My notes on Anthropic's substantial essay about how they built their multi-agent research system, which has finally talked me around to taking multi-agent LLM prompt engineering seriously simonwillison.net/2025/Jun/14/mu…

trying to get a spot at tony's pizza in north beach - apparently one of the best restaurants on the planet - ain't easy

When reality hits you with a funky building.

And then there are the moments that change everything. Hyped to join the YC AI Startup school tomorrow and Tuesday. Bring it on life ❤️. Now this once poor kid from nowhereville got new snicks.

So heart warming ❤️🥹

What happens when an LLM is asked to use information that contradicts its knowledge? We explore knowledge conflict in a new preprint📑 TLDR: Performance drops, and this could affect the overall performance of LLMs in model-based evaluation.📑🧵⬇️ 1/8 #NLProc #LLM #AIResearch

As someone trained in stats, the gender ratio of the YC ai summer school is stark😅.

took a bit longer than expected but here we are

Our notes on Anthropic's latest blog on how they built Claude Deep Research. Excellent insights on building a multi-agent system!

Excited to introduce LLM-First Search (LFS) - a new paradigm where the language model takes the lead in reasoning and search! LFS is a self-directed search method that empowers LLMs to guide the exploration process themselves, without relying on predefined heuristics or fixed

second day at Y Combinator ai sus - excited to see andrej karpathy 🫶

Learning from the legend Satya Nadella at the Y Combinator AI SUS school. The big Q: “What is the next step up?”🚀 Made us a #ycaisus tag ✨.

Talk 2 of Day 2 Y Combinator #ycaisus brings Andrej Karpathy and his brilliant insights into magical new computers: “Whenever I am talking to ChatGPT […], it feels like I’m talking to an operating system.”