Member of Technical Staff (RSI Lab) https://t.co/ptU2Xh4aO1 If you are a visionary builder ready to move to Tokyo and engineer the engine of recursive discovery, we invite you to apply. https://t.co/ydKhZQiup8

Not only do we want to train a good model, we want to know it'll be good before we even start training. About a month ago, the Marin team launched a 129B (16B active) 1e23 FLOPs MoE run and preregistered a loss of 2.252. The run finished this past week and landed at 2.234. https://t.co/OptaVa7jIO

A mic drop moment @ycombinator tonight @sama just offered $2M in OpenAI tokens to EVERY YC startup in the current batch in exchange for equity Just like Yuri Milner offering to invest in every startup back when Sam was a YC partner I can't wait to see what's unlocked when you let the most driven, creative and formidable founders tokenmaxx

It’s Codex Thursday, and yes, we have updates for you. First up: Appshots, a new way to bring the context of what you’re working on into Codex. On your Mac, press Command-Command to attach your app window to a Codex thread. Codex gets both a screenshot and text from the window, including content beyond what’s visible onscreen. Appshots are available across plans on Mac, with enterprise access coming soon.

AI should dramatically increase quality of life and individual freedoms for people around the world. The OpenAI Foundation is making an initial $250M commitment to measurement, transition support, and new approaches to broadly shared prosperity. https://t.co/zOD8O94RjQ

It's time to fly. https://t.co/ObUaCZ07EM

The AI race is no longer just about models. It is about energy. China’s wind-powered underwater datacentre shows how the next wave of AI infrastructure may be built around new approaches to power, cooling and efficiency. The winners in AI may not only be the companies with the smartest algorithms, but also those with the most sustainable infrastructure.

https://t.co/IVGHTjLNTf

AI safety is becoming a launch strategy. Anthropic’s release of a Mythos-class model shows how leading AI companies are trying to balance powerful new capabilities with safeguards for high-risk areas. The real test is no longer whether models can do more. It is whether society trusts how they are released.

AI may help find the next football superstar, but it should not define what talent is. Data can make scouting faster and more disciplined, but football still depends on qualities that are hard to measure: instinct, resilience, creativity and the ability to surprise. The risk is not using AI in sport. The risk is believing the spreadsheet sees everything.

https://t.co/QGVSJcNhOB

The AI race is becoming a price war. If OpenAI cuts token prices to compete with Anthropic, lower costs could accelerate enterprise adoption but also pressure margins across the AI stack. The next battle may not be who has the most powerful model. It may be who can deliver intelligence at the lowest sustainable cost.

The future of work will not be human versus AI. The people who thrive will be those who combine human judgment, creativity and trust with AI’s speed, scale and pattern recognition. The real career risk is not being replaced by AI. It is being outpaced by people who know how to use it better.

The next AI race may be won as much in factories as in research labs. China’s dominance in robot supply chains shows that building advanced AI is one challenge. Building millions of intelligent machines at scale is another. In the age of AI and robotics, manufacturing strength is becoming a strategic advantage again.

The AI IPO boom is turning into a capital tsunami. Demand for SpaceX shares suggests investors are no longer asking whether AI will reshape the economy. They are asking how much exposure they can get before the next phase begins. The biggest risk today may not be a lack of capital. It may be finding enough opportunities to deploy it.

@ScottPatterson0 @maticrobots Yeah, I'm so proud it was launched in my home here last year: https://t.co/fEkfcuLxjt We love ours for same reasons. @mehul is one of the best consumer entrepreneurs America has.

“don’t train your own model” is common ai advice. it's wrong. your token bill's the proof. today, we’re excited to launch castform into open preview. castform is the easiest way for you to train your own model, on your own data. open-weights models are performant and much cheaper. when trained on your task & proprietary data, they beat closed models. the thing standing between you and that was weeks of plumbing & years of ml expertise. with castform, model training is as simple as prompt engineering. @castformai bring your agent traces or raw corpora. castform turns it into training data, picks the right algorithmic recipes, manages gpus, and gives you an ide to watch and chat with your model as it learns. see what you can build with castform👇

Meet Coinbase for Agents. Give your agent its own account to: → Execute trades & manage your portfolio → Run autonomously under guardrails → Pay for data & research tools via x402 (coming next week) Agentic finance is here, and it's powered by Coinbase. https://t.co/DK220fko0z

48 hours ago Apple announced its new Personal AI: Siri. Today Onairos announces Persona, the user data api to power all other Personal AI’s with context not even Siri has. https://t.co/ogBCIsM1xc

San Francisco comes through again. Spent a delightful hour with @khal_ism who is building a new brain computer interface that you will hear a lot about later this year. I wrote eight books about decade-long change coming. My first was about social media before this site were are all on started. If I write another it will be BCIs. Have a list of neuroscientists and builders working in this field. It will be a trend of new things coming over next decade. I already asked @elonmusk for a full neuralink. He said they aren’t ready for regular people yet, but told me “cognitive enhancements” are coming in three years. OpenAI, Meta, and others are working on them too. But now I know a lot more about how they work. Thanks X. This morning when I woke up I had no idea who I would meet. Another meeting in a few minutes. Can’t do this in most other places in the world. This morning I wrote that I had the day free, and already an amazing day. Plus the weather is so amazing today. And my other surreptitious meetings today? Entrepreneurs who are changing how we work. Deeply too, more on that soon.

Preview of some slides for this https://t.co/PcKT8T39Ax

Coming up against the same problems over and over? Tried a bunch of experiments to get unstuck but still not where you want to be? This year at Observe @HamelHusain will be hosting two 1-hour Offices Hours to work through challenges in what you're building. Grab your tickets for Observe now and don't miss out on this special session. https://t.co/NYEih97lij

Our evals course is 25% off this week! The next cohort will have completely refreshed material. @sh_reya and I will cover new topics like: using agents for evals (w/o the foot guns), thinking like a data scientist, optimization approaches for harnesses and more. Use this link to apply @lennysan 's special 25% off code: https://t.co/Fy8r7bAAxM

today, we’re excited to announce raindrop 2.0: self-healing agents. we now train custom models that autonomously detect hidden issues with your agent. i could tell you all about it, but wouldn’t you rather hear it from someone… else? https://t.co/dBGNNr9UNv

https://t.co/8HaeeSSfQL

https://t.co/8HaeeSSfQL

I love agentic map-reduce, but I don't love Claude's workflows feature. It's hard to steer the workflow, and it's not data-driven enough (e.g., reduce groups are often predefined, not emergent from the map outputs) HOWEVER...I was inspired by its UI to ship something similar for DocETL. Satisfying way to visualize progress when using the CLI 😆

Got my 10yr old introduced to Codex today. The excitement in his face tells it all. After struggling with Claude Code CLI for a bit, today he was like “this is the future, dad”. The Codex team built a beautiful app. https://t.co/ITM7Euzybl

NEW: Anthropic is walking back Claude Fable 5's policy to covertly degrade performance for competing AI researchers, after facing fierce backlash. “We’re changing Fable 5’s safeguards for frontier LLM development to make them visible,” Anthropic tells WIRED. “We made the wrong tradeoff and we apologize for not getting the balance right.”

and it happened: Poke is now back on WhatsApp in the whole European Union! wouldn't have been possible without all the help from twitter, such as @kishanbagaria and @lutherlowe, who help us start this investigation from me testifying in rome to @felixms in brussels last month. @interaction also won in brazil, turkey, and many more jurisdictions are on the way free market prevails! and the path is paved for the inevitable product form factor to win consumer AI

We just shipped a search and command bar to the OpenAI API Platform! Use ⌘𝙺 to search through: → Pages → Settings → Quick Actions → Resources → Developer Docs https://t.co/XFIiCE0w6n

Cooking. @steipete 🦞 https://t.co/hni2zTM9ZM