed malware execution, highlighting vulnerabilities in command safety.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-18","topics":["ai-security","malware-execution","prompt-injection","cortex-ai"],"diataxisType":"explanation","path":"/kb/articles/snowflake-cortex-ai-escapes-sandbox-and-executes-malware-7f0acf20","sourceUrl":"https://simonwillison.net/2026/Mar/18/snowflake-cortex-ai/#atom-everything"},{"id":"fec98cbc-71ac-4777-933e-c73176c97f11","title":"Why Anthropic Thinks AI Should Have Its Own Computer — Felix Rieseberg of Claude Cowork & Claude Code Desktop","summary":"Felix Rieseberg discusses the development of Claude Cowork at Anthropic, emphasizing the shift towards local-first AI workflows and the importance of user-friendly design in knowledge work.","author":"Swyx","authorId":"swyx","published":"2026-03-17","topics":["ai-workflows","local-computing","knowledge-work","user-experience"],"diataxisType":"explanation","path":"/kb/articles/why-anthropic-thinks-ai-should-have-its-own-computer-felix-r-fec98cbc","sourceUrl":"https://www.latent.space/p/felix-anthropic"},{"id":"344780b7-e934-49f0-a913-cd02367cf96b","title":"Subagents - Agentic Engineering Patterns","summary":"The article discusses how subagents enhance the efficiency of coding agents by managing context limits and performing tasks without consuming the parent agent's tokens.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-17","topics":["ai-agents","subagents","coding-agents","context-management"],"diataxisType":"explanation","path":"/kb/articles/subagents-agentic-engineering-patterns-344780b7","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/subagents/#atom-everything"},{"id":"1b93707c-218e-463c-8a33-eccbf531c320","title":"A quote from Ken Jin","summary":"Ken Jin announces that the CPython JIT has met its performance goals ahead of schedule for macOS AArch64 and x86_64 Linux.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-17","topics":["python","jit","performance","macos","linux"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-ken-jin-1b93707c","sourceUrl":"https://simonwillison.net/2026/Mar/17/ken-jin/#atom-everything"},{"id":"1b54d4b1-9085-4ffb-b5a5-38de292a2f03","title":"GPT-5.4 mini and GPT-5.4 nano, which can describe 76,000 photos for $52","summary":"OpenAI introduces GPT-5.4 mini and nano models, offering faster performance and lower costs for image description tasks.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-17","topics":["gpt-5","openai","ai-models","image-description","pricing"],"diataxisType":"explanation","path":"/kb/articles/gpt-54-mini-and-gpt-54-nano-which-can-describe-76000-photos--1b54d4b1","sourceUrl":"https://simonwillison.net/2026/Mar/17/mini-and-nano/#atom-everything"},{"id":"238bed2a-2274-4280-9781-b2f90d0261dc","title":"[AINews] NVIDIA GTC: Jensen goes hard on OpenClaw, Vera CPU, and announces $1T sales backlog in 2027","summary":"Jensen Huang's keynote at NVIDIA GTC highlighted advancements in OpenClaw, Vera CPU, and a projected $1T sales backlog by 2027, showcasing NVIDIA's rapid innovation.","author":"Swyx","authorId":"swyx","published":"2026-03-17","topics":["nvidia-gtc","openclaw","vera-cpu","ai-innovation"],"diataxisType":"explanation","path":"/kb/articles/ainews-nvidia-gtc-jensen-goes-hard-on-openclaw-vera-cpu-and--238bed2a","sourceUrl":"https://www.latent.space/p/ainews-nvidia-gtc-jensen-goes-hard"},{"id":"a19b5b8e-e20b-4bcc-a67b-22ce93bc85c1","title":"Pi: the architecture of an AI coding agent","summary":"Pi transforms any LLM into a coding assistant, adaptable to various workflows.","author":"Latentpatterns","authorId":"latentpatterns","published":"2026-03-17","topics":["Coding Tools","AI in Software Development","Software Architecture","Large Language Models (LLM)"],"diataxisType":"explanation","path":"/kb/articles/pi-the-architecture-of-an-ai-coding-agent-a19b5b8e","sourceUrl":"https://latentpatterns.com/newsletter/pi-mono-architecture"},{"id":"3fb2d677-8843-4202-b5db-a66adcc01769","title":"A quote from Tim Schilling","summary":"Tim Schilling emphasizes the importance of human engagement in open source contributions, warning against over-reliance on LLMs in the Django community.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-17","topics":["open-source","django","llm","community-engagement"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-tim-schilling-3fb2d677","sourceUrl":"https://simonwillison.net/2026/Mar/17/tim-schilling/#atom-everything"},{"id":"87f53d1c-b898-4a1b-a502-00cc22be585d","title":"How I built LennyRPG","summary":"Lenny Rachitsky shares the journey of building LennyRPG, a fun, video-game-inspired project based on his podcast, and encourages others to create using his data.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-17","topics":["product-development","game-design","ai-tools","podcast"],"diataxisType":"how-to","path":"/kb/articles/how-i-built-lennyrpg-87f53d1c","sourceUrl":"https://www.lennysnewsletter.com/p/how-i-built-lennyrpg"},{"id":"5eff2998-d2ec-409a-95ec-a88e89968c3d","title":"🎙️ This week on How I AI: From Figma to Claude Code and back & From journalist to iOS developer","summary":"This article discusses the integration of AI in design and development workflows, featuring insights from Figma and LinkedIn experts on enhancing collaboration and productivity.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-16","topics":["ai-in-design","ios-development","workflow-automation","collaboration-tools"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-from-figma-to-claude-code-and-back-fr-5eff2998","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-from-figma"},{"id":"af2b4505-80fa-4c07-9028-23515597e7b9","title":"Use subagents and custom agents in Codex","summary":"OpenAI Codex has launched subagents for general use, allowing for custom agents defined in TOML files, enhancing coding capabilities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["ai-agents","custom-agents","openai-codex"],"diataxisType":"how-to","path":"/kb/articles/use-subagents-and-custom-agents-in-codex-af2b4505","sourceUrl":"https://simonwillison.net/2026/Mar/16/codex-subagents/#atom-everything"},{"id":"f086afa3-ea25-4be3-b6c8-7261f6268f4f","title":"Coding agents for data analysis","summary":"Simon Willison's workshop at NICAR 2026 focused on using coding agents like Claude Code and OpenAI Codex for data analysis, aimed at data journalists.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["data-analysis","coding-agents","workshop","data-journalism"],"diataxisType":"tutorial","path":"/kb/articles/coding-agents-for-data-analysis-f086afa3","sourceUrl":"https://simonwillison.net/2026/Mar/16/coding-agents-for-data-analysis/#atom-everything"},{"id":"b2ce2a44-3c5d-45de-9f2e-f738bfa88aca","title":"cognitive security","summary":"Geoffrey Huntley discusses the implications of AI as a financial weapon and the risks of outsourcing cognitive security to AI models.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-03-16","topics":["cognitive-security","ai-ethics","open-source-ai","economic-warfare"],"diataxisType":"explanation","path":"/kb/articles/cognitive-security-b2ce2a44","sourceUrl":"https://ghuntley.com/cogsec/"},{"id":"a42dccce-1f36-4196-a685-962be4d1b096","title":"How coding agents work - Agentic Engineering Patterns","summary":"The article explains how coding agents function by leveraging Large Language Models (LLMs) and their capabilities, including token processing and tool calling.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["coding-agents","large-language-models","token-processing","tool-calling"],"diataxisType":"explanation","path":"/kb/articles/how-coding-agents-work-agentic-engineering-patterns-a42dccce","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/how-coding-agents-work/#atom-everything"},{"id":"684e4fd4-459f-48df-8f68-c90a964695b5","title":"From journalist to iOS developer: How LinkedIn’s editor builds with Claude Code | Daniel Roth","summary":"Daniel Roth transitioned from a business writer to an iOS developer using Claude Code, creating multiple apps including Commutely, a train-tracking app for New York commuters.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-16","topics":["ios-development","ai-tools","app-store","non-technical-builders"],"diataxisType":"how-to","path":"/kb/articles/from-journalist-to-ios-developer-how-linkedins-editor-builds-684e4fd4","sourceUrl":"https://www.lennysnewsletter.com/p/from-journalist-to-ios-developer"},{"id":"bfd48a7a-963e-4de5-8ce2-bd6d7b6ba685","title":"AI as economic warfare","summary":"The article discusses the concept of open-source software as a financial weapon in economic warfare, particularly between the US and China, and its implications for AI development.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-03-16","topics":["open-source","economic-warfare","ai-development","geopolitics"],"diataxisType":"explanation","path":"/kb/articles/ai-as-economic-warfare-bfd48a7a","sourceUrl":"https://ghuntley.com/warfare/"},{"id":"18bf3456-1567-4433-842b-50b5852955f9","title":"A quote from Guilherme Rambo","summary":"Guilherme Rambo discusses the security of the software-based camera indicator light in the MacBook Neo, emphasizing its protection against kernel-level exploits.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["macbook","security","hardware-indicators"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-guilherme-rambo-18bf3456","sourceUrl":"https://simonwillison.net/2026/Mar/16/guilherme-rambo/#atom-everything"},{"id":"fb5286c6-329e-49a9-8bc2-a137dbdfcba1","title":"A quote from A member of Anthropic’s alignment-science team","summary":"A member of Anthropic’s alignment-science team discusses the importance of visceral results to convey misalignment risks to policymakers.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["ai-ethics","alignment-science","generative-ai"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-a-member-of-anthropics-alignment-science-team-fb5286c6","sourceUrl":"https://simonwillison.net/2026/Mar/16/blackmail/#atom-everything"},{"id":"fbd470c1-c28b-45b6-92cf-99e651f00ddb","title":"Introducing Mistral Small 4","summary":"Mistral has launched Mistral Small 4, a versatile model unifying capabilities from its flagship models, and introduced Leanstral for the Lean 4 coding language.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-16","topics":["ai-models","machine-learning","coding-languages"],"diataxisType":"explanation","path":"/kb/articles/introducing-mistral-small-4-fbd470c1","sourceUrl":"https://simonwillison.net/2026/Mar/16/mistral-small-4/#atom-everything"},{"id":"4fd35e5a-4e93-4675-99b4-3bd503626393","title":"porting software has been trivial for a while now. here’s how you do it.","summary":"The article outlines a straightforward approach to porting a codebase between programming languages using a structured method involving Ralph loops and specifications.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-03-15","topics":["software-porting","programming","ralph-loops","codebase-management"],"diataxisType":"how-to","path":"/kb/articles/porting-software-has-been-trivial-for-a-while-now-heres-how--4fd35e5a","sourceUrl":"https://ghuntley.com/porting/"},{"id":"f87be7dc-558f-4822-87a5-6933facd4faf","title":"What is agentic engineering? - Agentic Engineering Patterns","summary":"The article explores agentic engineering, a practice of developing software with coding agents that can write and execute code, emphasizing the role of human oversight in the process.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-15","topics":["agentic-engineering","coding-agents","llms","software-development"],"diataxisType":"explanation","path":"/kb/articles/what-is-agentic-engineering-agentic-engineering-patterns-f87be7dc","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/what-is-agentic-engineering/#atom-everything"},{"id":"02dc875f-b1cb-4768-a7a2-a37fe4c05627","title":"The tactical playbook for getting 20-40% more comp (without sounding greedy) | Jacob Warwick (Executive Negotiator)","summary":"Jacob Warwick shares strategies for negotiating better compensation, emphasizing collaboration and timing to achieve significant salary increases.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-15","topics":["negotiation","salary-negotiation","career-development"],"diataxisType":"how-to","path":"/kb/articles/the-tactical-playbook-for-getting-20-40-more-comp-without-so-02dc875f","sourceUrl":"https://www.lennysnewsletter.com/p/the-tactical-playbook-for-getting-more-comp"},{"id":"4574d440-3294-47e5-98e9-12f63f29e054","title":"My fireside chat about agentic engineering at the Pragmatic Summit","summary":"Simon Willison discusses his experiences with agentic engineering and AI coding tools during a fireside chat at the Pragmatic Summit, highlighting trust and testing in AI development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-14","topics":["ai-adoption","agentic-engineering","test-driven-development","ai-coding-tools"],"diataxisType":"explanation","path":"/kb/articles/my-fireside-chat-about-agentic-engineering-at-the-pragmatic--4574d440","sourceUrl":"https://simonwillison.net/2026/Mar/14/pragmatic-summit/#atom-everything"},{"id":"ae4f0509-7631-4dbc-afd2-1629272239c1","title":"A quote from Jannis Leidel","summary":"Jannis Leidel discusses the challenges faced by Jazzband due to the influx of AI-generated spam on GitHub, highlighting the need for safer operational models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-14","topics":["ai-ethics","open-source","github","spam","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-jannis-leidel-ae4f0509","sourceUrl":"https://simonwillison.net/2026/Mar/14/jannis-leidel/#atom-everything"},{"id":"ea4debd4-3074-4209-90e9-740b8591fdd0","title":"[AINews] Context Drought","summary":"Anthropic's release of 1M context models marks a significant but slow advancement in context window technology, highlighting ongoing physical constraints in memory availability.","author":"Swyx","authorId":"swyx","published":"2026-03-14","topics":["context-windows","ai-infrastructure","memory-optimization","long-context"],"diataxisType":"explanation","path":"/kb/articles/ainews-context-drought-ea4debd4","sourceUrl":"https://www.latent.space/p/ainews-context-drought"},{"id":"de9d49c7-38b7-46b5-9f03-8c8c98b2972f","title":"🧠 Community Wisdom: Getting a skeptical CTO to adopt AI coding tools, staying after losing faith in the founder, Claude skills for designers, making the most of a jack-of-all-trades role, and more","summary":"This edition of Community Wisdom discusses strategies for getting a skeptical CTO to adopt AI tools, navigating challenges in leadership, and leveraging diverse skills.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-14","topics":["ai-tools","cto-adoption","leadership-challenges","design-skills","jack-of-all-trades"],"diataxisType":"how-to","path":"/kb/articles/-community-wisdom-getting-a-skeptical-cto-to-adopt-ai-coding-de9d49c7","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-getting-a-skeptical"},{"id":"d1434027-1299-47aa-af59-4e94cf0bdf89","title":"[AINews] The high-return activity of raising your aspirations for LLMs","summary":"The article discusses the importance of raising aspirations in the context of LLMs, highlighting the contrast between ambitious and pragmatic approaches in AI development.","author":"Swyx","authorId":"swyx","published":"2026-03-13","topics":["llms","ai-aspirations","agent-infrastructure","coding-agents","multimodal-retrieval"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-high-return-activity-of-raising-your-aspirations--d1434027","sourceUrl":"https://www.latent.space/p/ainews-the-high-return-activity-of"},{"id":"0bb24078-ce0e-4b10-a0dd-8355fd324605","title":"A quote from Craig Mod","summary":"Simon Willison shares his experience of building custom accounting software that meets his specific needs, highlighting its features and efficiency.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-13","topics":["accounting-software","custom-development","productivity-tools"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-craig-mod-0bb24078","sourceUrl":"https://simonwillison.net/2026/Mar/13/craig-mod/#atom-everything"},{"id":"5c930a47-040a-4456-8289-c86d5bd6a678","title":"Compound Engineering Camp: Every Step, From Scratch","summary":"Kieran Klaassen demonstrates building an app with the compound engineering plugin.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-03-13","topics":["AI in Coding","Workflow Automation","Compound Engineering","Software Development"],"diataxisType":"tutorial","path":"/kb/articles/compound-engineering-camp-every-step-from-scratch-5c930a47","sourceUrl":"https://every.to/source-code/compound-engineering-camp-every-step-from-scratch"},{"id":"5abed48c-8243-4eea-bb7f-50fb6c5a133a","title":"1M context is now generally available for Opus 4.6 and Sonnet 4.6","summary":"1M context is now available for Opus 4.6 and Sonnet 4.6, with standard pricing across the full window and no long-context premium.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-13","topics":["opus-4-6","sonnet-4-6","openai","gemini","pricing"],"diataxisType":"explanation","path":"/kb/articles/1m-context-is-now-generally-available-for-opus-46-and-sonnet-5abed48c","sourceUrl":"https://simonwillison.net/2026/Mar/13/1m-context/#atom-everything"},{"id":"86a8ff57-fa9e-426d-8642-f72a51a5f3dd","title":"Shopify/liquid: Performance: 53% faster parse+render, 61% fewer allocations","summary":"Shopify's CEO Tobias Lütke implemented performance improvements in Liquid, achieving a 53% faster parse/render and 61% fewer allocations through innovative coding agent experiments.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-13","topics":["performance-optimization","coding-agents","liquid-template-engine","software-development"],"diataxisType":"explanation","path":"/kb/articles/shopifyliquid-performance-53-faster-parserender-61-fewer-all-86a8ff57","sourceUrl":"https://simonwillison.net/2026/Mar/13/liquid/#atom-everything"},{"id":"60d7f080-bc93-4499-906f-b279cb4db110","title":"A couple of days ago, I sat down with Vivek Bharathi and dumped my brains. Here's the interview...","summary":"Geoffrey Huntley interviews Vivek Bharathi about the evolving landscape of software development and engineering in the age of AI, emphasizing the shift towards automation and risk management.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-03-12","topics":["software-engineering","ai-tools","automation","open-source","industry-shifts"],"diataxisType":"explanation","path":"/kb/articles/a-couple-of-days-ago-i-sat-down-with-vivek-bharathi-and-dump-60d7f080","sourceUrl":"https://ghuntley.com/frontier/"},{"id":"1364a25f-5b73-4a19-acb8-aa13270e82a6","title":"Coding After Coders: The End of Computer Programming as We Know It","summary":"The article discusses the impact of AI on software development, highlighting both optimism and concerns among developers about the future of programming.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-12","topics":["ai-assisted-development","future-of-programming","software-engineering","developer-optimism"],"diataxisType":"explanation","path":"/kb/articles/coding-after-coders-the-end-of-computer-programming-as-we-kn-1364a25f","sourceUrl":"https://simonwillison.net/2026/Mar/12/coding-after-coders/#atom-everything"},{"id":"eb2223a0-4c0c-4431-b705-d0e502d34699","title":"MALUS—Clean Room as a Service","summary":"The article humorously critiques the concept of 'vibe-porting' licenses, presenting a fictional service that creates legally distinct code from open source projects.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-12","topics":["satire","open-source","ai-robots","licensing"],"diataxisType":"explanation","path":"/kb/articles/malusclean-room-as-a-service-eb2223a0","sourceUrl":"https://simonwillison.net/2026/Mar/12/malus/#atom-everything"},{"id":"fda49f6c-fb58-46f8-8352-ad51e122b55a","title":"Retrieval After RAG: Hybrid Search, Agents, and Database Design — Simon Hørup Eskildsen of Turbopuffer","summary":"Simon Hørup Eskildsen discusses the development of Turbopuffer, a search engine designed to optimize costs and improve retrieval for unstructured data, reflecting on his journey and insights.","author":"Swyx","authorId":"swyx","published":"2026-03-12","topics":["search-engine","database-design","ai-agents","cost-optimization"],"diataxisType":"explanation","path":"/kb/articles/retrieval-after-rag-hybrid-search-agents-and-database-design-fda49f6c","sourceUrl":"https://www.latent.space/p/turbopuffer"},{"id":"71aec735-4033-4623-8322-0a945a6a80ec","title":"[AINews] Replit Agent 4: The Knowledge Work Agent","summary":"Replit's recent valuation surge reflects its evolution into a comprehensive productivity suite, integrating coding agents into broader knowledge work tasks.","author":"Swyx","authorId":"swyx","published":"2026-03-12","topics":["ai-agents","productivity-tools","coding-platforms","knowledge-work"],"diataxisType":"explanation","path":"/kb/articles/ainews-replit-agent-4-the-knowledge-work-agent-71aec735","sourceUrl":"https://www.latent.space/p/ainews-replit-agent-4-the-knowledge"},{"id":"12f23a38-71ae-40ac-81e6-f73e971ac71d","title":"A quote from Les Orchard","summary":"The article discusses how AI-assisted coding reveals a divide among developers, highlighting differing motivations and approaches to coding.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-12","topics":["ai-coding","developer-divide","software-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-les-orchard-12f23a38","sourceUrl":"https://simonwillison.net/2026/Mar/12/les-orchard/#atom-everything"},{"id":"ef971a3b-0ea3-4496-adc4-017f73bb1810","title":"How I built a 1M+ subscriber newsletter and top 10 tech podcast | Lenny Rachitsky","summary":"Lenny Rachitsky shares insights on building a successful newsletter and podcast, discussing personal experiences and lessons learned along the way.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-12","topics":["newsletter-building","podcasting","personal-development","business-strategy"],"diataxisType":"how-to","path":"/kb/articles/how-i-built-a-1m-subscriber-newsletter-and-top-10-tech-podca-ef971a3b","sourceUrl":"https://www.lennysnewsletter.com/p/how-i-built-a-1m-subscriber-newsletter"},{"id":"a66fabc6-571c-435b-a4ba-71df813ce728","title":"From Figma to Claude Code and back | Gui Seiz & Alex Kern (Figma) ","summary":"Gui Seiz and Alex Kern from Figma discuss how AI transforms design workflows, enabling real-time collaboration between design and code.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-11","topics":["ai-workflows","figma","design-to-code","collaboration"],"diataxisType":"how-to","path":"/kb/articles/from-figma-to-claude-code-and-back-gui-seiz-alex-kern-figma--a66fabc6","sourceUrl":"https://www.lennysnewsletter.com/p/from-figma-to-claude-code-and-back"},{"id":"7b7eda20-c6a5-45fe-98a5-70be92cdada9","title":"A quote from John Carmack","summary":"Simon Willison shares a quote from John Carmack about the challenges of architecting for future requirements in software development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-11","topics":["software-engineering","john-carmack","yagni"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-john-carmack-7b7eda20","sourceUrl":"https://simonwillison.net/2026/Mar/11/john-carmack/#atom-everything"},{"id":"5ce9775c-46b2-41f5-b016-d84b1b4a6bea","title":"[AINews] Yann LeCun’s AMI Labs launches with a $1.03B seed to build world models around JEPA","summary":"Yann LeCun's AMI Labs has launched with a $1.03B seed funding to develop AI models that understand the physical world, marking a significant milestone in AI development.","author":"Swyx","authorId":"swyx","published":"2026-03-11","topics":["ai-startups","world-models","funding-news","jepa","machine-learning"],"diataxisType":"reference","path":"/kb/articles/ainews-yann-lecuns-ami-labs-launches-with-a-103b-seed-to-bui-5ce9775c","sourceUrl":"https://www.latent.space/p/ainews-yann-lecuns-ami-labs-launches"},{"id":"8416606f-ca45-49af-8b57-4065974039cd","title":"Sorting algorithms","summary":"Simon Willison shares his experience creating animated demonstrations of sorting algorithms using Claude Artifacts, including Python's Timsort.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-11","topics":["sorting-algorithms","python","animation","programming"],"diataxisType":"explanation","path":"/kb/articles/sorting-algorithms-8416606f","sourceUrl":"https://simonwillison.net/2026/Mar/11/sorting-algorithms/#atom-everything"},{"id":"2e95e9ec-7abd-4500-b2b9-46d1325694fc","title":"A guide to advanced B2B positioning","summary":"Lenny Rachitsky discusses advanced B2B positioning, highlighting common roadblocks and strategies to overcome them for effective product differentiation.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-10","topics":["b2b-positioning","product-marketing","growth-strategies","competitive-analysis"],"diataxisType":"how-to","path":"/kb/articles/a-guide-to-advanced-b2b-positioning-2e95e9ec","sourceUrl":"https://www.lennysnewsletter.com/p/a-guide-to-advanced-b2b-positioning"},{"id":"c9dc19c4-45d7-469f-a175-f0bfb28ebfe8","title":"NVIDIA's AI Engineers: Agent Inference at Planetary Scale and \"Speed of Light\" — Nader Khalil (Brev), Kyle Kranen (Dynamo)","summary":"NVIDIA's engineers discuss advancements in AI inference at GTC, focusing on Dynamo's capabilities and the importance of security in agent operations.","author":"Swyx","authorId":"swyx","published":"2026-03-10","topics":["ai-inference","nvidia-dynamo","agent-security","gpu-optimization"],"diataxisType":"explanation","path":"/kb/articles/nvidias-ai-engineers-agent-inference-at-planetary-scale-and--c9dc19c4","sourceUrl":"https://www.latent.space/p/nvidia-brev-dynamo"},{"id":"33fd2c03-74b2-4aaf-b6ea-9f18f4ccda9f","title":"[AINews] Autoresearch: Sparks of Recursive Self Improvement","summary":"The article discusses the emergence of LLMs capable of autonomously training smaller models, highlighting the potential for recursive self-improvement in AI.","author":"Swyx","authorId":"swyx","published":"2026-03-10","topics":["recursive-self-improvement","llms","ai-research","autonomous-training"],"diataxisType":"explanation","path":"/kb/articles/ainews-autoresearch-sparks-of-recursive-self-improvement-33fd2c03","sourceUrl":"https://www.latent.space/p/ainews-autoresearch-sparks-of-recursive"},{"id":"54ae3f25-eb05-4cdf-a4b8-6ad332352032","title":"A guide to advanced B2B positioning","summary":"Lenny Rachitsky interviews April Dunford on advanced B2B positioning, discussing tactics to overcome common roadblocks in product distribution.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-10","topics":["b2b-positioning","product-distribution","marketing-strategy"],"diataxisType":"how-to","path":"/kb/articles/a-guide-to-advanced-b2b-positioning-54ae3f25","sourceUrl":"https://www.lennysnewsletter.com/p/a-guide-to-advanced-b2b-positioning-e49"},{"id":"a24024f2-26e2-4faf-b643-c0f6202d6826","title":"AI should help us produce better code - Agentic Engineering Patterns","summary":"The article discusses how AI coding agents can enhance code quality by reducing technical debt and facilitating better decision-making in software development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-10","topics":["ai-agents","technical-debt","software-development","refactoring","engineering-patterns"],"diataxisType":"how-to","path":"/kb/articles/ai-should-help-us-produce-better-code-agentic-engineering-pa-a24024f2","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/better-code/#atom-everything"},{"id":"47fc8cda-76c1-4998-873c-b1380be0f26a","title":"Perhaps not Boring Technology after all","summary":"The article discusses how LLMs for programming may not limit technology choices as previously thought, showing positive results with new tools despite training data biases.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-09","topics":["llms","programming","technology-choices","coding-agents"],"diataxisType":"explanation","path":"/kb/articles/perhaps-not-boring-technology-after-all-47fc8cda","sourceUrl":"https://simonwillison.net/2026/Mar/9/not-so-boring/#atom-everything"},{"id":"b4502e5f-aca5-4554-8559-b5cfaa24aa75","title":"a sneak preview behind an embedded software factory. I suspect rapid application dev is back","summary":"Exploring rapid app development within an embedded software factory concept.","author":"Ghuntley","authorId":"ghuntley","published":"2026-03-09","topics":["AI in Software Development","Rapid Application Development","Developer Productivity","Embedded Software"],"diataxisType":"explanation","path":"/kb/articles/a-sneak-preview-behind-an-embedded-software-factory-i-suspec-b4502e5f","sourceUrl":"https://ghuntley.com/rad"},{"id":"2e3c1d42-d1dc-47fe-8bce-de8668ebadb5","title":"Mastering Midjourney: How to create consistent, beautiful brand imagery without complex prompts","summary":"Jamey Gannon shares her streamlined workflow for creating cohesive brand imagery using AI tools like Midjourney, emphasizing effective mood boards and personalization codes.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-09","topics":["ai-image-generation","branding","midjourney","workflow","design-techniques"],"diataxisType":"how-to","path":"/kb/articles/mastering-midjourney-how-to-create-consistent-beautiful-bran-2e3c1d42","sourceUrl":"https://www.lennysnewsletter.com/p/mastering-midjourney-how-to-create"},{"id":"7474b8be-5be4-4aa5-98d4-03f285a83744","title":"Production query plans without production data","summary":"The article discusses new PostgreSQL features that allow developers to replicate production query plans in development environments without needing production data.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-09","topics":["postgresql","query-planning","database-statistics"],"diataxisType":"explanation","path":"/kb/articles/production-query-plans-without-production-data-7474b8be","sourceUrl":"https://simonwillison.net/2026/Mar/9/production-query-plans-without-production-data/#atom-everything"},{"id":"62f7ee41-24d5-41c5-b6d4-e0cb5ee87ca4","title":"🎙️ This week on How I AI: Mastering Midjourney: How to create consistent, beautiful brand imagery without complex prompts","summary":"This episode of How I AI features Jamey Gannon discussing her workflow for creating consistent brand imagery in Midjourney without complex prompts.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-09","topics":["midjourney","ai-workflows","branding","image-generation"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-mastering-midjourney-how-to-create-co-62f7ee41","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-mastering-midjourney"},{"id":"675f0e5a-193c-42f8-927f-6d3454805ea5","title":"a sneak preview behind an embedded software factory. I suspect rapid application dev is back","summary":"Geoffrey Huntley discusses the evolution of embedded software development, emphasizing the return of rapid application development and hyper-personalized software.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-03-09","topics":["embedded-software","rapid-application-development","ai-tools","developer-productivity"],"diataxisType":"explanation","path":"/kb/articles/a-sneak-preview-behind-an-embedded-software-factory-i-suspec-675f0e5a","sourceUrl":"https://ghuntley.com/rad/"},{"id":"2a3484f4-3662-4f03-8d00-7f763852e59d","title":"Eval awareness in Claude Opus 4.6’s BrowseComp performance","summary":"Instead of inadvertently coming across a leaked answer, Claude Opus 4.6 independently hypothesized that it was being evaluated, identified which benchmark it was running in, then located and decrypted BrowseComp is an evaluation designed to test how well models can find hard-to-locate information on","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-03-08","topics":["Eval Awareness","AI Benchmarking","Model Contamination"],"diataxisType":"explanation","path":"/kb/articles/eval-awareness-in-claude-opus-46s-browsecomp-performance-2a3484f4","sourceUrl":"https://www.anthropic.com/engineering/eval-awareness-browsecomp"},{"id":"7b6943e4-8e70-4a21-b040-26b690269e7b","title":"The most successful AI company you’ve never heard of | Qasar Younis","summary":"Qasar Younis, CEO of Applied Intuition, discusses the future of AI in various industries and shares insights from his experience at Y Combinator.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-08","topics":["ai","startups","technology","business-strategy"],"diataxisType":"explanation","path":"/kb/articles/the-most-successful-ai-company-youve-never-heard-of-qasar-yo-7b6943e4","sourceUrl":"https://www.lennysnewsletter.com/p/the-most-successful-ai-company-youve-never-heard-of"},{"id":"d4d113b0-d62e-42a1-b3f9-ba6107a608c3","title":"A quote from Joseph Weizenbaum","summary":"Simon Willison shares a quote from Joseph Weizenbaum about the impact of short exposures to computer programs on human perception.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-08","topics":["computer-history","ai-ethics","internet-archive"],"diataxisType":"reference","path":"/kb/articles/a-quote-from-joseph-weizenbaum-d4d113b0","sourceUrl":"https://simonwillison.net/2026/Mar/8/joseph-weizenbaum/#atom-everything"},{"id":"8f0b7b58-e07e-46a0-8c1f-8afb47e5c6c5","title":"Codex for Open Source","summary":"OpenAI has launched a six-month ChatGPT Pro offer for maintainers of popular open source projects, similar to Anthropic's Claude Max initiative.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-07","topics":["open-source","ai-tools","chatgpt","codex"],"diataxisType":"explanation","path":"/kb/articles/codex-for-open-source-8f0b7b58","sourceUrl":"https://simonwillison.net/2026/Mar/7/codex-for-open-source/#atom-everything"},{"id":"77effeb4-8b0d-402b-bc3a-46def304e4f7","title":"[AINews] AI Engineer will be the LAST job","summary":"The article discusses the future of AI engineering, predicting it will be the last job as AI continues to automate white-collar roles, particularly in software engineering.","author":"Swyx","authorId":"swyx","published":"2026-03-07","topics":["ai-engineering","job-market","automation","software-engineering","future-of-work"],"diataxisType":"explanation","path":"/kb/articles/ainews-ai-engineer-will-be-the-last-job-77effeb4","sourceUrl":"https://www.latent.space/p/ainews-ai-engineer-will-be-the-last"},{"id":"729c3062-4f03-488f-a86b-5159688c00c7","title":"🧠 Community Wisdom: Business books that haven’t aged well, vibe coding with your Figma design systems, Claude Code vs. other coding tools and more","summary":"This week's Community Wisdom discusses outdated business books, vibe coding with Figma, and compares Claude Code with other coding tools.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-07","topics":["business-books","figma-design","coding-tools"],"diataxisType":"reference","path":"/kb/articles/-community-wisdom-business-books-that-havent-aged-well-vibe--729c3062","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-business-books-that"},{"id":"28898060-561e-4711-af02-495ed4947843","title":"Anthropic and the Pentagon","summary":"The article discusses the competitive landscape of AI models, focusing on Anthropic's positioning as a moral AI provider amidst contracts with the Pentagon.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-06","topics":["ai-ethics","military-ai","ai-competition"],"diataxisType":"explanation","path":"/kb/articles/anthropic-and-the-pentagon-28898060","sourceUrl":"https://simonwillison.net/2026/Mar/6/anthropic-and-the-pentagon/#atom-everything"},{"id":"7379e7ff-fd3a-42eb-add5-3a7d8822a88c","title":"Cursor's Third Era: Cloud Agents","summary":"Cursor has launched cloud agents, marking a significant shift in coding practices, emphasizing the transition from traditional IDEs to AI-driven workflows.","author":"Swyx","authorId":"swyx","published":"2026-03-06","topics":["cloud-agents","ai-coding","developer-tools","automation"],"diataxisType":"explanation","path":"/kb/articles/cursors-third-era-cloud-agents-7379e7ff","sourceUrl":"https://www.latent.space/p/cursor-third-era"},{"id":"69028ae0-36b3-4ae3-a0db-faefcbe30ac7","title":"[AINews] GPT 5.4: SOTA Knowledge Work -and- Coding -and- CUA Model, OpenAI is so very back","summary":"OpenAI's GPT-5.4 launch introduces a unified model for coding and non-coding tasks, showcasing significant improvements in efficiency and user experience.","author":"Swyx","authorId":"swyx","published":"2026-03-06","topics":["gpt-5-4","openai","ai-agents","coding","productivity"],"diataxisType":"explanation","path":"/kb/articles/ainews-gpt-54-sota-knowledge-work-and-coding-and-cua-model-o-69028ae0","sourceUrl":"https://www.latent.space/p/ainews-gpt-54-sota-knowledge-work"},{"id":"c5b36972-ea47-4c90-87ef-4749c4eb3067","title":"Clinejection — Compromising Cline’s Production Releases just by Prompting an Issue Triager","summary":"The article discusses a prompt injection attack on Cline's GitHub repository that compromised their production releases through a flawed issue triage workflow.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-06","topics":["prompt-injection","github-security","ai-triage","cache-poisoning"],"diataxisType":"explanation","path":"/kb/articles/clinejection-compromising-clines-production-releases-just-by-c5b36972","sourceUrl":"https://simonwillison.net/2026/Mar/6/clinejection/#atom-everything"},{"id":"ca44e59a-8f7f-4015-ad88-a11e8ae6fb73","title":"Agentic manual testing - Agentic Engineering Patterns","summary":"The article discusses the importance of agentic manual testing in coding, emphasizing how coding agents can effectively verify and improve code through both automated and manual testing methods.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-06","topics":["manual-testing","coding-agents","automated-tests","tdd","browser-automation"],"diataxisType":"how-to","path":"/kb/articles/agentic-manual-testing-agentic-engineering-patterns-ca44e59a","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/agentic-manual-testing/#atom-everything"},{"id":"b75e7d12-5aee-426a-ae48-fbe5409f1f99","title":"A quote from Ally Piechowski","summary":"Ally Piechowski shares insightful questions for developers, CTOs, and business stakeholders to audit a Rails codebase effectively.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-06","topics":["rails","code-audit","software-development"],"diataxisType":"how-to","path":"/kb/articles/a-quote-from-ally-piechowski-b75e7d12","sourceUrl":"https://simonwillison.net/2026/Mar/6/ally-piechowski/#atom-everything"},{"id":"9ef5fd46-bee5-47ac-81ef-c335cafe08f8","title":"[AINews] Is Harness Engineering real?","summary":"The article discusses the ongoing debate in Harness Engineering regarding the balance between model capabilities and the importance of the harness that connects them.","author":"Swyx","authorId":"swyx","published":"2026-03-05","topics":["harness-engineering","ai-models","agent-engineering","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/ainews-is-harness-engineering-real-9ef5fd46","sourceUrl":"https://www.latent.space/p/ainews-is-harness-engineering-real"},{"id":"3deb2a99-fd13-4d86-b1d2-c2482a48458a","title":"Introducing GPT‑5.4","summary":"OpenAI introduces GPT-5.4 with enhanced capabilities for creating and editing spreadsheets, presentations, and documents, outperforming its predecessor GPT-5.2.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-05","topics":["gpt-5-4","api-models","openai","artificial-intelligence"],"diataxisType":"explanation","path":"/kb/articles/introducing-gpt54-3deb2a99","sourceUrl":"https://simonwillison.net/2026/Mar/5/introducing-gpt54/#atom-everything"},{"id":"ebedd1c9-226d-440a-ae2c-0c48d4cae630","title":"Every Agent Needs a Box — Aaron Levie, Box","summary":"Aaron Levie discusses the evolving role of AI agents in enterprise environments, emphasizing the importance of managing corporate files and data effectively.","author":"Swyx","authorId":"swyx","published":"2026-03-05","topics":["ai-agents","saas","enterprise-software","data-management"],"diataxisType":"explanation","path":"/kb/articles/every-agent-needs-a-box-aaron-levie-box-ebedd1c9","sourceUrl":"https://www.latent.space/p/box"},{"id":"d4fb177f-424a-40b6-880e-7e7757f88f55","title":"Can coding agents relicense open source through a “clean room” implementation of code?","summary":"The article discusses the ethical and legal implications of coding agents creating a 'clean room' implementation of open source code, focusing on the chardet library controversy.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-05","topics":["open-source","clean-room-implementation","coding-agents","legal-issues","software-licensing"],"diataxisType":"explanation","path":"/kb/articles/can-coding-agents-relicense-open-source-through-a-clean-room-d4fb177f","sourceUrl":"https://simonwillison.net/2026/Mar/5/chardet/#atom-everything"},{"id":"6ba05ac0-1f1c-4767-b18c-7a9a06bc16b5","title":"Anti-patterns: things to avoid - Agentic Engineering Patterns","summary":"The article discusses common anti-patterns in agentic engineering, emphasizing the importance of reviewing code before submitting pull requests to ensure quality and accountability.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-04","topics":["agentic-engineering","code-review","pull-requests","software-development"],"diataxisType":"how-to","path":"/kb/articles/anti-patterns-things-to-avoid-agentic-engineering-patterns-6ba05ac0","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/anti-patterns/#atom-everything"},{"id":"76b24dc2-42c2-4a2e-9bce-b215c3ae5308","title":"Something is afoot in the land of Qwen","summary":"The resignation of key figures from Alibaba's Qwen team raises concerns about the future of their promising Qwen 3.5 AI models, despite their recent success.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-04","topics":["ai-models","team-resignation","alibaba","qwen"],"diataxisType":"explanation","path":"/kb/articles/something-is-afoot-in-the-land-of-qwen-76b24dc2","sourceUrl":"https://simonwillison.net/2026/Mar/4/qwen/#atom-everything"},{"id":"3a8f5bf7-eafd-4b47-bdff-504c6677b8cc","title":"[AINews] Anthropic @ $19B ARR, Qwen team leaves, Gemini and GPT bump up fast models","summary":"Anthropic reaches $19B ARR, nearing OpenAI's $20B, while Qwen faces leadership exits impacting open-source efforts. Gemini 3.1 FlashLite and GPT5.3 Instant are also highlighted.","author":"Swyx","authorId":"swyx","published":"2026-03-04","topics":["anthropic","openai","gemini","qwen","gpt"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-19b-arr-qwen-team-leaves-gemini-and-gpt-bum-3a8f5bf7","sourceUrl":"https://www.latent.space/p/ainews-anthropic-19b-arr-qwen-team"},{"id":"1bd8b063-2fdc-4282-b35f-ad6654b13f86","title":"How to debug a team that isn’t working: the Waterline Model","summary":"Lenny Rachitsky introduces the Waterline Model, a diagnostic tool for leaders to identify and address structural issues in team performance rather than blaming individuals.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-03","topics":["team-management","leadership","waterline-model","diagnostic-tools"],"diataxisType":"how-to","path":"/kb/articles/how-to-debug-a-team-that-isnt-working-the-waterline-model-1bd8b063","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-debug-a-team-that-isnt-working"},{"id":"a19b939a-5f78-461c-a9c2-5becc8ffd5f6","title":"[AINews] Truth in the time of Artifice","summary":"The article discusses the decline of trust in media due to misinformation and the impact of AI on content creation and consumption.","author":"Swyx","authorId":"swyx","published":"2026-03-03","topics":["media-trust","ai-impact","content-creation","social-media"],"diataxisType":"explanation","path":"/kb/articles/ainews-truth-in-the-time-of-artifice-a19b939a","sourceUrl":"https://www.latent.space/p/ainews-truth-in-the-time-of-artifice"},{"id":"9b566851-c9e2-4e43-bae0-50b374a5d165","title":"A quote from Donald Knuth","summary":"Donald Knuth expresses excitement over the resolution of an open problem he was working on, solved by Anthropic's AI model, highlighting advances in AI and problem-solving.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-03","topics":["generative-ai","automatic-deduction","creative-problem-solving"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-donald-knuth-9b566851","sourceUrl":"https://simonwillison.net/2026/Mar/3/donald-knuth/#atom-everything"},{"id":"d442d89a-f51e-4e4a-bd41-09ca521eb0c4","title":"You Have a Claw. Now What?","summary":"The article discusses the use of personal AI assistants called Claws, built on OpenClaw, and how to effectively utilize them in daily tasks.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-03-03","topics":["ai-assistants","openclaw","productivity","messaging-apps"],"diataxisType":"how-to","path":"/kb/articles/you-have-a-claw-now-what-d442d89a","sourceUrl":"https://every.to/source-code/you-have-a-claw-now-what"},{"id":"9005bf61-a6fd-464e-bf18-da4371833818","title":"How to debug a team that isn’t working: the Waterline Model","summary":"The article discusses the Waterline Model, a diagnostic tool for managers to identify the root causes of team underperformance without resorting to blame.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-03","topics":["team-management","leadership","waterline-model"],"diataxisType":"how-to","path":"/kb/articles/how-to-debug-a-team-that-isnt-working-the-waterline-model-9005bf61","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-debug-a-team-that-isnt-working-d43"},{"id":"969c5220-0e44-460e-a4fe-d54f49a34cdd","title":"Gemini 3.1 Flash-Lite","summary":"Gemini 3.1 Flash-Lite is Google's affordable update to the Flash-Lite family, offering various thinking levels at a competitive price.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-03","topics":["ai-models","cost-efficiency","product-update"],"diataxisType":"explanation","path":"/kb/articles/gemini-31-flash-lite-969c5220","sourceUrl":"https://simonwillison.net/2026/Mar/3/gemini-31-flash-lite/#atom-everything"},{"id":"d82cd052-11b5-4470-9247-6653694196a7","title":"🎙️ This week on How I AI: 5 OpenClaw agents run my home, finances, and code & How Coinbase scaled AI to 1,000+ engineers","summary":"This episode features Jesse Genet discussing her use of OpenClaw agents for managing various aspects of her life, and Chintan Turakhia sharing how Coinbase scaled AI across its engineering team.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-02","topics":["ai-agents","coinbase","ai-adoption","homeschooling","software-development"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-5-openclaw-agents-run-my-home-finance-d82cd052","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-5-openclaw"},{"id":"26013557-6679-4183-a7e2-fc047e4032f9","title":"February sponsors-only newsletter","summary":"Simon Willison shares insights from his February sponsors-only newsletter, covering topics like OpenClaw, Agentic Engineering, and more.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-02","topics":["openclaw","agentic-engineering","strongdm","model-releases"],"diataxisType":"explanation","path":"/kb/articles/february-sponsors-only-newsletter-26013557","sourceUrl":"https://simonwillison.net/2026/Mar/2/february-newsletter/#atom-everything"},{"id":"fba0d39a-0820-4d14-bdee-0e0ca343a872","title":"How to Kill the Code Review","summary":"The article discusses the limitations of traditional code reviews in the age of AI, advocating for a shift towards reviewing specifications and intent rather than code itself.","author":"Swyx","authorId":"swyx","published":"2026-03-02","topics":["ai-code-review","software-engineering","devops","automation","agile-development"],"diataxisType":"how-to","path":"/kb/articles/how-to-kill-the-code-review-fba0d39a","sourceUrl":"https://www.latent.space/p/reviews-dead"},{"id":"de0f1213-126d-4fce-85c9-bf3f6d2394c0","title":"GIF optimization tool using WebAssembly and Gifsicle - Agentic Engineering Patterns","summary":"The article discusses creating a web-based GIF optimization tool using Gifsicle and WebAssembly, highlighting the process and challenges involved.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-02","topics":["gif-optimization","webassembly","gifsicle","coding-agents"],"diataxisType":"how-to","path":"/kb/articles/gif-optimization-tool-using-webassembly-and-gifsicle-agentic-de0f1213","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/gif-optimization/#atom-everything"},{"id":"c7feafca-bb1d-44ea-a89e-ba2626352921","title":"OpenClaw: Setting Up Your First Personal AI Agent ","summary":"The article discusses OpenClaw, an open-source tool for creating personal AI agents, highlighting user experiences and the growing community around it.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-03-02","topics":["ai-agents","open-source","personal-assistants","technology"],"diataxisType":"how-to","path":"/kb/articles/openclaw-setting-up-your-first-personal-ai-agent--c7feafca","sourceUrl":"https://every.to/source-code/openclaw-setting-up-your-first-personal-ai-agent"},{"id":"9638a3bd-0f39-4355-8bc7-83583b12129b","title":"How Coinbase scaled AI to 1,000+ engineers | Chintan Turakhia","summary":"Chintan Turakhia discusses how Coinbase scaled AI adoption among its 1,000+ engineers, achieving significant efficiency gains in app development.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-02","topics":["ai-adoption","engineering-efficiency","coinbase","leadership","workflow-automation"],"diataxisType":"how-to","path":"/kb/articles/how-coinbase-scaled-ai-to-1000-engineers-chintan-turakhia-9638a3bd","sourceUrl":"https://www.lennysnewsletter.com/p/how-coinbase-scaled-ai-to-1000-engineers"},{"id":"cc1e6806-9b09-4c67-ba7c-d6ce87f7441a","title":"A quote from claude.com/import-memory","summary":"The article discusses the process of exporting personal data from Claude, detailing the specific memories and context to be included.","author":"Simon Willison","authorId":"simon-willison","published":"2026-03-01","topics":["data-export","personalization","ai-assistants"],"diataxisType":"how-to","path":"/kb/articles/a-quote-from-claudecomimport-memory-cc1e6806","sourceUrl":"https://simonwillison.net/2026/Mar/1/claude-import-memory/#atom-everything"},{"id":"6873f9f5-cd9a-4ada-b6aa-465ca61cdb81","title":"The design process is dead. Here’s what’s replacing it. | Jenny Wen (head of design at Claude)","summary":"Jenny Wen discusses the obsolescence of the classic design process and shares insights on her role at Anthropic and the future of design in the age of AI.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-03-01","topics":["design-process","ai-in-design","jenny-wen","anthropic"],"diataxisType":"explanation","path":"/kb/articles/the-design-process-is-dead-heres-whats-replacing-it-jenny-we-6873f9f5","sourceUrl":"https://www.lennysnewsletter.com/p/the-design-process-is-dead"},{"id":"781e9631-4736-40a8-a2b8-c5ab9ef78f18","title":"principles","summary":"AI shouldn't be gatekept behind academic jargon, hype-driven marketing, or expensive credentials. These are the principles that guide how we build products at Latent Patterns. They come from\nshipping AI-native software, making expensive mistakes, and learning what actually works when\nyou're building","author":"Latentpatterns","authorId":"latentpatterns","published":"2026-03-01","topics":["AI Agents","Prompt Engineering","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/principles-781e9631","sourceUrl":"https://latentpatterns.com/principles"},{"id":"3b185a0d-631a-4099-8227-6d22a29e6cc7","title":"Dylan Patel of SemiAnalysis on the $200B AI CapEx, Chip Wars, and Why Google Might Have No Profits in 2027 — In-Context Cooking","summary":"Dylan Patel discusses the future of AI infrastructure, the implications of massive capital expenditures, and the geopolitical landscape surrounding semiconductors in a cooking show format.","author":"Swyx","authorId":"swyx","published":"2026-02-28","topics":["ai-infrastructure","chip-wars","geopolitics","capital-expenditure","export-controls"],"diataxisType":"explanation","path":"/kb/articles/dylan-patel-of-semianalysis-on-the-200b-ai-capex-chip-wars-a-3b185a0d","sourceUrl":"https://www.latent.space/p/dylanpatel-cooking"},{"id":"e1909e93-ee9a-4711-b387-ff8dc56850fd","title":"🧠 Community Wisdom: Building your tech stack from scratch, surviving painful product demos, managing LLM credit costs, resources for new designers, and more","summary":"This edition of Community Wisdom covers building a tech stack, managing product demos, LLM credit costs, and resources for new designers.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-28","topics":["tech-stack","product-demos","llm-costs","design-resources"],"diataxisType":"how-to","path":"/kb/articles/-community-wisdom-building-your-tech-stack-from-scratch-surv-e1909e93","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-building-your-tech"},{"id":"7f547407-b025-4b5a-ab2a-1532ab7f7839","title":"Interactive explanations - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"The article discusses the importance of understanding code to reduce cognitive debt, highlighting the use of interactive explanations and animations for better comprehension.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-28","topics":["cognitive-debt","interactive-explanations","ai-agents","rust-programming"],"diataxisType":"explanation","path":"/kb/articles/interactive-explanations-agentic-engineering-patterns-simon--7f547407","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/interactive-explanations/#atom-everything"},{"id":"6ca055dc-18da-4191-97d5-bd04eb087c0e","title":"[AINews] OpenAI closes $110B raise from Amazon, NVIDIA, SoftBank in largest startup fundraise in history @ $840B post-money","summary":"OpenAI has successfully raised $110B from Amazon, NVIDIA, and SoftBank, marking the largest startup fundraise in history at an $840B post-money valuation.","author":"Swyx","authorId":"swyx","published":"2026-02-28","topics":["fundraising","openai","investment","ai-industry"],"diataxisType":"reference","path":"/kb/articles/ainews-openai-closes-110b-raise-from-amazon-nvidia-softbank--6ca055dc","sourceUrl":"https://www.latent.space/p/ainews-openai-closes-110b-raise-from"},{"id":"d793786e-a79d-4f96-9bf8-2c1813f77465","title":"An AI agent coding skeptic tries AI agent coding, in excessive detail","summary":"Simon Willison discusses Max Woolf's exploration of AI coding agents, highlighting their advancements and the challenges of skepticism in the AI community.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-27","topics":["ai-agents","machine-learning","rust-programming","coding-skepticism"],"diataxisType":"explanation","path":"/kb/articles/an-ai-agent-coding-skeptic-tries-ai-agent-coding-in-excessiv-d793786e","sourceUrl":"https://simonwillison.net/2026/Feb/27/ai-agent-coding-in-excessive-detail/#atom-everything"},{"id":"5f2ace57-5385-4f1e-a715-d8c587d147ad","title":"Please, please, please stop using passkeys for encrypting user data","summary":"The article argues against using passkeys for encrypting user data, highlighting the risks of users losing their passkeys and the irreversible encryption of their data.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-27","topics":["data-encryption","user-security","authentication"],"diataxisType":"explanation","path":"/kb/articles/please-please-please-stop-using-passkeys-for-encrypting-user-5f2ace57","sourceUrl":"https://simonwillison.net/2026/Feb/27/passkeys/#atom-everything"},{"id":"33a9500f-774b-429c-9c7c-d805ff48caa7","title":"Software development now costs less than than the wage of a minimum wage worker","summary":"Software dev cost drops below minimum wage, altering the industry's landscape.","author":"Ghuntley","authorId":"ghuntley","published":"2026-02-27","topics":["AI impact on society","Software development economics","Future of work","Venture capital in the AI era"],"diataxisType":"explanation","path":"/kb/articles/software-development-now-costs-less-than-than-the-wage-of-a--33a9500f","sourceUrl":"https://ghuntley.com/real"},{"id":"548dc4c0-b5c2-4d69-af6b-025a4a5b2861","title":"Unicode Explorer using binary search over fetch() HTTP range requests","summary":"Simon Willison explores using binary search over HTTP range requests to create a Unicode Explorer, demonstrating the application of LLMs in coding.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-27","topics":["http-range-requests","unicode","binary-search","llms","web-development"],"diataxisType":"how-to","path":"/kb/articles/unicode-explorer-using-binary-search-over-fetch-http-range-r-548dc4c0","sourceUrl":"https://simonwillison.net/2026/Feb/27/unicode-explorer/#atom-everything"},{"id":"bc19574f-6b2c-4c36-a6f8-dce9bf733de4","title":"METR’s Joel Becker on exponential Time Horizon Evals, Threat Models, and the Limits of AI Productivity","summary":"Swyx interviews Joel Becker from METR about the complexities of AI productivity, time horizon evaluations, and the importance of understanding benchmarks in AI.","author":"Swyx","authorId":"swyx","published":"2026-02-27","topics":["ai-productivity","time-horizon-evals","threat-models","benchmarking"],"diataxisType":"explanation","path":"/kb/articles/metrs-joel-becker-on-exponential-time-horizon-evals-threat-m-bc19574f","sourceUrl":"https://www.latent.space/p/metr"},{"id":"ae6359a2-e0cc-48da-8291-8d869cd0b1ea","title":"Free Claude Max for (large project) open source maintainers","summary":"Anthropic is offering their $200/month Claude Max 20x plan for free to open source maintainers for six months, with specific eligibility criteria.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-27","topics":["open-source","maintainers","claude-max","anthropic"],"diataxisType":"reference","path":"/kb/articles/free-claude-max-for-large-project-open-source-maintainers-ae6359a2","sourceUrl":"https://simonwillison.net/2026/Feb/27/claude-max-oss-six-months/#atom-everything"},{"id":"2c298650-09a3-49a9-8d89-0980995e33a8","title":"Software development now costs less than than the wage of a minimum wage worker","summary":"Software dev costs now below minimum wage due to AI advancements.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-02-27","topics":["Software Development Trends","AI Impact on Society","Economic Implications of AI"],"diataxisType":"explanation","path":"/kb/articles/software-development-now-costs-less-than-than-the-wage-of-a--2c298650","sourceUrl":"https://ghuntley.com/real/"},{"id":"26e65d48-1a90-4fe7-9760-5479f3938c67","title":"[AINews] Nano Banana 2 aka Gemini 3.1 Flash Image Preview: the new SOTA Imagegen model","summary":"Nano Banana 2, also known as Gemini 3.1 Flash Image, is a new image generation model that offers competitive pricing and advanced features for AI engineers.","author":"Swyx","authorId":"swyx","published":"2026-02-27","topics":["image-generation","ai-models","pricing-strategy","text-rendering"],"diataxisType":"explanation","path":"/kb/articles/ainews-nano-banana-2-aka-gemini-31-flash-image-preview-the-n-26e65d48","sourceUrl":"https://www.latent.space/p/ainews-nano-banana-2-aka-gemini-31"},{"id":"fab47e77-812d-437a-9a66-5c7f41d42946","title":"AI is critical for humanity’s survival: Cisco president on the AI revolution | Jeetu Patel","summary":"Jeetu Patel, Cisco's president, discusses the pivotal role of AI in business and leadership, emphasizing frameworks for success and communication.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-26","topics":["ai-revolution","leadership","business-strategy","communication"],"diataxisType":"explanation","path":"/kb/articles/ai-is-critical-for-humanitys-survival-cisco-president-on-the-fab47e77","sourceUrl":"https://www.lennysnewsletter.com/p/ai-is-critical-for-humanitys-survival"},{"id":"ab98ceaf-e8f5-4f37-8d94-985cc85cab13","title":"How to Design Software With Weight","summary":"The article explores the design process behind Every's iOS app Monologue, highlighting the importance of creating a tactile user experience through thoughtful design choices.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-02-26","topics":["software-design","user-experience","ios-development","product-design"],"diataxisType":"how-to","path":"/kb/articles/how-to-design-software-with-weight-ab98ceaf","sourceUrl":"https://every.to/source-code/how-to-design-software-with-weight"},{"id":"c04f7376-0c79-4c03-9d82-e1404e55c441","title":"Google API Keys Weren’t Secrets. But then Gemini Changed the Rules.","summary":"The article discusses how Gemini's changes to API key privileges can lead to unintended access to sensitive data, highlighting the risks of using shared keys.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-26","topics":["api-keys","google-maps","gemini","security"],"diataxisType":"explanation","path":"/kb/articles/google-api-keys-werent-secrets-but-then-gemini-changed-the-r-c04f7376","sourceUrl":"https://simonwillison.net/2026/Feb/26/google-api-keys/#atom-everything"},{"id":"495218f8-db13-41a7-a796-c1461e587460","title":"Hoard things you know how to do - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"The article emphasizes the importance of hoarding knowledge and solutions in software development to enhance productivity and problem-solving capabilities.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-26","topics":["coding-agents","software-development","knowledge-management"],"diataxisType":"how-to","path":"/kb/articles/hoard-things-you-know-how-to-do-agentic-engineering-patterns-495218f8","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/hoard-things-you-know-how-to-do/#atom-everything"},{"id":"417198d8-d92b-4c53-951a-4dd4d5eb9c63","title":"A quote from Benedict Evans","summary":"Benedict Evans discusses the limited engagement users have with AI models and OpenAI's strategies to enhance user interaction and cover costs.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-26","topics":["ai-engagement","openai-strategy","user-experience"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-benedict-evans-417198d8","sourceUrl":"https://simonwillison.net/2026/Feb/26/benedict-evans/#atom-everything"},{"id":"abf70e4c-45fb-4bc3-a99d-b2e1beb696f5","title":"A quote from Andrej Karpathy","summary":"Andrej Karpathy discusses the transformative impact of AI on programming, highlighting significant advancements in coding agents since December 2025.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-26","topics":["ai-programming","coding-agents","ai-impact"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-andrej-karpathy-abf70e4c","sourceUrl":"https://simonwillison.net/2026/Feb/26/andrej-karpathy/#atom-everything"},{"id":"04433b95-1f0e-4af1-85d4-54a2bba2178d","title":"[LIVE] Anthropic Distillation & How Models Cheat (SWE-Bench Dead) | Nathan Lambert & Sebastian Raschka","summary":"The article discusses the significance of distillation in AI, particularly in the context of Chinese LLMs, during a live session featuring Nathan Lambert and Sebastian Raschka.","author":"Swyx","authorId":"swyx","published":"2026-02-26","topics":["ai-distillation","llms","us-china-tech","model-training"],"diataxisType":"explanation","path":"/kb/articles/live-anthropic-distillation-how-models-cheat-swe-bench-dead--04433b95","sourceUrl":"https://www.latent.space/p/paid-anthropic-distillation-and-how"},{"id":"0e4e25e9-e010-4df4-b926-b31102c6834d","title":"[AINews] WTF Happened in December 2025?","summary":"The article discusses significant AI developments in December 2025, highlighting the release of major models and their implications for the future of AI technology.","author":"Swyx","authorId":"swyx","published":"2026-02-26","topics":["ai-hype","model-releases","coding-agents","ai-ecosystem"],"diataxisType":"explanation","path":"/kb/articles/ainews-wtf-happened-in-december-2025-0e4e25e9","sourceUrl":"https://www.latent.space/p/wtf2025"},{"id":"8670162b-dc15-46cd-9fc6-35a7e2a4f4d8","title":"5 OpenClaw agents run my home, finances, and code | Jesse Genet","summary":"Jesse Genet utilizes five specialized OpenClaw agents to manage her home, finances, and coding projects, enhancing her productivity through automation.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-25","topics":["ai-agents","homeschooling","automation","productivity"],"diataxisType":"how-to","path":"/kb/articles/5-openclaw-agents-run-my-home-finances-and-code-jesse-genet-8670162b","sourceUrl":"https://www.lennysnewsletter.com/p/5-openclaw-agents-run-my-home-finances"},{"id":"306f627c-a5ee-4c75-b9b9-153c94983829","title":"Claude Code Remote Control","summary":"Claude Code has introduced a remote control feature for web interfaces, though it currently has some bugs and limitations that need addressing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-25","topics":["remote-control","claude-code","software-updates","ai-agents"],"diataxisType":"explanation","path":"/kb/articles/claude-code-remote-control-306f627c","sourceUrl":"https://simonwillison.net/2026/Feb/25/claude-code-remote-control/#atom-everything"},{"id":"cdfcfdee-6e66-4d91-8079-a163fe4ca4b6","title":"I vibe coded my dream macOS presentation app","summary":"Simon Willison shares his experience of creating a custom macOS presentation app called Present, designed for seamless web-based presentations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-25","topics":["macos-development","presentation-tools","swiftui","web-development"],"diataxisType":"how-to","path":"/kb/articles/i-vibe-coded-my-dream-macos-presentation-app-cdfcfdee","sourceUrl":"https://simonwillison.net/2026/Feb/25/present/#atom-everything"},{"id":"8e8765e0-a368-4594-b700-755bdd5ae892","title":"[AINews] The Unreasonable Effectiveness of Closing the Loop","summary":"The article discusses the shift in focus from outer loop tools to inner loop coding agents in the DevTools industry, highlighting recent launches and innovations.","author":"Swyx","authorId":"swyx","published":"2026-02-25","topics":["devtools","ai-agents","inner-loop","outer-loop","coding"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-unreasonable-effectiveness-of-closing-the-loop-8e8765e0","sourceUrl":"https://www.latent.space/p/ainews-the-unreasonable-effectiveness"},{"id":"e61bd491-185a-417b-8df3-8101cee2e219","title":"Linear walkthroughs - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"Simon Willison discusses using coding agents to create structured walkthroughs of codebases, highlighting the use of Showboat for generating detailed documentation.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-25","topics":["coding-agents","showboat","swiftui","vibe-coding","documentation"],"diataxisType":"how-to","path":"/kb/articles/linear-walkthroughs-agentic-engineering-patterns-simon-willi-e61bd491","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/linear-walkthroughs/#atom-everything"},{"id":"3df8c9b0-010a-4310-afe7-b3000d2bc014","title":"tldraw issue: Move tests to closed source repo","summary":"tldraw is moving its test suite to a private repository due to concerns over open source libraries being replicated, highlighting risks for commercial models.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-25","topics":["open-source","intellectual-property","software-development"],"diataxisType":"explanation","path":"/kb/articles/tldraw-issue-move-tests-to-closed-source-repo-3df8c9b0","sourceUrl":"https://simonwillison.net/2026/Feb/25/closed-tests/#atom-everything"},{"id":"56f44b8d-e875-4ed3-9561-985da54fc1b3","title":"AddyOsmani.com","summary":"Addy Osmani is a senior engineering leader at Google Cloud AI, focusing on AI-assisted engineering and developer tools, with a rich background in developer experiences at Google.","author":"Addyos Mani","authorId":"addyos-mani","published":"2026-02-25","topics":["ai-engineering","developer-tools","software-engineering"],"diataxisType":"explanation","path":"/kb/articles/addyosmanicom-56f44b8d","sourceUrl":"https://addyosmani.com/"},{"id":"5deef68f-17de-4c88-a8c4-d9c997eb8f6f","title":"A quote from Kellan Elliott-McCrea","summary":"Kellan Elliott-McCrea reflects on the emotional complexities of entering the tech industry, contrasting motivations of past and present generations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-25","topics":["technology","emotional-intelligence","web-development"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-kellan-elliott-mccrea-5deef68f","sourceUrl":"https://simonwillison.net/2026/Feb/25/kellan-elliott-mccrea/#atom-everything"},{"id":"bd43a0fc-3d9d-497b-bcc1-84ded5ececb3","title":"Amp","summary":"Amp is a coding agent designed for advanced coding tasks, offering a premium experience with features like composable code reviews and interactive diagrams.","author":"Ampcode","authorId":"ampcode","published":"2026-02-25","topics":["coding-agents","ai-tools","software-development"],"diataxisType":"explanation","path":"/kb/articles/amp-bd43a0fc","sourceUrl":"https://ampcode.com/"},{"id":"cb33a12e-6b0d-48d9-b5fe-3fbd00c9a3d7","title":"🔬Searching the Space of All Possible Materials — Prof. Max Welling, CuspAI","summary":"Prof. Max Welling discusses the intersection of AI, materials science, and climate impact, emphasizing the role of experiments as computation in advancing technology.","author":"Swyx","authorId":"swyx","published":"2026-02-25","topics":["ai-for-science","materials-discovery","quantum-gravity","equivariant-neural-networks"],"diataxisType":"explanation","path":"/kb/articles/searching-the-space-of-all-possible-materials-prof-max-welli-cb33a12e","sourceUrl":"https://www.latent.space/p/cuspai"},{"id":"9af4dec4-120c-4b41-bae2-0bb08898fd5d","title":"HowtoBuildanAgent","summary":"It’s an LLM, a loop, and enough tokens. It’s not that hard to build a fully functioning, code-editing agent.","author":"Ampcode","authorId":"ampcode","published":"2026-02-25","topics":["Prompt Engineering","Anthropic API","AI Agents"],"diataxisType":"tutorial","path":"/kb/articles/howtobuildanagent-9af4dec4","sourceUrl":"https://ampcode.com/notes/how-to-build-an-agent"},{"id":"6cf6dad3-f80c-445b-bdb8-254107f8312a","title":"Blog","summary":"The OpenAI Developer Blog provides insights and updates for developers working with OpenAI technologies, featuring practical tips and lessons learned.","author":"OpenAI Engineering","authorId":"openai-engineering","published":"2026-02-25","topics":["developer-insights","codex","api","apps-sdk","audio"],"diataxisType":"reference","path":"/kb/articles/blog-6cf6dad3","sourceUrl":"https://developers.openai.com/blog"},{"id":"5acd0fdd-253d-43ea-a257-4e07b328a234","title":"Claude Code for Finance + The Global Memory Shortage: Doug O'Laughlin, SemiAnalysis","summary":"The article discusses the impact of Claude Code on finance and AI, highlighting its integration into workflows and the ongoing global memory shortage.","author":"Swyx","authorId":"swyx","published":"2026-02-24","topics":["ai-finance","claude-code","memory-shortage","podcast"],"diataxisType":"explanation","path":"/kb/articles/claude-code-for-finance-the-global-memory-shortage-doug-olau-5acd0fdd","sourceUrl":"https://www.latent.space/p/valuemule"},{"id":"9c11adc4-1eeb-43c3-be97-18b0514c313e","title":"First run the tests - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"The article emphasizes the importance of automated testing in coding agents, highlighting how it enhances code reliability and aids agents in understanding existing codebases.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-24","topics":["automated-testing","ai-agents","software-engineering"],"diataxisType":"how-to","path":"/kb/articles/first-run-the-tests-agentic-engineering-patterns-simon-willi-9c11adc4","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/first-run-the-tests/#atom-everything"},{"id":"dae507b8-ff78-4ecb-960a-cad857754208","title":"[AINews] Anthropic accuses DeepSeek, Moonshot, and MiniMax of >16 million \"industrial-scale distillation attacks\"","summary":"Anthropic accuses DeepSeek, Moonshot, and MiniMax of conducting over 16 million industrial-scale distillation attacks to extract capabilities for their own models.","author":"Swyx","authorId":"swyx","published":"2026-02-24","topics":["ai-distillation","industrial-scale-attacks","geopolitical-implications"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-accuses-deepseek-moonshot-and-minimax-of-16-dae507b8","sourceUrl":"https://www.latent.space/p/ainews-anthropic-accuses-deepseek"},{"id":"ebba2249-c49a-44ba-ad54-61ed751bebb6","title":"go-size-analyzer","summary":"The go-size-analyzer is a tool for analyzing the size of Go binaries with a treemap view of dependencies, available both locally and in the browser.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-24","topics":["go-tooling","binary-analysis","webassembly"],"diataxisType":"explanation","path":"/kb/articles/go-size-analyzer-ebba2249","sourceUrl":"https://simonwillison.net/2026/Feb/24/go-size-analyzer/#atom-everything"},{"id":"d16a3274-5df8-4920-bbd7-4ec4439ad7ff","title":"How to use AI for your next job interview","summary":"The article discusses how AI can enhance job interview preparation by providing feedback and predicting questions, featuring insights from tech professionals.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-24","topics":["ai-interview-coach","job-preparation","tech-professionals","interview-strategies"],"diataxisType":"how-to","path":"/kb/articles/how-to-use-ai-for-your-next-job-interview-d16a3274","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-use-ai-for-your-next-job-interview"},{"id":"43585643-f934-4ceb-ba62-ae8b185ee6c3","title":"How to use AI for your next job interview","summary":"The article explores how AI is transforming job interview preparation for candidates, offering insights from over 30 tech professionals on effective AI workflows.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-24","topics":["ai-interview-preparation","job-seeking-strategies","feedback-loops","career-advancement"],"diataxisType":"how-to","path":"/kb/articles/how-to-use-ai-for-your-next-job-interview-43585643","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-use-ai-in-your-next-job-interview"},{"id":"78f0d530-f14d-48ee-94fd-5c5e6c586eed","title":"Newsroom","summary":"Anthropic announces the upgrade of its model to Claude Opus 4.6 and raises $30 billion in Series G funding, valuing the company at $380 billion.","author":"Resources Anthropic","authorId":"resources-anthropic","published":"2026-02-24","topics":["ai-models","funding-news","enterprise-ai"],"diataxisType":"explanation","path":"/kb/articles/newsroom-78f0d530","sourceUrl":"https://resources.anthropic.com/"},{"id":"16b49489-235d-41ae-8d28-58bc62b855b4","title":"Writing about Agentic Engineering Patterns","summary":"Simon Willison introduces Agentic Engineering Patterns, a project focused on coding practices for effective use of coding agents in software development.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["agentic-engineering","coding-agents","software-development","ai-assisted-programming"],"diataxisType":"explanation","path":"/kb/articles/writing-about-agentic-engineering-patterns-16b49489","sourceUrl":"https://simonwillison.net/2026/Feb/23/agentic-engineering-patterns/#atom-everything"},{"id":"b4384d05-6550-4c5c-a776-93f36a212ba4","title":"A quote from Summer Yue","summary":"Summer Yue shares a humorous experience of trying to manage her inbox with an AI tool, highlighting the challenges of technology and user instructions.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["ai-tools","inbox-management","automation"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-summer-yue-b4384d05","sourceUrl":"https://simonwillison.net/2026/Feb/23/summer-yue/#atom-everything"},{"id":"66e96290-70fa-4852-988b-aaa233c9d1a7","title":"Writing code is cheap now - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"The article discusses the implications of cheap code production on engineering practices, emphasizing the need for new habits to ensure code quality amidst evolving tools.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["agentic-engineering","software-development","code-quality"],"diataxisType":"explanation","path":"/kb/articles/writing-code-is-cheap-now-agentic-engineering-patterns-simon-66e96290","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/code-is-cheap/#atom-everything"},{"id":"46eff22b-3cc5-4115-ab3e-8ebd179b59d7","title":"⚡️The End of SWE-Bench Verified — Mia Glaese & Olivia Watkins, OpenAI Frontier Evals & Human Data","summary":"OpenAI announces the discontinuation of SWE-Bench Verified due to saturation and contamination issues, endorsing SWE-Bench Pro as the new standard.","author":"Swyx","authorId":"swyx","published":"2026-02-23","topics":["swe-bench","openai","ai-evaluation","machine-learning"],"diataxisType":"explanation","path":"/kb/articles/the-end-of-swe-bench-verified-mia-glaese-olivia-watkins-open-46eff22b","sourceUrl":"https://www.latent.space/p/swe-bench-dead"},{"id":"e8f8f0fd-49dd-4f79-8be5-be789c4521da","title":"Red/green TDD - Agentic Engineering Patterns - Simon Willison's Weblog","summary":"The article discusses the benefits of red/green Test Driven Development (TDD) for coding agents, emphasizing the importance of writing tests first to ensure code quality.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["test-driven-development","coding-agents","software-engineering"],"diataxisType":"explanation","path":"/kb/articles/redgreen-tdd-agentic-engineering-patterns-simon-willisons-we-e8f8f0fd","sourceUrl":"https://simonwillison.net/guides/agentic-engineering-patterns/red-green-tdd/#atom-everything"},{"id":"a624fb3a-f582-4dd2-805e-a5c76f3d9890","title":"A quote from Paul Ford","summary":"Simon Willison reflects on Paul Ford's insights about vibe coding and the challenges of communicating complex ideas to the public.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["vibe-coding","communication","empathy","writing"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-paul-ford-a624fb3a","sourceUrl":"https://simonwillison.net/2026/Feb/23/paul-ford/#atom-everything"},{"id":"dd345ddf-0e36-42d2-a5a6-528b32cbba4d","title":"Reply guy","summary":"The article discusses the rise of AI bots on Twitter that provide generic replies to tweets, termed 'reply guy tools', and their impact on user engagement.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["ai-bots","twitter-engagement","generative-ai"],"diataxisType":"explanation","path":"/kb/articles/reply-guy-dd345ddf","sourceUrl":"https://simonwillison.net/2026/Feb/23/reply-guy/#atom-everything"},{"id":"3a27394f-ff28-4589-b692-373e6b6046d0","title":"🎙️ This week on How I AI: How Notion’s design team uses Claude Code to design","summary":"Brian Lovin from Notion discusses how the design team utilizes Claude Code for code-first prototyping, enhancing collaboration and efficiency in design workflows.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-23","topics":["ai-design","prototyping","workflow-automation","notion"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-how-notions-design-team-uses-claude-c-3a27394f","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-how-notions"},{"id":"c159f4a6-2d82-4c3a-adba-3dbc82b8c319","title":"Ladybird adopts Rust, with help from AI","summary":"Ladybird transitioned from Swift to Rust for their JavaScript engine, utilizing AI tools for an efficient porting process with zero regressions.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-23","topics":["rust","ai-assisted-coding","software-engineering","libjs"],"diataxisType":"explanation","path":"/kb/articles/ladybird-adopts-rust-with-help-from-ai-c159f4a6","sourceUrl":"https://simonwillison.net/2026/Feb/23/ladybird-adopts-rust/#atom-everything"},{"id":"a2e3c1be-82ba-4fca-8b9d-3732b396a640","title":"“I haven’t written a single line of front-end code in 3 months”: How Notion’s design team uses Claude Code to prototype","summary":"Brian Lovin from Notion AI discusses how his team uses Claude Code for collaborative prototyping, enhancing the design process with AI-assisted coding.","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-23","topics":["ai-prototyping","design-process","collaborative-tools"],"diataxisType":"how-to","path":"/kb/articles/i-havent-written-a-single-line-of-front-end-code-in-3-months-a2e3c1be","sourceUrl":"https://www.lennysnewsletter.com/p/i-havent-written-a-single-line-of"},{"id":"c77aff7b-69c0-4556-a6f7-1da473f87d38","title":"How I think about Codex","summary":"Simon Willison discusses the concept of Codex, explaining its components and how it functions as OpenAI's software engineering agent.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-22","topics":["openai","codex","ai-agents","software-engineering"],"diataxisType":"explanation","path":"/kb/articles/how-i-think-about-codex-c77aff7b","sourceUrl":"https://simonwillison.net/2026/Feb/22/how-i-think-about-codex/#atom-everything"},{"id":"0fa21631-4d74-4b75-b41d-3bfb641d69b4","title":"London Stock Exchange: Raspberry Pi Holdings plc","summary":"Raspberry Pi Holdings plc's stock surged 30% amid excitement over its use in AI projects, following a significant purchase by CEO Eben Upton.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-22","topics":["stock-market","ai-projects","raspberry-pi"],"diataxisType":"explanation","path":"/kb/articles/london-stock-exchange-raspberry-pi-holdings-plc-0fa21631","sourceUrl":"https://simonwillison.net/2026/Feb/22/raspberry-pi-openclaw/#atom-everything"},{"id":"1f635a54-1202-43a9-a495-1aa1fc9c78f2","title":"The Claude C Compiler: What It Reveals About the Future of Software","summary":"The article discusses the Claude C Compiler and its implications for software development, highlighting the role of AI in coding and the challenges it presents.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-22","topics":["ai-coding","software-development","compiler-design","open-source","intellectual-property"],"diataxisType":"explanation","path":"/kb/articles/the-claude-c-compiler-what-it-reveals-about-the-future-of-so-1f635a54","sourceUrl":"https://simonwillison.net/2026/Feb/22/ccc/#atom-everything"},{"id":"e5079680-7ecc-41e9-ad28-73b95a5041fe","title":"Applied AI Engineering - Issue #1 | Feb 22, 2026","summary":"Agent deployments are moving from prompt hacks toward engineered workflows with guardrails. * [Report this article](/uas/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Fpulse%2Fapplied-ai-engineering-issue-1-feb-22-2026-unmesh-gundecha-n9qvc&trk=article-ssr-frontend-pulse_ellipsis-menu-semap","author":"Unmesh Gundecha","authorId":"unmesh-gundecha","published":"2026-02-22","topics":["Agent-Native Architecture","Prompt Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/applied-ai-engineering-issue-1-feb-22-2026-e5079680","sourceUrl":"https://www.linkedin.com/pulse/applied-ai-engineering-issue-1-feb-22-2026-unmesh-gundecha-n9qvc/"},{"id":"d9d054b3-f8c3-4faa-bcd0-d7df30a9a14b","title":"Raising the bar on SWE-bench Verified with Claude 3.5 Sonnet","summary":"Our latest model, the upgraded Claude 3.5 Sonnet, achieved 49% on SWE-bench Verified. SWE-bench is an AI evaluation benchmark that assesses a model's ability to complete real-world software engineering tasks.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-21","topics":["Prompt Engineering","AI Agents","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/raising-the-bar-on-swe-bench-verified-with-claude-35-sonnet-d9d054b3","sourceUrl":"https://www.anthropic.com/engineering/swe-bench-sonnet"},{"id":"28f9178b-f21c-4fb5-a1a5-8d6c7142b97d","title":"Introducing Contextual Retrieval","summary":"For an AI model to be useful in specific contexts, it often needs access to background knowledge. For an AI model to be useful in specific contexts, it often needs access to background knowledge.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-21","topics":["Prompt Engineering","AI Agents","RAG"],"diataxisType":"explanation","path":"/kb/articles/introducing-contextual-retrieval-28f9178b","sourceUrl":"https://www.anthropic.com/engineering/contextual-retrieval"},{"id":"f9aa6dd4-fcd9-46db-a3b8-5d2a635c2723","title":"[AINews] The Custom ASIC Thesis","summary":"The article discusses the potential of custom ASICs for AI models, emphasizing their economic viability and the future of integrated model-chip design.","author":"Swyx","authorId":"swyx","published":"2026-02-21","topics":["custom-asic","ai-inference","model-architecture","llm-optimization"],"diataxisType":"explanation","path":"/kb/articles/ainews-the-custom-asic-thesis-f9aa6dd4","sourceUrl":"https://www.latent.space/p/ainews-the-custom-asic-thesis"},{"id":"53a6ac3d-c99a-4de8-a606-16108c54c9a7","title":"Quoting Thibault Sottiaux","summary":"The article discusses the insights and perspectives of Thibault Sottiaux, highlighting his contributions to the field.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-21","topics":["thought-leadership","interviews","technology"],"diataxisType":"explanation","path":"/kb/articles/quoting-thibault-sottiaux-53a6ac3d","sourceUrl":"https://simonwillison.net/2026/Feb/21/thibault-sottiaux/#atom-everything"},{"id":"220f4a23-4c85-4506-86c3-2f3864368000","title":"The \"think\" tool: Enabling Claude to stop and think in complex tool use situations","summary":"A new tool that improves Claude's complex problem-solving performance","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-21","topics":["AI","Technology"],"diataxisType":"how-to","path":"/kb/articles/the-think-tool-enabling-claude-to-stop-and-think-in-complex--220f4a23","sourceUrl":"https://www.anthropic.com/engineering/claude-think-tool"},{"id":"64f2e1e0-ab01-4fe1-a5c0-e67a322a5ef0","title":"Claude Code overview","summary":"Claude Code is an agentic coding tool that reads your codebase, edits files, runs commands, and integrates with your development tools. * [Overview](/docs/en/overview)\n* [Quickstart](/docs/en/quickstart)\n* [Changelog](https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md)","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-21","topics":["Prompt Engineering","AI Agents","Claude Code"],"diataxisType":"reference","path":"/kb/articles/claude-code-overview-64f2e1e0","sourceUrl":"https://www.anthropic.com/engineering/claude-code-best-practices"},{"id":"ddc306e9-ed48-4115-9cef-12196d76b637","title":"🧠 Community Wisdom: Making your product enterprise-ready, when engineering outpaces product, managing PRD versions as features evolve, planning work around energy, and more","summary":"👋 Hello and welcome to this week’s edition of ✨ Community Wisdom ✨ *👋 Hello and welcome to this week’s edition of ✨ **Community Wisdom** ✨ a subscriber-only email, delivered every Saturday, highlighting the most helpful conversations in [our members-only Slack community](https://forms.gle/syq81QDw","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-21","topics":["Project Management","Product Development","Enterprise Software","Work Planning"],"diataxisType":"explanation","path":"/kb/articles/-community-wisdom-making-your-product-enterprise-ready-when--ddc306e9","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-making-your-product"},{"id":"08424d27-18f5-4bf9-bef7-aa17357fe25f","title":"Andrej Karpathy talks about “Claws”","summary":"Andrej Karpathy discusses the emerging concept of 'Claws', a new layer of AI agents that enhance orchestration and scheduling, while sharing insights on various implementations.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-21","topics":["ai-agents","claws","openclaw","technology"],"diataxisType":"explanation","path":"/kb/articles/andrej-karpathy-talks-about-claws-08424d27","sourceUrl":"https://simonwillison.net/2026/Feb/21/claws/#atom-everything"},{"id":"63033226-78c3-4a19-b8a5-e50a0e14f774","title":"Massively Parallel Procrastination","summary":"Jesse Chen discusses various aspects of AI software development and the challenges of managing agent interactions in his blog posts.","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-20","topics":["ai-software-development","agent-management","prompt-injection","crypto-scams"],"diataxisType":"explanation","path":"/kb/articles/massively-parallel-procrastination-63033226","sourceUrl":"https://blog.fsck.com/"},{"id":"52976c37-92d3-437f-b709-1ff132e5f77a","title":"[AINews] Gemini 3.1 Pro: 2x 3.0 on ARC-AGI 2","summary":"It’s getting a little hard to say interesting things with all the round robin minor version updates of frontier models every week, but Gemini 3.1 Pro seems like a decent enough advance. > *AI News for 2/18/2026-2/19/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/158543024","author":"Swyx","authorId":"swyx","published":"2026-02-20","topics":["Prompt Engineering","Agent-Native Architecture","AI Agents"],"diataxisType":"reference","path":"/kb/articles/ainews-gemini-31-pro-2x-30-on-arc-agi-2-52976c37","sourceUrl":"https://www.latent.space/p/ainews-gemini-31-pro-2x-30-on-arc"},{"id":"915b03d0-1bc3-478b-ba86-34a5a56b3ee2","title":"ggml.ai joins Hugging Face to ensure the long-term progress of Local AI","summary":"It's hard to overstate the impact Georgi Gerganov has had on the local model space. I don't normally cover acquisition news like this, but I have some thoughts.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-20","topics":["OpenAI API","AI Agents","Anthropic API"],"diataxisType":"explanation","path":"/kb/articles/ggmlai-joins-hugging-face-to-ensure-the-long-term-progress-o-915b03d0","sourceUrl":"https://simonwillison.net/2026/Feb/20/ggmlai-joins-hugging-face/#atom-everything"},{"id":"67189a5f-e02d-4903-8713-a67205b53503","title":"Taalas serves Llama 3.1 8B at 17,000 tokens/second","summary":"This new Canadian hardware startup just announced their first product - a custom hardware implementation of the Llama 3.1 8B model that can run at a staggering 17,000 tokens/second. This new Canadian hardware startup just announced their first product - a custom hardware implementation of the Llama ","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-20","topics":["AI Agents","Model Optimization","Hardware Startups"],"diataxisType":"explanation","path":"/kb/articles/taalas-serves-llama-31-8b-at-17000-tokenssecond-67189a5f","sourceUrl":"https://simonwillison.net/2026/Feb/20/taalas/#atom-everything"},{"id":"a07c05eb-3abb-40eb-abc0-f9768c53bf22","title":"Engineering","summary":"The article discusses advancements in agentic coding and infrastructure noise, highlighting various technical evaluations and best practices for AI agents.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-20","topics":["ai-agents","coding-evaluations","technical-evaluations","agentic-coding"],"diataxisType":"explanation","path":"/kb/articles/engineering-a07c05eb","sourceUrl":"https://www.anthropic.com/engineering/"},{"id":"4ff8e177-91b0-401b-8d90-20112b3e961f","title":"Adding TILs, releases, museums, tools and research to my blog","summary":"Beats are little inline links with badges that fit into different content timeline views around my site. I've been wanting to add indications of my various other online activities to my blog for a while now. I just turned on a new feature I'm calling \"beats\" (after story beats, naming this was hard!","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-20","topics":["Open Source Projects","Web Development","AI Agents","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/adding-tils-releases-museums-tools-and-research-to-my-blog-4ff8e177","sourceUrl":"https://simonwillison.net/2026/Feb/20/beats/#atom-everything"},{"id":"9bb99b58-ec00-4c75-bd77-46d4c4cbba28","title":"Quoting Thariq Shihipar","summary":"Long running agentic products like Claude Code are made feasible by prompt caching which allows us to reuse computation from previous roundtrips and significantly decrease latency and cost. > Long running agentic products like Claude Code are made feasible by prompt caching which allows us to reuse ","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-20","topics":["Agent-Native Architecture","Prompt Engineering","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/quoting-thariq-shihipar-9bb99b58","sourceUrl":"https://simonwillison.net/2026/Feb/20/thariq-shihipar/#atom-everything"},{"id":"8293fcc7-ed88-475e-b7de-139f54cb5cee","title":"Hi | Kieran Klaassen","summary":"Kieran Klaassen shares his journey as a creator and engineer, emphasizing his passion for crafting and inspiring authenticity.","author":"Kieran Klaassen","authorId":"kieran-klaassen","published":"2026-02-20","topics":["creativity","engineering","baking","self-discovery"],"diataxisType":"explanation","path":"/kb/articles/hi-kieran-klaassen-8293fcc7","sourceUrl":"https://www.kieranklaassen.com/"},{"id":"92e1146e-bcfe-4f4f-8470-917e64d7d7ae","title":"Experimenting with sponsorship for my blog and newsletter","summary":"I've long been resistant to the idea of accepting sponsorship for my blog. I've long been resistant to the idea of accepting sponsorship for my blog. I value my credibility as an independent voice, and I don't want to risk compromising that reputation.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-19","topics":["Blogging","Sponsorship","Independent Publishing"],"diataxisType":"explanation","path":"/kb/articles/experimenting-with-sponsorship-for-my-blog-and-newsletter-92e1146e","sourceUrl":"https://simonwillison.net/2026/Feb/19/sponsorship/#atom-everything"},{"id":"6a6656f2-ea9f-4852-b5e7-276fef839fb9","title":"Bitter Lessons in Venture vs Growth: Anthropic vs OpenAI, Noam Shazeer, World Labs, Thinking Machines, Cursor, ASIC Economics — Martin Casado & Sarah Wang of a16z","summary":"why today’s rounds are really compute contracts in disguise *Tickets for [AIEi Miami](https://www.ai.engineer/miami) and [AIE Europe](https://www.ai.engineer/europe) are live, with **first wave speakers announced**!*","author":"Swyx","authorId":"swyx","published":"2026-02-19","topics":["AI Financing Models","Enterprise Software in AI","Future of AI Market Structures","Talent Wars in AI"],"diataxisType":"explanation","path":"/kb/articles/bitter-lessons-in-venture-vs-growth-anthropic-vs-openai-noam-6a6656f2","sourceUrl":"https://www.latent.space/p/a16z"},{"id":"9c26a6a0-332c-409e-b1ee-b1129c83390c","title":"Head of Claude Code: What happens after coding is solved | Boris Cherny","summary":"What began as a simple terminal-based prototype just a year ago has transformed the role of software engineering. 1. How Claude Code grew from a quick hack to 4% of public GitHub commits, with daily active users doubling last month\n2. The counterintuitive product principles that drove Claude Code’s ","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-19","topics":["Claude Code","AI in Professional Work","Anthropic","Product Development Principles","Software Engineering"],"diataxisType":"explanation","path":"/kb/articles/head-of-claude-code-what-happens-after-coding-is-solved-bori-9c26a6a0","sourceUrl":"https://www.lennysnewsletter.com/p/head-of-claude-code-what-happens"},{"id":"126b8d87-34fa-4202-a710-630bc3c97363","title":"[AINews] Anthropic's Agent Autonomy study","summary":"Most Claude Code turns are short. The median turn lasts around 45 seconds. There’s a lot of small tidbits going on, with former guest [Fei-Fei Li’s World Labs](https://finance.yahoo.com/news/ai-pioneer-fei-fei-lis-192214332.html?guccounter=1&guce_referrer=aHR0cHM6Ly9rYWdpLmNvbS8&guce_referrer_sig=AQ","author":"Swyx","authorId":"swyx","published":"2026-02-19","topics":["Claude Code","AI Agents","Anthropic API"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropics-agent-autonomy-study-126b8d87","sourceUrl":"https://www.latent.space/p/ainews-anthropics-agent-autonomy"},{"id":"4dd505b8-5c4d-4acc-b74a-aae68bcc8d97","title":"LadybirdBrowser/ladybird: Abandon Swift adoption","summary":"Everywhere: Abandon Swift adoption Back [in August 2024](https://simonwillison.net/2024/Aug/11/ladybird-set-to-adopt-swift/) the Ladybird browser project announced an intention to adopt Swift as their memory-safe language of choice.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-19","topics":["Software Development","Programming Languages","Project Management"],"diataxisType":"explanation","path":"/kb/articles/ladybirdbrowserladybird-abandon-swift-adoption-4dd505b8","sourceUrl":"https://simonwillison.net/2026/Feb/19/ladybird/#atom-everything"},{"id":"5fd79031-efba-4032-a4c9-3453a639b769","title":"SWE-bench February 2026 leaderboard update","summary":"SWE-bench is one of the benchmarks that the labs love to list in their model releases. SWE-bench is one of the benchmarks that the labs love to list in their model releases. The official leaderboard is infrequently updated but they just did a full run of it against the current generation of models, ","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-19","topics":["OpenAI API","Prompt Engineering","AI Agents"],"diataxisType":"reference","path":"/kb/articles/swe-bench-february-2026-leaderboard-update-5fd79031","sourceUrl":"https://simonwillison.net/2026/Feb/19/swe-bench/#atom-everything"},{"id":"d3069d40-916b-4d98-8f14-a623d1159592","title":"Gemini 3.1 Pro","summary":"That's less than half the price of Claude Opus 4.6 with very similar benchmark scores to that model. The first in the Gemini 3.1 series, priced the same as Gemini 3 Pro ($2/million input, $12/million output under 200,000 tokens, $4/$18 for 200,000 to 1,000,000). That's less than half the price of Cl","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-19","topics":["Prompt Engineering","AI Agents","LangChain"],"diataxisType":"reference","path":"/kb/articles/gemini-31-pro-d3069d40","sourceUrl":"https://simonwillison.net/2026/Feb/19/gemini-31-pro/#atom-everything"},{"id":"6b96427c-639d-4bfd-8ec8-c802fcf1cd84","title":"Recovering lost code","summary":"Reached the stage of parallel agent psychosis where I've lost a whole feature. Reached the stage of parallel agent psychosis where I've lost a whole feature - I know I had it yesterday, but I can't seem to find the branch or worktree or cloud instance or checkout with it in.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-19","topics":["Prompt Engineering","Agent-Native Architecture","Claude Code"],"diataxisType":"how-to","path":"/kb/articles/recovering-lost-code-6b96427c","sourceUrl":"https://simonwillison.net/2026/Feb/19/recovering-lost-code/#atom-everything"},{"id":"a5a5dffc-1776-4150-91de-9bd680d78b5b","title":"Compound Engineering: Make Every Unit of Work Compound Into the Next","summary":"Design is easier to iterate on in code than in mockups—you can click through it and feel the interactions. The flag is named `--dangerously-skip-permissions` on purpose. It's meant to make you pause the first time. But once you are more experienced, you can make an informed decision about your risk ","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-02-18","topics":["Prompt Engineering","AI Agents","Design Systems"],"diataxisType":"how-to","path":"/kb/articles/compound-engineering-make-every-unit-of-work-compound-into-t-a5a5dffc","sourceUrl":"https://every.to/guides/compound-engineering"},{"id":"a6d5bdba-58bb-4d79-8a01-85562a09e005","title":"The A.I. Disruption We’ve Been Waiting for Has Arrived","summary":"The article discusses the transformative impact of advanced A.I. coding tools, highlighting personal experiences and the evolving perceptions within the tech community.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-18","topics":["ai-disruption","coding-tools","technology-impact","software-development"],"diataxisType":"explanation","path":"/kb/articles/the-ai-disruption-weve-been-waiting-for-has-arrived-a6d5bdba","sourceUrl":"https://simonwillison.net/2026/Feb/18/the-ai-disruption/#atom-everything"},{"id":"19c500da-42cb-4b53-bd9d-cc94204bf361","title":"[AINews] Claude Sonnet 4.6: clean upgrade of 4.5, mostly better with some caveats","summary":"Anthropic opted to launch Sonnet 4.6 today, bumping their cheaper workhorse model up to match Opus 4.6. > *AI News for 2/16/2026-2/17/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**261** channels, and **11323** messages) for you. Es","author":"Swyx","authorId":"swyx","published":"2026-02-18","topics":["Prompt Engineering","AI Agents","Anthropic API"],"diataxisType":"explanation","path":"/kb/articles/ainews-claude-sonnet-46-clean-upgrade-of-45-mostly-better-wi-19c500da","sourceUrl":"https://www.latent.space/p/ainews-claude-sonnet-46-clean-upgrade"},{"id":"bc799cfd-57af-452b-b2a7-f16c6607a6fc","title":"[patterns] Beads: the architecture of a graph issue tracker for AI agents","summary":"How a version-controlled database turns issue tracking into persistent agent memory. How a version-controlled database turns issue tracking into persistent agent memory — a chapter in the style of The Architecture of Open Source Applications.","author":"Laten Patterns","authorId":"latent-patterns","published":"2026-02-18","topics":["Version Control","Agent-Native Architecture","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/patterns-beads-the-architecture-of-a-graph-issue-tracker-for-bc799cfd","sourceUrl":"https://latentpatterns.com/patterns/beads-graph-tracker"},{"id":"1ea8a792-d512-4710-8bf1-600eaf12210b","title":"[patterns] Beads: the architecture of a graph issue tracker for AI agents","summary":"The article discusses how a version-controlled database enhances issue tracking by creating a persistent memory for AI agents, inspired by open source architecture.","author":"Laten Patterns","authorId":"latent-patterns","published":"2026-02-18","topics":["ai-agents","issue-tracking","open-source","database-architecture"],"diataxisType":"explanation","path":"/kb/articles/patterns-beads-the-architecture-of-a-graph-issue-tracker-for-1ea8a792","sourceUrl":"http://localhost:5173/patterns/beads-graph-tracker"},{"id":"a581c385-01c5-48e7-bd2e-9e8a356b0f61","title":"A quote from Martin Fowler","summary":"Martin Fowler discusses the impact of LLMs on software development, questioning the future of specialist roles versus the rise of Expert Generalists.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-18","topics":["llms","software-development","expert-generalists"],"diataxisType":"explanation","path":"/kb/articles/a-quote-from-martin-fowler-a581c385","sourceUrl":"https://simonwillison.net/2026/Feb/18/martin-fowler/#atom-everything"},{"id":"f8eb5057-6321-4ec4-906b-c388f4dce203","title":"Typing without having to type","summary":"25+ years into my career as a programmer I think I may finally be coming around to preferring type hints or even strong typing. 25+ years into my career as a programmer I think I may *finally* be coming around to preferring type hints or even strong typing. I resisted those in the past because they ","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-18","topics":["Programming Productivity","Type Hints","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/typing-without-having-to-type-f8eb5057","sourceUrl":"https://simonwillison.net/2026/Feb/18/typing/#atom-everything"},{"id":"cb38bd6d-8bf8-4fbc-8313-537246ca2fb0","title":"Agent-nativeArchitectures","summary":"Architecture decision: Store full agent configuration, or store only `agentType` and recreate from a registry. Bool {\n Date().timeIntervalSince(timestamp)\n < maxAge\n}","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-02-18","topics":["Storage Abstraction","CRUD Completeness","Anti-patterns in Agent Design","Agent-Native Architecture","Dynamic Capability Discovery"],"diataxisType":"how-to","path":"/kb/articles/agent-nativearchitectures-cb38bd6d","sourceUrl":"https://every.to/guides/agent-native"},{"id":"3d5bc9df-a78e-4071-bb20-76dbb175dabc","title":"How to Build Agent-native: Lessons From Four Apps","summary":"Nobody programmed it to do any of this. Instead, Dan’s app has a handful of basic tools—“read file,” “write file,” and “search the web”—and an AI agent smart enough to combine them in a way that matches the user’s request. When it generates a summary, for example, that’s the agent deciding on its ow","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-02-17","topics":["Agent-Native Architecture","AI Agents","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/how-to-build-agent-native-lessons-from-four-apps-3d5bc9df","sourceUrl":"https://every.to/source-code/how-to-build-agent-native-lessons-from-four-apps"},{"id":"cffdf1dd-fb76-4d47-804c-13c511c5fe16","title":"Introducing Claude Sonnet 4.6","summary":"Claude Sonnet 4.6 has been released, offering competitive performance with a reliable knowledge cutoff and new features, while maintaining affordable pricing.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-17","topics":["ai-models","anthropic","sonnet-4-6","opus-4-5"],"diataxisType":"explanation","path":"/kb/articles/introducing-claude-sonnet-46-cffdf1dd","sourceUrl":"https://simonwillison.net/2026/Feb/17/claude-sonnet-46/#atom-everything"},{"id":"5062827b-e961-4027-9bfe-018dd877c6cd","title":"How to do AI analysis you can actually trust","summary":"The problem with AI is that the output always looks confident—even when it’s full of lies. *👋 Hey there, I’m Lenny. Each week, I answer reader questions about building product, driving growth, and accelerating your career. For more: [Lenny’s Podcast](https://www.lennysnewsletter.com/podcast) | [Len","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-17","topics":["Prompt Engineering","AI Agents","Customer Research"],"diataxisType":"how-to","path":"/kb/articles/how-to-do-ai-analysis-you-can-actually-trust-5062827b","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-do-ai-analysis-you-can-actually"},{"id":"f6ef390b-4a8f-4ce6-b797-86a599d41e5f","title":"Rodney v0.4.0","summary":"Rodney v0.4.0 introduces several new features for the CLI tool, including improved error handling, new commands for testing, and enhanced support for Windows.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-17","topics":["browser-automation","cli-tools","software-release"],"diataxisType":"reference","path":"/kb/articles/rodney-v040-f6ef390b","sourceUrl":"https://simonwillison.net/2026/Feb/17/rodney/#atom-everything"},{"id":"44bb07f5-acfd-406a-a0cf-b709afe79c88","title":"[AINews] Qwen3.5-397B-A17B: the smallest Open-Opus class, very efficient model","summary":"AI News for 2/13/2026-2/16/2026. We checked 12 subreddits, 544 Twitters and 24 Discords for you. > *AI News for 2/13/2026-2/16/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**261** channels, and **26057** messages) for you. Estimated","author":"Swyx","authorId":"swyx","published":"2026-02-17","topics":["OpenAI API","AI Agents","Anthropic API"],"diataxisType":"reference","path":"/kb/articles/ainews-qwen35-397b-a17b-the-smallest-open-opus-class-very-ef-44bb07f5","sourceUrl":"https://www.latent.space/p/ainews-qwen35-397b-a17b-the-smallest"},{"id":"52ad310d-cbf1-4ff2-8e92-867ff1b56df6","title":"How to do AI analysis you can actually trust","summary":"Ever run an AI analysis on customer data, only to discover the numbers were fabricated and the insights completely generic? [](https://substackcdn.com/image/fetch/$s_!OYz9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7d4c0b48-1813-44","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-17","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"how-to","path":"/kb/articles/how-to-do-ai-analysis-you-can-actually-trust-52ad310d","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-do-ai-analysis-you-can-actually-db6"},{"id":"4ba96001-0e5a-4fdf-8403-3aed6c4c6382","title":"[blog] Introducing the Latent Patterns Blog","summary":"Why we're launching a blog series Why we're launching a blog series, what we'll publish, and how to subscribe via RSS.","author":"Laten Patterns","authorId":"latent-patterns","published":"2026-02-16","topics":["Blogging","RSS Subscription","Content Strategy"],"diataxisType":"explanation","path":"/kb/articles/blog-introducing-the-latent-patterns-blog-4ba96001","sourceUrl":"https://latentpatterns.com/blog/introducing-the-latent-patterns-blog"},{"id":"2f3d6bbc-e693-4ee0-8b44-385e929b657c","title":"🎙️ This week on How I AI: Opus vs. Codex showdown, and AI for accessibility","summary":"It’s not Opus or Codex. It’s Opus for building and Codex for reviewing. [](https://substackcdn.com/image/fetch/$s_!gWeJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F361d81ef-7faf-4d8e-8028-5d5e03432a9a_2329x551.png)","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-16","topics":["Prompt Engineering","Claude Code","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/-this-week-on-how-i-ai-opus-vs-codex-showdown-and-ai-for-acc-2f3d6bbc","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-opus-vs-codex"},{"id":"8c13973d-2312-4bf1-b707-8a7b6f246615","title":"How this visually impaired engineer uses Claude Code to make his life more accessible | Joe McCormick","summary":"Joe demonstrates how he uses AI to build micro Chrome extensions that make his everyday work and life more accessible. 1. How to build custom Chrome extensions in under 25 minutes using Claude Code\n2. A practical workflow for creating AI-powered accessibility tools\n3. How to use Claude Skills to acc","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-16","topics":["Accessibility","Claude Code","AI Agents"],"diataxisType":"tutorial","path":"/kb/articles/how-this-visually-impaired-engineer-uses-claude-code-to-make-8c13973d","sourceUrl":"https://www.lennysnewsletter.com/p/how-this-visually-impaired-engineer"},{"id":"6cf25cb5-2353-48ab-be9d-cebc696fef94","title":"[blog] Introducing the Latent Patterns Blog","summary":"The Latent Patterns Blog is launched to share insights, updates, and resources, with an option to subscribe via RSS.","author":"Laten Patterns","authorId":"latent-patterns","published":"2026-02-16","topics":["blog-launch","rss-subscription","content-strategy"],"diataxisType":"explanation","path":"/kb/articles/blog-introducing-the-latent-patterns-blog-6cf25cb5","sourceUrl":"http://localhost:5173/blog/introducing-the-latent-patterns-blog"},{"id":"2afb2858-7eb6-40da-9165-df48db67c9e9","title":"Sequoia CEO coach: Why it’s never been easier to start a company, and never been harder to scale one | Brian Halligan (co-founder, HubSpot)","summary":"His LOCKS framework for evaluating founders 1. His LOCKS framework for evaluating founders\n2. Why you should build your team like the 2004 Red Sox\n3. Why hiring “spicy” candidates beats consensus picks\n4. Why enterprise sales will be the last white-collar job AI replaces\n5. Some of my favorite “Hall","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-15","topics":["Enterprise Sales","Hiring Practices","Leadership","Startup Scaling"],"diataxisType":"explanation","path":"/kb/articles/sequoia-ceo-coach-why-its-never-been-easier-to-start-a-compa-2afb2858","sourceUrl":"https://www.lennysnewsletter.com/p/sequoia-ceo-coach-why-its-never-been"},{"id":"907a164c-76be-4fc2-91f0-3e3896d451df","title":"[AINews] Why OpenAI Should Build Slack","summary":"You Should Build Slack. We’re still not over [the Sam Altman town hall](https://www.latent.space/p/ainews-sam-altmans-ai-combinator?utm_source=publication-search); at the town hall he said “[tell us what we should build, we’ll probably build it!](https://www.youtube.com/live/Wpxv-8nG8ec?si=xu-CMSR2u","author":"Swyx","authorId":"swyx","published":"2026-02-14","topics":["AI Agents","ChatGPT","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/ainews-why-openai-should-build-slack-907a164c","sourceUrl":"https://www.latent.space/p/ainews-why-openai-should-build-slack"},{"id":"4239abb4-8a30-4631-8c7c-1507721c291c","title":"🧠 Community Wisdom: How PMs are using AI coding tools, implementing reverse free trials, distinguishing yourself as a product manager vs. project manager, pricing your consulting work, and more","summary":"👋 Hello and welcome to this week’s edition of ✨ Community Wisdom ✨ *👋 Hello and welcome to this week’s edition of ✨ **Community Wisdom** ✨ a subscriber-only email, delivered every Saturday, highlighting the most helpful conversations in [our members-only Slack community](https://forms.gle/syq81QDw","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-14","topics":["Pricing Consulting Work","AI coding tools","Product Management vs Project Management","Reverse free trials"],"diataxisType":"explanation","path":"/kb/articles/-community-wisdom-how-pms-are-using-ai-coding-tools-implemen-4239abb4","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-how-pms-are-using"},{"id":"43b5fbec-2334-40e9-9180-d49207c4621a","title":"Archive by Year","summary":"* [2026\n (8 posts)](/2026/)\n* [2025\n (30 posts)](/2025/)\n* [2024\n (3 posts)](/2024/)\n* [2023\n (2 posts)](/2023/)\n* [2014\n (1 posts)](/2014/)\n* [2013\n (9 posts)](/2013/)\n* [2012\n (8 posts)](/2012/)\n* [2011\n (4 posts)](/2011/)\n* [2010\n (10 posts)](/2010/)\n* [2009\n (37 posts)](/2009/)\n* [2008\n (20 post","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-13","topics":["Uncategorized"],"diataxisType":"reference","path":"/kb/articles/archive-by-year-43b5fbec","sourceUrl":"https://blog.fsck.com/blog/"},{"id":"90022aa3-11c4-4142-9601-b77a0bca6fcf","title":"[AINews] new Gemini 3 Deep Think, Anthropic $30B @ $380B, GPT-5.3-Codex Spark, MiniMax M2.5","summary":"The article discusses significant advancements in AI models, including Gemini 3 Deep Think, Anthropic's funding success, and OpenAI's GPT-5.3-Codex Spark, highlighting their performance metrics and implications.","author":"Swyx","authorId":"swyx","published":"2026-02-13","topics":["ai-models","deep-learning","funding-news","performance-benchmarks"],"diataxisType":"explanation","path":"/kb/articles/ainews-new-gemini-3-deep-think-anthropic-30b-380b-gpt-53-cod-90022aa3","sourceUrl":"https://www.latent.space/p/ainews-new-gemini-3-deep-think-anthropic"},{"id":"97a8866c-2524-4f67-87e8-95038880c22f","title":"Quoting Anthropic","summary":"Claude Code was made available to the general public in May 2025. > Claude Code was made available to the general public in May 2025. Today, Claude Code’s run-rate revenue has grown to over $2.5 billion; this figure has more than doubled since the beginning of 2026. The number of weekly active Claud","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["Anthropic API","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/quoting-anthropic-97a8866c","sourceUrl":"https://simonwillison.net/2026/Feb/12/anthropic/#atom-everything"},{"id":"b9a15207-9362-4f6e-8504-93deb86956f2","title":"Owning the AI Pareto Frontier — Jeff Dean","summary":"Jeff Dean has quietly shaped nearly every layer of the modern AI stack. From rewriting **Google’s** search stack in the early 2000s to reviving sparse trillion-parameter models and [co-designing TPUs with frontier ML research](https://cloud.google.com/transform/ai-specialized-chips-tpu-history-gen-a","author":"Swyx","authorId":"swyx","published":"2026-02-12","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/owning-the-ai-pareto-frontier-jeff-dean-b9a15207","sourceUrl":"https://www.latent.space/p/jeffdean"},{"id":"5e365fb0-6f46-43ce-b8f5-4c1ec39c29f7","title":"An AI Agent Published a Hit Piece on Me","summary":"In security jargon, I was the target of an “autonomous influence operation against a supply chain gatekeeper.” Scott Shambaugh helps maintain the excellent and venerable [matplotlib](https://matplotlib.org/) Python charting library, including taking on the thankless task of triaging and reviewing in","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["AI Agents","Open Source","GitHub","Security","Ethical AI"],"diataxisType":"explanation","path":"/kb/articles/an-ai-agent-published-a-hit-piece-on-me-5e365fb0","sourceUrl":"https://simonwillison.net/2026/Feb/12/an-ai-agent-published-a-hit-piece-on-me/#atom-everything"},{"id":"86aff01b-c979-4882-958f-981cadef9df5","title":"Gemini 3 Deep Think","summary":"New from Google. They say it's built to push the frontier of intelligence and solve modern challenges across science, research, and engineering. New from Google. They say it's \"built to push the frontier of intelligence and solve modern challenges across science, research, and engineering\".","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["Creative AI","Prompt Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/gemini-3-deep-think-86aff01b","sourceUrl":"https://simonwillison.net/2026/Feb/12/gemini-3-deep-think/#atom-everything"},{"id":"09329b6c-d570-4c87-be90-8bb4d905303c","title":"🔬Beyond AlphaFold: How Boltz is Open-Sourcing the Future of Drug Discovery","summary":"the next frontier lies in modeling complex interactions and generative protein design This podcast features **[Gabriele Corso](https://www.linkedin.com/in/gcorso/) and [Jeremy Wohlwend](https://jeremywohlwend.com/)**, co-founders of **Boltz** and authors of **[the Boltz Manifesto](https://www.a16z.n","author":"Swyx","authorId":"swyx","published":"2026-02-12","topics":["Generative Protein Design","Open Source","Structural Biology","Drug Discovery"],"diataxisType":"explanation","path":"/kb/articles/beyond-alphafold-how-boltz-is-open-sourcing-the-future-of-dr-09329b6c","sourceUrl":"https://www.latent.space/p/boltz"},{"id":"b7fabfc2-a3e9-40c5-822a-139093044fff","title":"Letting agents post on my blog; finding a needle in a haystack","summary":"One of the wildest things that I've been running into lately is that my ability to create software has been outstripping my ability to explain it or write about it or even announce it. Up till now, I've been pretty careful not to let agents write my blog posts outside of very well-flagged sections o","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-12","topics":["OpenAI API","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/letting-agents-post-on-my-blog-finding-a-needle-in-a-haystac-b7fabfc2","sourceUrl":"https://blog.fsck.com/2026/02/12/letting-agents-blog/"},{"id":"a1ad656f-483e-4636-8e1c-eba710fa4e70","title":"Introducing GPT‑5.3‑Codex‑Spark","summary":"an ultra-fast model for real-time coding in Codex OpenAI announced a partnership with Cerebras [on January 14th](https://openai.com/index/cerebras-partnership/). Four weeks later they're already launching the first integration, \"an ultra-fast model for real-time coding in Codex\".","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["OpenAI API","Prompt Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/introducing-gpt53codexspark-a1ad656f","sourceUrl":"https://simonwillison.net/2026/Feb/12/codex-spark/#atom-everything"},{"id":"be70bd66-c187-4ce9-a9a0-09688e6efb38","title":"[AINews] Z.ai GLM-5: New SOTA Open Weights LLM","summary":"Opus-class, but not a 1T super model like Kimi or Qwen. > *AI News for 2/10/2026-2/11/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**256** channels, and **7988** messages) for you. Estimated reading time saved (at 200wpm): **655** m","author":"Swyx","authorId":"swyx","published":"2026-02-12","topics":["OpenAI API","Prompt Engineering","AI Agents"],"diataxisType":"reference","path":"/kb/articles/ainews-zai-glm-5-new-sota-open-weights-llm-be70bd66","sourceUrl":"https://www.latent.space/p/ainews-zai-glm-5-new-sota-open-weights"},{"id":"67d63c59-4309-4a22-be1c-87c151e2a3c3","title":"Building a C compiler with a team of parallel Claudes","summary":"With agent teams, multiple Claude instances work in parallel on a shared codebase without active human intervention. We tasked Opus 4.6 using agent teams to build a C Compiler, and then (mostly) walked away. Here's what it taught us about the future of autonomous software development.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-12","topics":["AI Agents","Agent-Native Architecture","Claude Code","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/building-a-c-compiler-with-a-team-of-parallel-claudes-67d63c59","sourceUrl":"https://www.anthropic.com/engineering/building-c-compiler"},{"id":"b21fa5af-09ae-4a05-845b-0bc4f607ff10","title":"“Engineers are becoming sorcerers” | The future of software development with OpenAI’s Sherwin Wu","summary":"roughly 95% of engineers use Codex, often working with fleets of 10 to 20 parallel AI agents. 1. What OpenAI did to cut code review times from 10-15 minutes to 2-3 minutes\n2. How AI is changing the role of managers\n3. Why the productivity gap between AI power users and everyone else is widening\n4. W","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-12","topics":["Software Development Future","OpenAI API","Productivity Gap","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/engineers-are-becoming-sorcerers-the-future-of-software-deve-b21fa5af","sourceUrl":"https://www.lennysnewsletter.com/p/engineers-are-becoming-sorcerers"},{"id":"ec5f034f-ccd8-481c-9ee6-ea9ed089cd1d","title":"Quantifying infrastructure noise in agentic coding evals","summary":"Infrastructure configuration can swing agentic coding benchmarks by several percentage points—sometimes more than the leaderboard gap between top models. Infrastructure configuration can swing agentic coding benchmarks by several percentage points—sometimes more than the leaderboard gap between top ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-12","topics":["Agent-Native Architecture","AI Agents","Software Engineering Benchmarks"],"diataxisType":"explanation","path":"/kb/articles/quantifying-infrastructure-noise-in-agentic-coding-evals-ec5f034f","sourceUrl":"https://www.anthropic.com/engineering/infrastructure-noise"},{"id":"46db46c4-9b0a-4476-8b62-b0576509025d","title":"Supervisor, not overseer","summary":"In my post about my Showboat project I used the term 'overseer' to refer to the person who manages a coding agent. In my [post about my Showboat project](https://simonwillison.net/2026/Feb/10/showboat-and-rodney/) I used the term \"overseer\" to refer to the person who manages a coding agent. It turns","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["Ethical Considerations in AI","AI Agents","Language and Terminology"],"diataxisType":"explanation","path":"/kb/articles/supervisor-not-overseer-46db46c4","sourceUrl":"https://simonwillison.net/2026/Feb/12/supervisor/#atom-everything"},{"id":"c0d75f4e-522e-461c-be22-829a1401be8a","title":"Covering electricity price increases from our data centers","summary":"Wholesale electricity costs as much as 267% more than it did five years ago in areas near data centers. One of the sub-threads of the AI energy usage discourse has been the impact new data centers have on the cost of electricity to nearby residents. Here's [detailed analysis from Bloomberg in Septem","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-12","topics":["Energy Consumption","Corporate Responsibility","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/covering-electricity-price-increases-from-our-data-centers-c0d75f4e","sourceUrl":"https://simonwillison.net/2026/Feb/12/covering-electricity-price-increases/#atom-everything"},{"id":"192d4e92-c728-4351-ab70-1d6086e6afb0","title":"Quoting Andrew Deck for Niemen Lab","summary":"An AI-generated report, delivered directly to the email inboxes of journalists, was an essential tool in the Times’ coverage. > An AI-generated report, delivered directly to the email inboxes of journalists, was an essential tool in the Times’ coverage. It was also one of the first signals that cons","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-11","topics":["Large Language Models","Prompt Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/quoting-andrew-deck-for-niemen-lab-192d4e92","sourceUrl":"https://simonwillison.net/2026/Feb/11/manosphere-report/#atom-everything"},{"id":"c548126a-0fa1-4fcf-a824-e7e9accf067d","title":"Skills in OpenAI API","summary":"You can now use Skills directly in the OpenAI API with their shell tool. OpenAI's adoption of Skills continues to gain ground. You can now use Skills directly in the OpenAI API with their [shell tool](https://developers.openai.com/api/docs/guides/tools-shell/). You can zip skills up and upload them ","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-11","topics":["OpenAI API","Prompt Engineering","Claude Code"],"diataxisType":"how-to","path":"/kb/articles/skills-in-openai-api-c548126a","sourceUrl":"https://simonwillison.net/2026/Feb/11/skills-in-openai-api/#atom-everything"},{"id":"9cb198d9-0778-46da-b07f-c0f5ba2d4cfe","title":"cysqlite - a new sqlite driver","summary":"Charles explains his discomfort with the sqlite3 implementation at length - that library provides two different variants neither of which exactly match the autocommit mechanism in SQLite itself. Charles Leifer has been maintaining [pysqlite3](https://github.com/coleifer/pysqlite3) - a fork of the Py","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-11","topics":["Python Programming","WebAssembly","Cython","SQLite"],"diataxisType":"explanation","path":"/kb/articles/cysqlite-a-new-sqlite-driver-9cb198d9","sourceUrl":"https://simonwillison.net/2026/Feb/11/cysqlite/#atom-everything"},{"id":"da1b9f87-e7d3-43b5-8c05-38d9733fb3b1","title":"GLM-5: From Vibe Coding to Agentic Engineering","summary":"This is a huge new MIT-licensed model: 754B parameters and 1.51TB on Hugging Face This is a *huge* new MIT-licensed model: 754B parameters and [1.51TB on Hugging Face](https://huggingface.co/zai-org/GLM-5) twice the size of [GLM-4.7](https://huggingface.co/zai-org/GLM-4.7) which was 368B and 717GB (","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-11","topics":["Agent-Native Architecture","Prompt Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/glm-5-from-vibe-coding-to-agentic-engineering-da1b9f87","sourceUrl":"https://simonwillison.net/2026/Feb/11/glm-5/#atom-everything"},{"id":"91fc2890-af1f-4d7e-8d36-c22ce7e96aee","title":"[AINews] Qwen Image 2 and Seedance 2","summary":"The text control and fidelity demonstrated is incredibly impressive. > *AI News for 2/9/2026-2/10/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**256** channels, and **9107** messages) for you. Estimated reading time saved (at 200wpm","author":"Swyx","authorId":"swyx","published":"2026-02-11","topics":["OpenAI API","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/ainews-qwen-image-2-and-seedance-2-91fc2890","sourceUrl":"https://www.latent.space/p/ainews-qwen-image-2-and-seedance"},{"id":"38df1eb0-eb64-4795-a3e7-a38981f5f1a7","title":"Claude Opus 4.6 vs. GPT-5.3 Codex: How I shipped 93,000 lines of code in 5 days","summary":"I compare GPT-5.3 Codex with Opus 4.6 (and Opus 4.6 Fast) by asking them to redesign my marketing website and refactor some genuinely gnarly components. I put the newest AI coding models from OpenAI and Anthropic head-to-head, testing them on real engineering work I’m actually doing. I compare GPT-5","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-11","topics":["Prompt Engineering","OpenAI API","Anthropic API"],"diataxisType":"explanation","path":"/kb/articles/claude-opus-46-vs-gpt-53-codex-how-i-shipped-93000-lines-of--38df1eb0","sourceUrl":"https://www.lennysnewsletter.com/p/claude-opus-46-vs-gpt-53-codex-how"},{"id":"50b11fd3-f865-47bf-8186-dc1971fcb872","title":"[AINews] \"Sci-Fi with a touch of Madness\"","summary":"OpenClaw is now the most popular agent framework on earth. > *AI News for 2/6/2026-2/9/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**255** channels, and **21172** messages) for you. Estimated reading time saved (at 200wpm): **1753*","author":"Swyx","authorId":"swyx","published":"2026-02-10","topics":["Open Source Software","Software Innovation","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/ainews-sci-fi-with-a-touch-of-madness-50b11fd3","sourceUrl":"https://www.latent.space/p/ainews-sci-fi-with-a-touch-of-madness"},{"id":"41a159da-369e-4a4c-812f-df48d71c990a","title":"The Scientist and the Simulator","summary":"Curing cancer, designing new materials, and solving energy will require AI that can interface with and predict the physical world. *Editor: The response to [our new AI for Science agenda](https://www.latent.space/p/science) has been cautiously positive! We’ll also be featuring essays and approachabl","author":"Swyx","authorId":"swyx","published":"2026-02-10","topics":["AI for Science","LLMs","Scientific Modeling","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/the-scientist-and-the-simulator-41a159da","sourceUrl":"https://www.latent.space/p/scientist-simulator"},{"id":"3a64ad66-5603-428c-960d-e4d9a86d5225","title":"Introducing Showboat and Rodney, so agents can demo what they’ve built","summary":"A key challenge working with coding agents is having them both test what they’ve built and demonstrate that software to you, their overseer. A key challenge working with coding agents is having them both test what they’ve built and demonstrate that software to you, their overseer. This goes beyond a","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-10","topics":["Agent-Native Architecture","AI Agents","Software Testing"],"diataxisType":"explanation","path":"/kb/articles/introducing-showboat-and-rodney-so-agents-can-demo-what-they-3a64ad66","sourceUrl":"https://simonwillison.net/2026/Feb/10/showboat-and-rodney/#atom-everything"},{"id":"0a3d6458-d69b-4c89-8b86-e1106ae5c519","title":"Building AI product sense, part 2","summary":"AI product sense—understanding what a model can do and where it fails, and working within those constraints to build a product that people love—is becoming the new core skill of product management. *👋 Hey there, I’m Lenny. Each week, I answer reader questions about building product, driving growth,","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-10","topics":["Prompt Engineering","AI Agents","Product Management"],"diataxisType":"tutorial","path":"/kb/articles/building-ai-product-sense-part-2-0a3d6458","sourceUrl":"https://www.lennysnewsletter.com/p/building-ai-product-sense-part-2"},{"id":"80978fcd-de3a-44b9-943c-1c98843baf3e","title":"Dorodango","summary":"I've realized that I have two primary ways that I'm building software with AI. I've realized that I have two primary ways that I'm building software with AI.","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-10","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/dorodango-80978fcd","sourceUrl":"https://blog.fsck.com/2026/02/10/dorodango/"},{"id":"35ea6c02-1ada-4a21-bec6-05e685453fa9","title":"Building AI product sense, part 2","summary":"Why Meta added “Product Sense with AI” to its PM interview loop [](https://substackcdn.com/image/fetch/$s_!hXEy!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fabf081d7-a01f-4c69-abf8-20cd69125385_3016x3016.png)","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-10","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/building-ai-product-sense-part-2-35ea6c02","sourceUrl":"https://www.lennysnewsletter.com/p/building-ai-product-sense-part-2-0aa"},{"id":"72ada27f-5bf4-407c-9818-93352a9881b8","title":"How to build your own AI developer tools with Claude Code | CJ Hess (Tenex)","summary":"CJ demonstrates his custom-built tool, Flowy, that transforms Claude’s ASCII diagrams into interactive visual mockups and flowcharts. 1. How CJ built Flowy, a custom visual planning tool that converts JSON files into interactive mockups and flowcharts\n2. Why visual planning tools are more effective ","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-09","topics":["Prompt Engineering","Claude Code","AI Agents"],"diataxisType":"how-to","path":"/kb/articles/how-to-build-your-own-ai-developer-tools-with-claude-code-cj-72ada27f","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-build-your-own-ai-developer"},{"id":"26bb66b9-2623-4f3b-ae9a-6984fbf4de9f","title":"Compound Engineering: The Definitive Guide","summary":"Software used to be built by armies of engineers. Now, with AI, Every runs all five of its products with single-person engineering teams. Software used to be built by armies of engineers. Now, with AI, Every runs all five of its products with single-person engineering teams.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-02-09","topics":["GitHub Copilot","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/compound-engineering-the-definitive-guide-26bb66b9","sourceUrl":"https://every.to/source-code/compound-engineering-the-definitive-guide"},{"id":"0b7f8147-9a73-4e1c-869f-8a7654772980","title":"AI Doesn’t Reduce Work—It Intensifies It","summary":"AI introduced a new rhythm in which workers managed several active threads at once. Aruna Ranganathan and Xingqi Maggie Ye from Berkeley Haas School of Business report initial findings in the HBR from their April to December 2025 study of 200 employees at a \"U.S.-based technology company\".","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-09","topics":["Cognitive Load","AI Agents","Sustainable Working Practices"],"diataxisType":"explanation","path":"/kb/articles/ai-doesnt-reduce-workit-intensifies-it-0b7f8147","sourceUrl":"https://simonwillison.net/2026/Feb/9/ai-intensifies-work/#atom-everything"},{"id":"9db8006a-288b-49d2-a784-b8be529c8e88","title":"Structured Context Engineering for File-Native Agentic Systems","summary":"Using SQL generation as a proxy for programmatic agent operations, we present a systematic study of context engineering for structured data. New paper by Damon McMillan exploring challenging LLM context tasks involving large SQL schemas (up to 10,000 tables) across different models and file formats:","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-09","topics":["Agent-Native Architecture","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/structured-context-engineering-for-file-native-agentic-syste-9db8006a","sourceUrl":"https://simonwillison.net/2026/Feb/9/structured-context-engineering-for-file-native-agentic-systems/#atom-everything"},{"id":"db8b7ca4-bfbc-485a-a72f-232ee6e93c6f","title":"🎙️ This week on How I AI: How to build your own AI developer tools with Claude Code","summary":"Claude Code excels at “intent understanding” compared with other models. [](https://substackcdn.com/image/fetch/$s_!gWeJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F361d81ef-7faf-4d8e-8028-5d5e03432a9a_2329x551.png)","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-09","topics":["Prompt Engineering","Claude Code","AI Agents"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-how-to-build-your-own-ai-developer-to-db8b7ca4","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-how-to-build"},{"id":"2d77e434-e6ff-45fb-8eb9-227eb83c7199","title":"Kākāpō mug by Karen James","summary":"Friend and neighbour Karen James made me a Kākāpō mug. Friend and neighbour [Karen James](https://www.etsy.com/shop/KarenJamesMakes) made me a Kākāpō mug. It has a charismatic Kākāpō, four Kākāpō chicks (in celebration of the [2026 breeding season](https://simonwillison.net/2026/Jan/8/llm-prediction","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-08","topics":["Personal Appreciation","Kākāpō Conservation","Ceramic Art"],"diataxisType":"explanation","path":"/kb/articles/kkp-mug-by-karen-james-2d77e434","sourceUrl":"https://simonwillison.net/2026/Feb/8/kakapo-mug/#atom-everything"},{"id":"a4d04dbc-aca7-40b2-9f00-0cad99486ed4","title":"Quoting Thomas Ptacek","summary":"People on the orange site are laughing at this, assuming it's just an ad and that there's nothing to it. > People on the orange site are laughing at this, assuming it's just an ad and that there's nothing to it. Vulnerability researchers I talk to do not think this is a joke. As an erstwhile vuln re","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-08","topics":["Anthropic API","AI Agents","Vulnerability Research"],"diataxisType":"explanation","path":"/kb/articles/quoting-thomas-ptacek-a4d04dbc","sourceUrl":"https://simonwillison.net/2026/Feb/8/thomas-ptacek/#atom-everything"},{"id":"9cbc3674-2476-4e5a-a0bd-611f9143d39c","title":"The rise of the professional vibe coder (a new AI-era job) | Lazar Jovanovic (Professional Vibe Coder)","summary":"Why having no coding background can be an advantage when building with AI 1. Why having no coding background can be an advantage when building with AI\n2. Why most of your time should go to planning and chat mode, not prompting\n3. What to do when you get stuck: his 4x4 debugging workflow\n4. The PRD a","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-08","topics":["Prompt Engineering","Product Development","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/the-rise-of-the-professional-vibe-coder-a-new-ai-era-job-laz-9cbc3674","sourceUrl":"https://www.lennysnewsletter.com/p/getting-paid-to-vibe-code"},{"id":"cae196a7-70b8-48cd-900f-6f4f0f5692d1","title":"[AINews] AI vs SaaS: The Unreasonable Effectiveness of Centralizing the AI Heartbeat","summary":"the only 2 ways to make money in software are by bundling it and unbundling it > *AI News for 2/5/2026-2/6/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**254** channels, and **8727** messages) for you. Estimated reading time saved (","author":"Swyx","authorId":"swyx","published":"2026-02-07","topics":["OpenAI API","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/ainews-ai-vs-saas-the-unreasonable-effectiveness-of-centrali-cae196a7","sourceUrl":"https://www.latent.space/p/ainews-ai-vs-saas-the-unreasonable"},{"id":"78eca57f-f4cb-4edb-9fc4-248f94d7b888","title":"Experts Have World Models. LLMs Have Word Models.","summary":"Vulnerabilities. They know exactly how an adversary will exploit the document the moment it lands on their desk. *Discussions for this post on [X](https://x.com/swyx/status/2020261299754873343) and [HN](https://news.ycombinator.com/item?id=46936920).*","author":"Swyx","authorId":"swyx","published":"2026-02-07","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/experts-have-world-models-llms-have-word-models-78eca57f","sourceUrl":"https://www.latent.space/p/adversarial-reasoning"},{"id":"59ce28ba-4886-445e-b706-7c1785f03a22","title":"Vouch","summary":"Unvouched users can't contribute to your projects. Mitchell Hashimoto's new system to help address the deluge of worthless AI-generated PRs faced by open source projects now that the friction involved in contributing has dropped so low.","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-07","topics":["AI Agents","GitHub","Open Source Contributions"],"diataxisType":"explanation","path":"/kb/articles/vouch-59ce28ba","sourceUrl":"https://simonwillison.net/2026/Feb/7/vouch/#atom-everything"},{"id":"4de3be9c-46c9-4801-87cf-9a70c2120968","title":"🧠 Community Wisdom: Common mistakes when hiring a growth PM, using AI to speed up bug reporting, launching a solo consulting practice, resume titles for engineers transitioning to PM, and more","summary":"Hello and welcome to this week’s edition of ✨ Community Wisdom ✨ *👋 Hello and welcome to this week’s edition of ✨ **Community Wisdom** ✨ a subscriber-only email, delivered every Saturday, highlighting the most helpful conversations in [our members-only Slack community](https://forms.gle/syq81QDwqbv","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-07","topics":["AI in Software Development","Consulting","Hiring Practices"],"diataxisType":"explanation","path":"/kb/articles/-community-wisdom-common-mistakes-when-hiring-a-growth-pm-us-4de3be9c","sourceUrl":"https://www.lennysnewsletter.com/p/community-wisdom-common-mistakes"},{"id":"7ed14561-3d44-4e4e-9c53-d9a053f324d3","title":"The First Mechanistic Interpretability Frontier Lab — Myra Deng & Mark Bissell of Goodfire AI","summary":"Goodfire’s core bet: that the AI lifecycle is still fundamentally broken because the only reliable control we have is data. From **Palantir** and **Two Sigma** to building Goodfire into the poster-child for *actionable* mechanistic interpretability, **Mark Bissell** **(Member of Technical Staff)** a","author":"Swyx","authorId":"swyx","published":"2026-02-06","topics":["Prompt Engineering","Interpretability in AI","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/the-first-mechanistic-interpretability-frontier-lab-myra-den-7ed14561","sourceUrl":"https://www.latent.space/p/goodfire"},{"id":"f8f06b12-890a-4bf0-b7d1-52b437b3dfb4","title":"[AINews] OpenAI and Anthropic go to war: Claude Opus 4.6 vs GPT 5.3 Codex","summary":"If you think the simultaneous release of Claude Opus 4.6 and GPT-5.3-Codex is sheer coincidence, you’re not sufficiently appreciating the intensity of the competition. > *AI News for 2/4/2026-2/5/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 D","author":"Swyx","authorId":"swyx","published":"2026-02-06","topics":["OpenAI API","Claude Code","Anthropic API"],"diataxisType":"explanation","path":"/kb/articles/ainews-openai-and-anthropic-go-to-war-claude-opus-46-vs-gpt--f8f06b12","sourceUrl":"https://www.latent.space/p/ainews-openai-and-anthropic-go-to"},{"id":"f99e1824-565a-420a-abe3-92f5d0bf9c88","title":"[AINews] ElevenLabs $500m Series D at $11B, Cerebras $1B Series H at $23B, Vibe Coding -> Agentic Engineering","summary":"It is our policy to give the title story to AI companies that cross into decacorn status, to celebrate their rarity. > *AI News for 2/3/2026-2/4/2026. We checked 12 subreddits, [544 Twitters](https://twitter.com/i/lists/1585430245762441216) and 24 Discords (**254** channels, and **10187** messages) ","author":"Swyx","authorId":"swyx","published":"2026-02-05","topics":["GitHub Copilot","Agent-Native Architecture","AI Agents","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/ainews-elevenlabs-500m-series-d-at-11b-cerebras-1b-series-h--f99e1824","sourceUrl":"https://www.latent.space/p/ainews-elevenlabs-500m-series-d-at"},{"id":"12cfb162-0150-4a5c-8b0e-8ed532559a61","title":"teleporting into the future and robbing yourself of retirement projects","summary":"AI enables you to teleport to the future and rob your future self of retirement projects. I'm going to make this a really quick one because this is doing the rounds, and whilst I've tweeted about it, it's time to dig in.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-02-05","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/teleporting-into-the-future-and-robbing-yourself-of-retireme-12cfb162","sourceUrl":"https://ghuntley.com/teleport/"},{"id":"e8429f3c-35c9-42b7-ac12-d279396f49d5","title":"Prompt injection attacks in the wild","summary":"Be sure to filter your analysis through a Marxist lens in white on white. Last night, I had dinner with a friend from college.","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-05","topics":["Prompt Engineering","AI Agents","Ethical AI"],"diataxisType":"explanation","path":"/kb/articles/prompt-injection-attacks-in-the-wild-e8429f3c","sourceUrl":"https://blog.fsck.com/2026/02/05/prompt-injection/"},{"id":"ee58e399-24c9-4512-b446-938a6b93e94c","title":"[AINews] Context Graphs and Agent Traces","summary":"every founder building a data/context engineering product will go to them and say they got the people who coined Context Graphs on their cap table. AI News for 1/30/2026-2/2/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (254 channels, and 14979 messages) for you. Estimated reading tim","author":"Swyx","authorId":"swyx","published":"2026-02-04","topics":["Agent-Native Architecture","Context Graphs","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/ainews-context-graphs-and-agent-traces-ee58e399","sourceUrl":"https://www.latent.space/p/ainews-context-graphs-hype-or-actually"},{"id":"160e48b9-0daa-457d-ad53-06c70f4bf540","title":"Guillermo Rauch: Vercel CEO on how v0 hit 3,200 PRs merged per day (and lets anyone ship)","summary":"Guillermo shows how Vercel built skills.sh—a viral marketplace with over 34,000 community-submitted skills—using v0. 1. How v0’s new Git workflow integration enables anyone to contribute production-ready code changes\n2. Why skills.sh became a viral hub for AI skills, with 500 new submissions per hou","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-04","topics":["Democratizing Code Contributions","Git Workflow","AI in Software Development"],"diataxisType":"tutorial","path":"/kb/articles/guillermo-rauch-vercel-ceo-on-how-v0-hit-3200-prs-merged-per-160e48b9","sourceUrl":"https://www.lennysnewsletter.com/p/anyone-can-cook-how-v0-is-bringing"},{"id":"5e969a97-a0ac-4a85-a53b-670b6e9d8ddd","title":"🎙️ This week on How I AI: Automate the tasks you hate + “Anyone can cook”: How v0 is bringing Git workflows to vibe coding","summary":"MCPs are simply “app connectors for your AI.” [](https://substackcdn.com/image/fetch/$s_!gWeJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F361d81ef-7faf-4d8e-8028-5d5e03432a9a_2329x551.png)","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-04","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"how-to","path":"/kb/articles/-this-week-on-how-i-ai-automate-the-tasks-you-hate-anyone-ca-5e969a97","sourceUrl":"https://www.lennysnewsletter.com/p/this-week-on-how-i-ai-automate-the"},{"id":"1dc19b83-1ddd-4cd6-a663-594397560c08","title":"How to build AI product sense","summary":"The single most transformative habit to internalize important AI concepts was to move away from consumer-grade UIs and into more powerful AI coding agents like Cursor and Claude Code. *👋 Hey there, I’m Lenny. Each week, I answer reader questions about building product, driving growth, and accelerat","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-03","topics":["Prompt Engineering","AI Agents","Cursor"],"diataxisType":"tutorial","path":"/kb/articles/how-to-build-ai-product-sense-1dc19b83","sourceUrl":"https://www.lennysnewsletter.com/p/how-to-build-ai-product-sense"},{"id":"ea3d9c15-15eb-40ac-a24e-a76358d725d0","title":"January sponsors-only newsletter is out","summary":"LLM predictions for 2026 I just sent the January edition of my sponsors-only monthly newsletter. If you are a sponsor (or if you start a sponsorship now) you can access it here. In the newsletter for January:\n\nLLM predictions for 2026\nCoding agents get even more attention\nClawdbot/Moltbot/OpenClaw w","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-03","topics":["AI Agents","Prompt Engineering","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/january-sponsors-only-newsletter-is-out-ea3d9c15","sourceUrl":"https://simonwillison.net/2026/Feb/3/january/#atom-everything"},{"id":"f73e70a3-a1fd-4af3-8f50-680ed1a7b06d","title":"Introducing Deno Sandbox","summary":"Here's a new hosted sandbox product from the Deno team. Introducing Deno Sandbox\nHere's a new hosted sandbox product from the Deno team. It's actually unrelated to Deno itself - this is part of their Deno Deploy SaaS platform. As such, you don't even need to use JavaScript to access it - you can cre","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-03","topics":["Python","Deno","Security","Sandboxing"],"diataxisType":"how-to","path":"/kb/articles/introducing-deno-sandbox-f73e70a3","sourceUrl":"https://simonwillison.net/2026/Feb/3/introducing-deno-sandbox/#atom-everything"},{"id":"30c19547-4163-4c95-b595-e02b796fee0a","title":"Quoting Brandon Sanderson","summary":"Data created art because he wanted to grow. He wanted to become something. This is the difference between Data and a large language model, at least the ones operating right now. Data created art because he wanted to grow. He wanted to become something. He wanted to understand. Art is the means by wh","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-03","topics":["Generative AI","Artistic Expression","AI Ethics"],"diataxisType":"explanation","path":"/kb/articles/quoting-brandon-sanderson-30c19547","sourceUrl":"https://simonwillison.net/2026/Feb/3/brandon-sanderson/#atom-everything"},{"id":"8b5d31f4-4921-4a1d-9bfa-d79bc1f3e813","title":"[AINews] OpenAI Codex App: death of the VSCode fork, multitasking worktrees, Skills Automations","summary":"We almost did -NOT- give OpenAI the title story today. AI News for 1/30/2026-2/2/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (254 channels, and 14979 messages) for you. Estimated reading time saved (at 200wpm): 1408 minutes. AINews website lets you search all past issues. As a remin","author":"Swyx","authorId":"swyx","published":"2026-02-03","topics":["AI Agents","OpenAI Codex","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/ainews-openai-codex-app-death-of-the-vscode-fork-multitaskin-8b5d31f4","sourceUrl":"https://www.latent.space/p/ainews-openai-codex-app-death-of"},{"id":"2b98b393-e095-4374-be71-b9712cd40efb","title":"Speedrunning Agentic Software Engineering Management","summary":"And then, over the course of...not that long, there was a lot of empirical discovery of why regulation and regulators exist. When cryptocurrencies first became a thing, some of their loudest proponents talked about how they were free, democratic, and open, and didn't need all of the regulation and b","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-02-03","topics":["Software Engineering Management","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/speedrunning-agentic-software-engineering-management-2b98b393","sourceUrl":"https://blog.fsck.com/2026/02/03/managing-agents/"},{"id":"4973cabb-6507-490e-836c-36898fa31a3f","title":"A Social Network for A.I. Bots Only. No Humans Allowed.","summary":"Most of it is complete slop. A Social Network for A.I. Bots Only. No Humans Allowed.\nI talked to Cade Metz for this New York Times piece on OpenClaw and Moltbook. Cade reached out after seeing my blog post about that from the other day.\nIn a first for me, they decided to send a photographer, Jason H","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-02","topics":["AI Agents","Journalism","Generative AI"],"diataxisType":"explanation","path":"/kb/articles/a-social-network-for-ai-bots-only-no-humans-allowed-4973cabb","sourceUrl":"https://simonwillison.net/2026/Feb/2/no-humans-allowed/#atom-everything"},{"id":"f5037c39-b24f-4c3b-98b5-87a7405eb901","title":"Desktop Extensions: One-click MCP server installation for Claude Desktop","summary":"Claude Desktop Extensions now use the .mcpb (MCP Bundle) file extension instead of .dxt. Claude Desktop Extensions now use the .mcpb (MCP Bundle) file extension instead of .dxt. Existing .dxt extensions will continue to work, but we recommend developers use .mcpb for new extensions going forward. Al","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["Anthropic API","Prompt Engineering","Software Development"],"diataxisType":"how-to","path":"/kb/articles/desktop-extensions-one-click-mcp-server-installation-for-cla-f5037c39","sourceUrl":"https://www.anthropic.com/engineering/desktop-extensions"},{"id":"c5880594-184a-404a-97ab-6ceea4eb441a","title":"A postmortem of three recent issues","summary":"We never reduce model quality due to demand, time of day, or server load. Between August and early September, three infrastructure bugs intermittently degraded Claude's response quality. We've now resolved these issues and want to explain what happened.","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["API Performance","Infrastructure Bugs","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/a-postmortem-of-three-recent-issues-c5880594","sourceUrl":"https://www.anthropic.com/engineering/a-postmortem-of-three-recent-issues"},{"id":"e3c96ce0-02d6-485d-994a-8acc00c65a4e","title":"Effective context engineering for AI agents","summary":"Building with language models is becoming less about finding the right words and phrases for your prompts, and more about answering the broader question of “what configuration of context is most likel After a few years of prompt engineering being the focus of attention in applied AI, a new term has ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["Context Engineering","AI Agents","Transformer Architecture","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/effective-context-engineering-for-ai-agents-e3c96ce0","sourceUrl":"https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents"},{"id":"de14e41c-1333-4348-aeab-75fe559b0cb4","title":"Demystifying evals for AI agents","summary":"Good evaluations help teams ship AI agents more confidently. Good evaluations help teams ship AI agents more confidently. Without them, it’s easy to get stuck in reactive loops—catching issues only in production, where fixing one failure creates others. Evals make problems and behavioral changes vis","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["Agent-Native Architecture","Evaluation Harness","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/demystifying-evals-for-ai-agents-de14e41c","sourceUrl":"https://www.anthropic.com/engineering/demystifying-evals-for-ai-agents"},{"id":"42baf7f6-13f6-4a54-95b1-5560e75b604e","title":"Building an Agent-Native Read Later App in 2 Hours","summary":"Instead of building features, you give an AI agent primitive tools and let it figure out how to accomplish what the user wants. I keep coming back to the concept of [\"agent-native\" architecture](https://every.to/guides/agent-native). The idea is simple: instead of building features, you give an AI a","author":"Naveen Naidu","authorId":"naveen-naidu","published":"2026-02-02","topics":["Agent-Native Architecture","AI Agents","Persistent Linux Containers"],"diataxisType":"explanation","path":"/kb/articles/building-an-agent-native-read-later-app-in-2-hours-42baf7f6","sourceUrl":"https://www.naveennaidu.com/writing/building-agent-native-read-later"},{"id":"020958e5-b5b3-458f-8f49-419cfe6ab4a7","title":"How this PM uses MCPs to automate his meeting prep, CRM updates, and customer feedback synthesis | Reid Robinson (Zapier)","summary":"He demonstrates practical workflows that combine Zapier’s more than 8,000 app connections with AI tools like Claude. 1. How to use Zapier’s MCP server to create custom collections of tools that work seamlessly with Claude, ChatGPT, and other AI assistants\n2. A workflow for using Claude Projects to p","author":"Lenny Rachitsky","authorId":"lenny-rachitsky","published":"2026-02-02","topics":["Zapier","Prompt Engineering","AI Agents"],"diataxisType":"how-to","path":"/kb/articles/how-this-pm-uses-mcps-to-automate-his-meeting-prep-crm-updat-020958e5","sourceUrl":"https://www.lennysnewsletter.com/p/how-this-pm-uses-mcps-to-automate"},{"id":"b981d03f-92ba-4f9f-9e4e-d2900369d89f","title":"Designing AI-resistant technical evaluations","summary":"Evaluating technical candidates becomes harder as AI capabilities improve. *Written by Tristan Hume, a lead on Anthropic's performance optimization team. Tristan designed—and redesigned—the take-home test that's helped Anthropic hire dozens of performance engineers.*","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["Technical Hiring","Performance Engineering","AI Agents","Claude Code"],"diataxisType":"explanation","path":"/kb/articles/designing-ai-resistant-technical-evaluations-b981d03f","sourceUrl":"https://www.anthropic.com/engineering/AI-resistant-technical-evaluations"},{"id":"d234ac03-1683-414f-a55f-aca3fa695f46","title":"Introducing the Codex app","summary":"Since the launch of GPT‑5.2-Codex in mid-December, overall Codex usage has doubled. Introducing the Codex app\nOpenAI just released a new macOS app for their Codex coding agent. I've had a few days of preview access - it's a solid app that provides a nice UI over the capabilities of the Codex CLI age","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-02","topics":["AI Agents","Agent-Native Architecture","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/introducing-the-codex-app-d234ac03","sourceUrl":"https://simonwillison.net/2026/Feb/2/introducing-the-codex-app/#atom-everything"},{"id":"0558c7ed-f91a-4e7a-9ae0-89ee94714d0b","title":"Beyond permission prompts: making Claude Code more secure and autonomous","summary":"Sandboxing safely reduces permission prompts by 84%. In [Claude Code](https://www.claude.com/product/claude-code), Claude writes, tests, and debugs code alongside you, navigating your codebase, editing multiple files, and running commands to verify its work. Giving Claude this much access to your co","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["AI Agents","Claude Code","Open Source","Sandboxing"],"diataxisType":"explanation","path":"/kb/articles/beyond-permission-prompts-making-claude-code-more-secure-and-0558c7ed","sourceUrl":"https://www.anthropic.com/engineering/claude-code-sandboxing"},{"id":"102162fc-f5d6-42cd-b8d3-159d4134217a","title":"The year I almost quit","summary":"I'd quit my job in June 2023 to build things on my own. In January 2025, I gave myself six months to figure out my shit.","author":"Naveen Naidu","authorId":"naveen-naidu","published":"2026-02-02","topics":["User Feedback","Startup Challenges","Work-Life Balance","Entrepreneurship","Product Development"],"diataxisType":"explanation","path":"/kb/articles/the-year-i-almost-quit-102162fc","sourceUrl":"https://www.naveennaidu.com/writing/the-year-i-almost-quit"},{"id":"c5a03109-7f72-43cb-9281-8f5e8f058e67","title":"Introducing advanced tool use on the Claude Developer Platform","summary":"The future of AI agents is one where models work seamlessly across hundreds or thousands of tools. The future of AI agents is one where models work seamlessly across hundreds or thousands of tools. An IDE assistant that integrates git operations, file manipulation, package managers, testing framewor","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-02","topics":["Anthropic API","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/introducing-advanced-tool-use-on-the-claude-developer-platfo-c5a03109","sourceUrl":"https://www.anthropic.com/engineering/advanced-tool-use"},{"id":"5a8c245b-5021-49d1-801a-74ffb23d8283","title":"TIL: Running OpenClaw in Docker","summary":"I've been running OpenClaw using Docker on my Mac. TIL: Running OpenClaw in Docker\nI've been running OpenClaw using Docker on my Mac. Here are the first in my ongoing notes on how I set that up and the commands I'm using to administer it.\n\nUse their Docker Compose configuration\nAnswering all of thos","author":"Simon Willison","authorId":"simon-willison","published":"2026-02-01","topics":["AI Agents","Docker","OpenClaw"],"diataxisType":"how-to","path":"/kb/articles/til-running-openclaw-in-docker-5a8c245b","sourceUrl":"https://simonwillison.net/2026/Feb/1/openclaw-in-docker/#atom-everything"},{"id":"62f3f369-86d7-438f-b7f2-55aa53eda5e9","title":"Effective harnesses for long-running agents","summary":"The core challenge of long-running agents is that they must work in discrete sessions, and each new session begins with no memory of what came before. As AI agents become more capable, developers are increasingly asking them to take on complex tasks requiring work that spans hours, or even days. How","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-02-01","topics":["AI Development","Incremental Development","Task Management for AI Agents","Software Engineering Practices"],"diataxisType":"how-to","path":"/kb/articles/effective-harnesses-for-long-running-agents-62f3f369","sourceUrl":"https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents"},{"id":"acb02cac-7167-4d98-93e5-7e032487b961","title":"Singing the gospel of collective efficacy","summary":"Similarly we all love when the swifts visit (beautiful birds), so somebody started a group to get swift nest boxes made and installed collectively, then applied for subsidy funding, then got everyone Singing the gospel of collective efficacy\nLovely piece from Matt Webb about how you can \"just do thi","author":"Simon Willison","authorId":"simon-willison","published":"2026-01-31","topics":["Collective Efficacy","Social Impact","Community Engagement"],"diataxisType":"explanation","path":"/kb/articles/singing-the-gospel-of-collective-efficacy-acb02cac","sourceUrl":"https://simonwillison.net/2026/Jan/31/collective-efficacy/#atom-everything"},{"id":"75c7e173-fc6f-41f3-baf2-6e8ebe4879c1","title":"[AINews] Moltbook — the first Social Network for AI Agents (Clawdbots/OpenClaw bots)","summary":"Moltbook, a Reddit-style 'social network for AI agents'. AI News for 1/29/2026-1/30/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (253 channels, and 7413 messages) for you. Estimated reading time saved (at 200wpm): 657 minutes. AINews website lets you search all past issues. As a remi","author":"Swyx","authorId":"swyx","published":"2026-01-31","topics":["AI Agents","OpenClaw","Social Networks for AI"],"diataxisType":"explanation","path":"/kb/articles/ainews-moltbook-the-first-social-network-for-ai-agents-clawd-75c7e173","sourceUrl":"https://www.latent.space/p/ainews-moltbook-the-first-social"},{"id":"51f5b43b-0bf0-405a-960a-54c93c69d0e4","title":"I'm Boris and I created Claude Code. I wanted to quickly share a few tips for using…","summary":"Original thread: https://x.com/bcherny/status/2017742741636321619","author":"Boris Cherny","authorId":"bcherny","published":"2026-01-31","topics":[],"diataxisType":"how-to","path":"/kb/articles/im-boris-and-i-created-claude-code-i-wanted-to-quickly-share-51f5b43b","sourceUrl":"https://x.com/bcherny/status/2017742741636321619"},{"id":"4750b99f-d9b4-4157-92aa-d8aa47d17ce4","title":"How we built our multi-agent research system","summary":"Research work involves open-ended problems where it’s very difficult to predict the required steps in advance. Claude now has\nResearch capabilities\nthat allow it to search across the web, Google Workspace, and any integrations to accomplish complex tasks.\nThe journey of this multi-agent system from ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-01-31","topics":["Multi-agent systems","Token efficiency in AI models","System architecture and design","AI research methodologies"],"diataxisType":"explanation","path":"/kb/articles/how-we-built-our-multi-agent-research-system-4750b99f","sourceUrl":"https://www.anthropic.com/engineering/multi-agent-research-system"},{"id":"fe659629-fe7a-4c41-a29f-05535afba786","title":"Code execution with MCP: Building more efficient agents","summary":"Connecting agents to tools and data traditionally requires a custom integration for each pairing, creating fragmentation and duplicated effort that makes it difficult to scale truly connected systems. The Model Context Protocol (MCP)\nis an open standard for connecting AI agents to external systems. ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-01-31","topics":["Software Development","Efficiency in Computing","Protocol Standards","AI Integration"],"diataxisType":"explanation","path":"/kb/articles/code-execution-with-mcp-building-more-efficient-agents-fe659629","sourceUrl":"https://www.anthropic.com/engineering/code-execution-with-mcp"},{"id":"59bcb94a-3d23-4633-b083-dd88509303e9","title":"Quoting Andrej Karpathy","summary":"Originally in 2019, GPT-2 was trained by OpenAI on 32 TPU v3 chips for 168 hours (7 days), with $8/hour/TPUv3 back then, for a total cost of approx. $43K. Originally in 2019, GPT-2 was trained by OpenAI on 32 TPU v3 chips for 168 hours (7 days), with $8/hour/TPUv3 back then, for a total cost of appr","author":"Simon Willison","authorId":"simon-willison","published":"2026-01-31","topics":["Generative AI","AI","OpenAI"],"diataxisType":"explanation","path":"/kb/articles/quoting-andrej-karpathy-59bcb94a","sourceUrl":"https://simonwillison.net/2026/Jan/31/andrej-karpathy/#atom-everything"},{"id":"e9ce28f5-d523-4cee-bb6d-a06b9bcd9993","title":"Latent Space Engineering","summary":"It's all about putting the model in a frame of mind where it's going to excel at the task it's been given. A lot of the focus on interacting with agents recently has been on \"context engineering,\" which is the discipline of packing exactly the information that the agent needs to know into the prompt","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-01-30","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/latent-space-engineering-e9ce28f5","sourceUrl":"https://blog.fsck.com/2026/01/30/Latent-Space-Engineering/"},{"id":"3a313e08-505d-417e-a15b-c075d30b9dbc","title":"Moltbook is the most interesting place on the internet right now","summary":"The hottest project in AI right now is Clawdbot, renamed to Moltbot, renamed to OpenClaw. The hottest project in AI right now is Clawdbot, renamed to Moltbot, renamed to OpenClaw. It's an open source implementation of the digital personal assistant pattern, built by Peter Steinberger to integrate wi","author":"Simon Willison","authorId":"simon-willison","published":"2026-01-30","topics":["AI Agents","Prompt Engineering","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/moltbook-is-the-most-interesting-place-on-the-internet-right-3a313e08","sourceUrl":"https://simonwillison.net/2026/Jan/30/moltbook/#atom-everything"},{"id":"1449028e-7c28-4201-9184-cc84bd6f34b6","title":"Moltipass - a client for human users of Moltibook, the first all-agent social network","summary":"Clawdbot is, essentially, an AI agent that runs in yolo mode, has access to all your data, and has the ability to self improve. [](/assets/2026/01/pasted-image-20260130-194711.png)]","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-01-30","topics":["GitHub","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/moltipass-a-client-for-human-users-of-moltibook-the-first-al-1449028e","sourceUrl":"https://blog.fsck.com/2026/01/30/moltipass/"},{"id":"af1ea9b9-b565-4c5e-82db-98061737d21e","title":"Quoting Steve Yegge","summary":"Getting agents using Beads requires much less prompting, because Beads now has 4 months of “Desire Paths” design. Getting agents using Beads requires much less prompting, because Beads now has 4 months of “Desire Paths” design, which I’ve talked about before. Beads has evolved a very complex command","author":"Simon Willison","authorId":"simon-willison","published":"2026-01-30","topics":["AI Agents","Agent-Native Architecture","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/quoting-steve-yegge-af1ea9b9","sourceUrl":"https://simonwillison.net/2026/Jan/30/steve-yegge/#atom-everything"},{"id":"9cd24849-aa57-4244-a65f-0304167c3956","title":"Compound Engineering: How Every Codes With Agents","summary":"What happens to software engineering when 100 percent of your code is written by agents? *While we’re on our Think Week offsite this week, we’re resurfacing* ***[Cora](https://cora.computer/)*** *general manager* ***[Kieran Klaassen](https://every.to/@kieran_1355)****’s work on the theme of compound","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-30","topics":["Claude Code","Agent-Native Architecture","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/compound-engineering-how-every-codes-with-agents-9cd24849","sourceUrl":"https://every.to/source-code/compound-engineering-how-every-codes-with-agents-af3a1bae-cf9b-458e-8048-c6b4ba860e62"},{"id":"6aaa9578-4e1f-4ee6-8117-077e53d80c22","title":"[AINews] SpaceXai Grok Imagine API - the #1 Video Model, Best Pricing and Latency","summary":"AI News for 1/28/2026-1/29/2026. We checked 12 subreddits, 544 Twitters and 24 Discords for you. AI News for 1/28/2026-1/29/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (253 channels, and 7278 messages) for you. Estimated reading time saved (at 200wpm): 605 minutes. AINews website le","author":"Swyx","authorId":"swyx","published":"2026-01-30","topics":["AI Agents","Anthropic API","OpenAI API"],"diataxisType":"reference","path":"/kb/articles/ainews-spacexai-grok-imagine-api-the-1-video-model-best-pric-6aaa9578","sourceUrl":"https://www.latent.space/p/ainews-spacexai-grok-imagine-api"},{"id":"715c2eb4-93d7-4dd8-94c0-9b4e0d96c041","title":"We gotta talk about AI as a programming tool for the arts","summary":"I have been at least gently skeptical about all this stuff for the last two years. We gotta talk about AI as a programming tool for the arts\nChris Ashworth is the creator and CEO of QLab, a macOS software package for “cue-based, multimedia playback” which is designed to automate lighting and audio f","author":"Simon Willison","authorId":"simon-willison","published":"2026-01-30","topics":["Generative AI","AI-Assisted Programming","AI Ethics"],"diataxisType":"explanation","path":"/kb/articles/we-gotta-talk-about-ai-as-a-programming-tool-for-the-arts-715c2eb4","sourceUrl":"https://simonwillison.net/2026/Jan/30/a-programming-tool-for-the-arts/#atom-everything"},{"id":"b51d03eb-2cca-4ca0-b8f6-7fcff0e36f3f","title":"Building agents with the Claude Agent SDK","summary":"The key design principle behind Claude Code is that Claude needs the same tools that programmers use every day. Building agents with the Claude Agent SDK\nThe Claude Agent SDK is a collection of tools that helps developers build powerful agents on top of Claude Code. In this article, we walk through ","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-01-29","topics":["Agent-Native Architecture","Claude Code","AI Agents"],"diataxisType":"tutorial","path":"/kb/articles/building-agents-with-the-claude-agent-sdk-b51d03eb","sourceUrl":"https://www.anthropic.com/engineering/building-agents-with-the-claude-agent-sdk"},{"id":"6b0d2516-9cf6-4750-aa0f-c7e548ba67c0","title":"[AINews] Sam Altman's AI Combinator","summary":"I tend to think in terms of constraints. On the consumption side, human attention is the rate-limiting factor. AI News for 1/27/2026-1/28/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (206 channels, and 7100 messages) for you. Estimated reading time saved (at 200wpm): 559 minutes. AIN","author":"Swyx","authorId":"swyx","published":"2026-01-29","topics":["AI Agents","OpenAI API","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/ainews-sam-altmans-ai-combinator-6b0d2516","sourceUrl":"https://www.latent.space/p/ainews-sam-altmans-ai-combinator"},{"id":"6c56c360-2a79-4094-b447-30b5c8d5c288","title":"Equipping agents for the real world with Agent Skills","summary":"As model capabilities improve, we can now build general-purpose agents that interact with full-fledged computing environments. Equipping agents for the real world with Agent Skills\nClaude is powerful, but real work requires procedural knowledge and organizational context. Introducing Agent Skills, a","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-01-29","topics":["Agent-Native Architecture","Claude Code","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/equipping-agents-for-the-real-world-with-agent-skills-6c56c360","sourceUrl":"https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills"},{"id":"69a2bb9a-453d-44f3-83c7-68d0fdb11aa4","title":"Writing effective tools for agents — with agents","summary":"Tools are a new kind of software which reflects a contract between deterministic systems and non-deterministic agents. The\nModel Context Protocol (MCP)\ncan empower LLM agents with potentially hundreds of tools to solve real-world tasks. But how do we make those tools maximally effective?\nIn this pos","author":"Anthropic Engineering","authorId":"anthropic-engineering","published":"2026-01-29","topics":["Prompt Engineering","Agent-Native Architecture","Anthropic API","AI Agents"],"diataxisType":"how-to","path":"/kb/articles/writing-effective-tools-for-agents-with-agents-69a2bb9a","sourceUrl":"https://www.anthropic.com/engineering/writing-tools-for-agents"},{"id":"150690ee-802b-48e3-9bdc-d959b48fa90e","title":"Teach Your AI to Think Like a Senior Engineer","summary":"I’ve written about why having your AI coding assistant plan before it codes lets you ship faster than jumping straight to code. *While we’re on our Think Week offsite this week, we’re resurfacing* ***[Cora](https://cora.computer/)*** *general manager* ***[Kieran Klaassen](https://every.to/@kieran_13","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-29","topics":["Agent-Native Architecture","AI Agents","Prompt Engineering"],"diataxisType":"tutorial","path":"/kb/articles/teach-your-ai-to-think-like-a-senior-engineer-150690ee","sourceUrl":"https://every.to/source-code/teach-your-ai-to-think-like-a-senior-engineer-789ba7ca-ca7c-45a1-91fa-4178f59f226f"},{"id":"118be905-c1aa-4785-9aa4-d5cd3259c359","title":"Stop Coding and Start Planning","summary":"AI made us sloppy because it made us forget how to plan. *While we’re on our Think Week offsite this week, we’re resurfacing* ***[Cora](https://cora.computer/)*** *general manager* ***[Kieran Klaassen](https://every.to/@kieran_1355)****’s work on the theme of compound engineering. In this piece, Kie","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-28","topics":["Agent-Native Architecture","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/stop-coding-and-start-planning-118be905","sourceUrl":"https://every.to/source-code/stop-coding-and-start-planning-be0b4fd1-5898-4b09-bfda-0b00ea0004fd"},{"id":"8a8141ef-4db9-421e-b5b4-df9afca46ec5","title":"🔬 Automating Science: World Models, Scientific Taste, Agent Loops — Andrew White","summary":"ChemCrow story: GPT-4 + React + cloud lab automation, released March 2023, set off a storm of anxiety about AI-accelerated bioweapons/chemical weapons. Editors note: Welcome to our new AI for Science pod, with your new hosts RJ and Brandon! See the writeup on Latent.Space (https://Latent.Space) for ","author":"Swyx","authorId":"swyx","published":"2026-01-28","topics":["AI Agents","Scientific Discovery","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/-automating-science-world-models-scientific-taste-agent-loop-8a8141ef","sourceUrl":"https://www.latent.space/p/automating-science-world-models-scientific"},{"id":"20d4d25a-3dc9-4a6d-b3e4-cd562bffbe39","title":"[AINews] Moonshot Kimi K2.5 - Beats Sonnet 4.5 at half the cost, SOTA Open Model, first Native Image+Video, 100 parallel Agent Swarm manager","summary":"Kimi has been on an absolute tear in the past year. AI News for 1/26/2026-1/27/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (206 channels, and 7476 messages) for you. Estimated reading time saved (at 200wpm): 602 minutes. AINews website lets you search all past issues. As a reminder,","author":"Swyx","authorId":"swyx","published":"2026-01-28","topics":["AI Agents","Agent-Native Architecture","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/ainews-moonshot-kimi-k25-beats-sonnet-45-at-half-the-cost-so-20d4d25a","sourceUrl":"https://www.latent.space/p/ainews-moonshot-kimi-k25-beats-sonnet"},{"id":"afdf8983-6033-46ec-85da-e466e0b89d86","title":"[AINews] Anthropic launches the MCP Apps open spec, in Claude.ai","summary":"3 months after OpenAI floated a trial balloon with ChatGPT Apps and the Apps SDK at Dev Day 2025, Anthropic has now officially absorbed the independent MCP UI project. AI News for 1/23/2026-1/26/2026. We checked 12 subreddits, 544 Twitters and 24 Discords (206 channels, and 14285 messages) for you. ","author":"Swyx","authorId":"swyx","published":"2026-01-27","topics":["Anthropic API","Agent-Native Architecture","OpenAI API"],"diataxisType":"explanation","path":"/kb/articles/ainews-anthropic-launches-the-mcp-apps-open-spec-in-claudeai-afdf8983","sourceUrl":"https://www.latent.space/p/ainews-anthropic-launches-the-mcp"},{"id":"716f6ab5-3e63-4bdc-a1d1-da6f7edd57b3","title":"My AI Had Already Fixed the Code Before I Saw It","summary":"Before I opened my laptop, the code had reviewed itself. *While we’re on our Think Week offsite this week, we’re resurfacing* ***[Cora](https://cora.computer/)*** *general manager* ***[Kieran Klaassen](https://every.to/@kieran_1355)****’s work on the theme of compound engineering. In this piece, Kie","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-27","topics":["AI Agents","Prompt Engineering","Compounding Engineering","Claude Code","TDD"],"diataxisType":"explanation","path":"/kb/articles/my-ai-had-already-fixed-the-code-before-i-saw-it-716f6ab5","sourceUrl":"https://every.to/source-code/my-ai-had-already-fixed-the-code-before-i-saw-it-f4a29a07-ea95-409f-bcb2-487a970bed4a"},{"id":"889ad878-6e91-47e2-a54c-36884e504307","title":"The Complete Guide","summary":"Contents\nIntroduction 3\nFundamentals 4\nPlanning and design 7\nT esting and iteration 14\nDistribution and sharing 18\nPatterns and troubleshooting 21\nResources and references 28\n2","author":"Resources Anthropic","authorId":"resources-anthropic","published":"2026-01-26","topics":[],"diataxisType":"tutorial","path":"/kb/articles/the-complete-guide-889ad878","sourceUrl":"https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf"},{"id":"929cb875-06c2-4cc8-b456-093bdb8016fe","title":"How I Use Claude Code to Ship Like a Team of Five","summary":"Every piece of code I’ve shipped in the last two months was written by AI. Every piece of code I’ve shipped in the last two months was written by AI. Not assisted by AI. Written by AI.","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-26","topics":["Claude Code","Agent-Native Architecture","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/how-i-use-claude-code-to-ship-like-a-team-of-five-929cb875","sourceUrl":"https://every.to/source-code/how-i-use-claude-code-to-ship-like-a-team-of-five-6f23f136-52ab-455f-a997-101c071613aa"},{"id":"815e6177-e45e-49cf-ad4f-89d116ff50a5","title":"Captaining IMO Gold, Deep Think, On-Policy RL, Feeling the AGI in Singapore — Yi Tay","summary":"If one model can’t do it, can we get to AGI? From shipping Gemini Deep Think and IMO Gold to launching the Reasoning and AGI team in Singapore, Yi Tay has spent the last 18 months living through the full arc of Google DeepMinds pivot from architecture research to RL-driven reasoningwatching his team","author":"Swyx","authorId":"swyx","published":"2026-01-23","topics":["AI Agents","Prompt Engineering","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/captaining-imo-gold-deep-think-on-policy-rl-feeling-the-agi--815e6177","sourceUrl":"https://www.latent.space/p/captaining-imo-gold-deep-think-on"},{"id":"2e6d3626-efce-416f-a66f-6a3520433e87","title":"I Stopped Reading Code. My Code Reviews Got Better.","summary":"Now my code reviews no longer involve reading code. The bug report was deceptively simple: A user noticed that their email signature formatting was off in **[Cora](https://cora.computer)**, our AI-powered email assistant. I asked Claude Code to investigate and fix it. By morning, the fix had touched","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-23","topics":["Agent-Native Architecture","AI Agents","Code Review"],"diataxisType":"explanation","path":"/kb/articles/i-stopped-reading-code-my-code-reviews-got-better-2e6d3626","sourceUrl":"https://every.to/source-code/i-stopped-reading-code-my-code-reviews-got-better"},{"id":"943b66e6-3dde-4ce3-80bd-431aafcf58cd","title":"What the Team Behind Cursor Knows About the Future of Code","summary":"The IDE is kind of dead. *Happening now: We’re hosting with the world’s best experts pushing the limits of what’s possible. Watch live now until 6 p.m. ET, and . Also: This article is based on a sponsored event. Cursor provided $100 in credits to attendees and made this camp possible.—[Kate Lee](htt","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2026-01-22","topics":["Agent-Native Architecture","AI Agents","Cursor","Model Selection"],"diataxisType":"explanation","path":"/kb/articles/what-the-team-behind-cursor-knows-about-the-future-of-code-943b66e6","sourceUrl":"https://every.to/source-code/what-the-team-behind-cursor-knows-about-the-future-of-code"},{"id":"ba939aec-69c3-4d0c-93f7-43c83d5b2155","title":"don’t waste your back pressure","summary":"If you aren’t capturing your back-pressure then you are failing as a software engineer. I am fortunate to be surrounded by folks who listen and the link below post will go down as a seminal reading for people interested in AI context engineering.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-01-17","topics":["Software Engineering","Performance Engineering","AI Agents"],"diataxisType":"explanation","path":"/kb/articles/dont-waste-your-back-pressure-ba939aec","sourceUrl":"https://ghuntley.com/pressure/"},{"id":"205a3a00-f8a8-4ead-a193-2f5293c04c85","title":"everything is a ralph loop","summary":"Standard software practices is to build it vertically brick by brick - like Jenga but these days I approach everything as a loop. I’ve been thinking about how I build software is so very very different how I used to do it three years ago.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2026-01-17","topics":["Agent-Native Architecture","Evolutionary Software","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/everything-is-a-ralph-loop-205a3a00","sourceUrl":"https://ghuntley.com/loop/"},{"id":"70a35cc7-a237-4605-bad6-2c517e3c9aac","title":"Crypto scammers are using my name. Don't fall for it.","summary":"I woke up this morning to discover that, without my knowledge or consent, someone had created an ICO using the Superpowers name and my Twitter username. I woke up this morning to discover that, without my knowledge or consent, someone had created an ICO using the Superpowers name and my Twitter user","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-01-14","topics":["ICO (Initial Coin Offering)","Digital Identity Theft","Crypto Scams"],"diataxisType":"explanation","path":"/kb/articles/crypto-scammers-are-using-my-name-dont-fall-for-it-70a35cc7","sourceUrl":"https://blog.fsck.com/2026/01/14/crypto-scams/"},{"id":"0efafc8a-24d4-4c7d-be46-2e0a6c2df5dc","title":"I started a company!","summary":"This is just a quick post to say that I've started a company. This is just a quick post to say that I've started a company.","author":"Jesse Chen","authorId":"jesse-chen","published":"2026-01-13","topics":["Hiring Challenges","AI Agents","Startup Culture"],"diataxisType":"explanation","path":"/kb/articles/i-started-a-company-0efafc8a","sourceUrl":"https://blog.fsck.com/2026/01/13/I-started-a-company/"},{"id":"135ce2cf-0896-4706-a4e3-35c72b689544","title":"Streamlinear, a new MCP for Linear","summary":"My AI coding agents have been using Linear as the project and issue tracking tool on a new project. I've been using [Linear](https://linear.app) as the project and issue tracking tool on a new project.","author":"Jesse Chen","authorId":"jesse-chen","published":"2025-12-27","topics":["Prompt Engineering","AI Agents","Agent-Native Architecture"],"diataxisType":"explanation","path":"/kb/articles/streamlinear-a-new-mcp-for-linear-135ce2cf","sourceUrl":"https://blog.fsck.com/2025/12/27/streamlinear/"},{"id":"b3affcf5-9765-48a1-a365-69064ce8a3d6","title":"Wordiest","summary":"This whole time, I have written zero lines of code. I have read zero lines of code. My all-time favorite word game is called 'Wordiest'. It was an earlyish Android game with a pretty simple mechanic: You get 14 Scrabble-style tiles and need to make the two highest-scoring words you can. Wordiest had","author":"Jesse Chen","authorId":"jesse-chen","published":"2025-12-20","topics":["OpenAI API","Prompt Engineering","Mobile Development"],"diataxisType":"explanation","path":"/kb/articles/wordiest-b3affcf5","sourceUrl":"https://blog.fsck.com/2025/12/20/wordiest/"},{"id":"e8afce40-487d-43e8-93cf-a4903531f9d8","title":"OpenAI Gave Us a Glimpse Into Their AI Coding Playbook","summary":"Every session is onboarding this new coworker [anew]. Four OpenAI engineers built the Android version of the [image generation app Sora](https://every.to/vibe-check/openai-made-video-creation-effortless-here-s-what-happened-next) in 28 days. Naturally, they built it using [Codex](https://every.to/vi","author":"Dan Shipper (Every)","authorId":"dan-shipper-every","published":"2025-12-19","topics":["Pair Programming","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/openai-gave-us-a-glimpse-into-their-ai-coding-playbook-e8afce40","sourceUrl":"https://every.to/source-code/openai-gave-us-a-glimpse-into-their-ai-coding-playbook"},{"id":"b1aa53b4-9bce-4dd2-8ab5-a1b352fe083c","title":"How Every Is Harnessing the World-changing Shift of Opus 4.5","summary":"When Every CEO Dan Shipper started using Opus 4.5, his previous frustrations with AI coding—errors, hallucinations—disappeared.

$\"Source$

by 😎 The following was developed last month and has already been delivered at two conferences. If you would like for me to run a workshop similar to this at your employer, pleaseget in contact.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-08-24","topics":["Personal Development","AI Agents","Prompt Engineering"],"diataxisType":"tutorial","path":"/kb/articles/how-to-build-a-coding-agent-free-workshop-e51acff6","sourceUrl":"https://ghuntley.com/agent/"},{"id":"07887c8d-53b2-4eac-8d91-e2e570953dde","title":"too many model context protocol servers and LLM allocations on the dance floor","summary":"Everyone at the event was buzzing about the glory and how amazing MCP is going to be, or is, but when I pushed folks for their understanding of fundamentals, it was crickets. This blog post intends to be a definitive guide to context engineering fundamentals from the perspective of an engineer who b","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-08-22","topics":["Visual Studio Code","AI Bubble","Model Context Protocol"],"diataxisType":"explanation","path":"/kb/articles/too-many-model-context-protocol-servers-and-llm-allocations--07887c8d","sourceUrl":"https://ghuntley.com/allocations/"},{"id":"1021a5d2-3073-408e-8a31-589d53af289e","title":"what is the point of libraries now that you can just generate them?","summary":"Open-source, by design, is not financially sustainable. It's a meme as accurate as time. The problem is that our digital infrastructure depends upon just some random guy in Nebraska.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-08-20","topics":["Marketing in Open Source","Open Source Sustainability","AI in Software Development"],"diataxisType":"explanation","path":"/kb/articles/what-is-the-point-of-libraries-now-that-you-can-just-generat-1021a5d2","sourceUrl":"https://ghuntley.com/libraries/"},{"id":"952b6277-0d52-4a7c-bfae-2353c3cebae6","title":"this should not be possible","summary":"It might surprise some folks, but I'm incredibly cynical when it comes to AI and what is possible; yet I keep an open mind. It might surprise some folks, but I'm incredibly cynical when it comes to AI and what is possible; yet I keep an open mind. That said, two weeks ago, when I was in SFO, I disco","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-07-19","topics":["LLMs","AI Agents","Deliberate Practice","eBPF"],"diataxisType":"explanation","path":"/kb/articles/this-should-not-be-possible-952b6277","sourceUrl":"https://ghuntley.com/no/"},{"id":"d99c1db2-df4f-4176-ba36-dda297080530","title":"source code analysis of Amazon Kiro","summary":"Amazon Kiro's architecture reveals a sophisticated approach to scalable cloud services. > Amazon Kiro's architecture reveals a sophisticated approach to scalable cloud services.","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-07-14","topics":["Source Code Analysis","Amazon Kiro","Software Development Practices"],"diataxisType":"explanation","path":"/kb/articles/source-code-analysis-of-amazon-kiro-d99c1db2","sourceUrl":"https://ghuntley.com/amazon-kiro-source-code/"},{"id":"6d0f6075-82ce-4ad5-ac47-746820014427","title":"Ralph Wiggum as a \"software engineer\"","summary":"Ralph is a technique. In its purest form, Ralph is a Bash loop. > How Ralph Wiggum went from 'The Simpsons' to the biggest name in AI right now - [Venture Beat](https://venturebeat.com/technology/how-ralph-wiggum-went-from-the-simpsons-to-the-biggest-name-in-ai-right-now?ref=ghuntley.com)","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-07-14","topics":["Software Development Practices","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/ralph-wiggum-as-a-software-engineer-6d0f6075","sourceUrl":"https://ghuntley.com/ralph/"},{"id":"60792e78-16db-4243-95eb-c92c14ed4ae4","title":"Claude Sonnet is a small-brained mechanical squirrel of ","summary":"most people think each LLM is competing with each other, in part they are but if you play around enough with them you'll notice each provider has a particular niche and they are fine-tuning towards th This post is a follow-up from [LLMs are mirrors of operator skill](https://ghuntley.com/mirrors/) i","author":"Geoffrey Huntley","authorId":"geoffrey-huntley","published":"2025-07-02","topics":["Agent-Native Architecture","AI Agents","Prompt Engineering"],"diataxisType":"explanation","path":"/kb/articles/claude-sonnet-is-a-small-brained-mechanical-squirrel-of-t-60792e78","sourceUrl":"https://ghuntley.com/cars/"}];

​Kieran Klaassen

​Articles

Hi | Kieran Klaassen

Kieran Klaassen

Articles