Claim: gpt-5-pro can prove new interesting mathematics. Proof: I took a convex optimization paper with a clean open problem in it and asked gpt-5-pro to work on it. It proved a better bound than what is in the paper, and I checked the proof it's correct. Details below.

Looks like a pretty long CoT

Can’t beat them? Connect your broker to trade like them.

Over the past few months, OpenAI models crossed a threshold: we’re seeing early/small-scale but repeated examples of GPT-5 meaningfully contributing to novel research. AI is the next great scientific instrument, and it benefits every field. Progress accelerates when researchers

It was an amazing experience visiting @OpenAI - chatting and exchanging ideas with brilliant researchers. Many of them strong chess players too!♟️🤖 After our discussions, I played an 8-board clock simul, and they really made me sweat! 😅 They played so well. I won 7 games and

yes

What’s your top storage priority for the next 12–18 months?

yes

Here is the story of a remarkable, independent treatment suggestion by GPT-5 Pro: repurposing a known drug for a patient with food protein–induced enterocolitis syndrome (FPIES). First, how we came to test this. My close friend, physician-scientist Dr. Oral Alpan, treated the

100% agree on the productivity boost. One just needs patience to correct mistakes, which are more subtle than before imo. I had a nice interaction with GPT-5-pro while proving a convex analysis lemma: https://t.co/N4oEv0S37X The model didn’t write the full proof, but the

I firmly believe we are at a watershed moment in the history of mathematics. In the coming years, using LLMs for math research will become mainstream, and so will Lean formalization, made easier by LLMs. (1/4)

Great profile on my OpenAI for Science colleague @ALupsasca, his quest to understand photon rings around black holes, and the opportunity he sees in using AI to accelerate science:

Wow!! I tried to prove this in 2019, without success. This is embarrassing lol I remember in 2023 @SebastienBubeck told me something like "Can an LLM rediscover Nesterov's acceleration? Not now, but certainly soon"

Following defense stocks ? AI Defense Stock @VWAVInc VisionWave Holdings ( Nasdaq: $VWAV) Unveils the Varan UGV — Leading the Ground Autonomy Revolution Varan UGV, a next-generation, highly autonomous and modular ground platform engineered from the ground up with the goal to

We ❤️ our Chief Scientist Jakub (@merettm)! Happy Halloween!

One year at OpenAI today ... It has been quite a ride, even more exciting than I expected! I joined as GPT-4.5 was being prepared, got some minor action on o3, and dove deeper for GPT-5. Being a spectator to the multi-modal releases (sora-1, imagegen, sora-2) was also absolutely

I used ChatGPT to solve an open problem in convex optimization. *Part I* (1/N)

Are you ready for another twist in the Erdos problems story? Today a $1000 prize open problem (#707 https://t.co/mqJnSZEUvh) was solved by two humans who … realized that the problem was solved 30 years before being posed 🤣. Maybe lit search is not so simple after all 😅. But

My posts last week created a lot of unnecessary confusion*, so today I would like to do a deep dive on one example to explain why I was so excited. In short, it’s not about AIs discovering new results on their own, but rather how tools like GPT-5 can help researchers navigate,

I deleted the post, I didn't mean to mislead anyone obviously, I thought the phrasing was clear, sorry about that. Only solutions in the literature were found that's it, and I find this very accelerating because I know how hard it is to search the literature.

Definitely!

I don't do podcasts very often - in reality this is my first one ever, but if anyone wants to listen to someone talk about RL for an hour, this is it