We wrapped up this year's competition circuit with a full score on the ICPC, after achieving 6th in the IOI, a gold medal at the IMO, and 2nd in the AtCoder Heuristic contest!

Instead of trying to prove it, I asked GPT5 about it, and in about 20 seconds received a proof. The proof relied on a lemma that I had not heard of (the statement was a bit outside my main areas), so although I am confident I'd have got there in the end, 2/3

Codex had its strongest growth in one day yesterday since the launch of gpt-5-codex. Way to motivate the team during a gnarly investigation that's making us go through every piece of infra, hardware and line of code in our system.

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

We often talk about big leaps in AI for mathematics, but I think the small steps are equally impressive. The future of mathematics is now. I was working on a particular task: finding a case-free proof of the representability of the local Néron function correction using a

@SebastienBubeck @kevinweil ChatGPT is now at the level of solving some math research questions, but you do need an expert guiding it. This exercise was a lot of fun and was highly productive. I also feel I'm getting better at prompting ChatGPT. I'll also try other open and unsolved problems. (16/N, N=16)

Excited to start OpenAI for Physics w/ @ALupsasca @kevinweil @aleks_madry and @merettm! I sat with @ALupsasca when GPT-5 reproduced his latest research paper, and we both felt parallels to watching AlphaGo play move 37. It's nearly impossible to be a world class chess player

The Training team @OpenAI is hiring researchers in London 🚀 Our twin missions are to train better LLMs, and serve them more cheaply Get in touch if you are excited to collaborate on architecture design, reliable scaling, and faster optimization

I guess it's now every day until the end of time

AI’s next leap isn’t “bigger models”, it’s more learnable ones What if models could learn faster, with less data, even while solving a task? Join @lukaszkaiser (co‑creator of the Transformer) for a map of how AI is evolving from RNNs to Reasoners to Researchers Link in thread

We aim to build the most intelligent and useful AI. But “useful” is a fuzzy word. GDPval (consisting of tasks spanning 44 occupations across the top 9 sectors contributing to US GDP) makes "usefulness" more concrete. Would love to saturate this one!

“The default way to code is vibecoding.” OpenAI chief research officer Mark Chen (@markchen90) says high schoolers already see coding from scratch as weird. If coding is vibes, what does great research look like?

Introducing Among AIs, a social reasoning benchmark where embodied models play Among Us to test social intelligence: deception, persuasion, and coordination. We put 6 SOTA models in a live arena and GPT-5 came out on top by leading in Impostor & Crewmate wins. Why did GPT-5 get

GPT-5 (the one we can actually use) solved 11/12 ICPC problems. Only the final hardest one needed OpenAI’s internal reasoning model. So the public already has access to a math/coding genius that just outperformed Google’s internal Gemini Deep Think. 🤯

take it from a goat:

.@OpenAI joined the experiment to incorporate AI development tools into the 49th Annual ICPC World Finals Local Judge. OpenAI's models solved all 12 problems – a milestone akin to achieving a gold medal at the highest level of achievement! https://t.co/ntUHfRAQzO

Any past contestant will tell you how much hard work, dedication, and creativity is needed to achieve these results. We're excited to bring these capabilities to bear on solving real world problems and to advance the scientific frontier.

Alignment is arguably the most important AI research frontier. As we scale reasoning, models gain situational awareness and a desire for self-preservation. Here, a model identifies it shouldn’t be deployed, considers covering it up, but then realizes it might be in a test.

Give GPT-5-Codex a try! Huge props to Andrey Mishchenko, @katyhshi, @hansonwng, @ssnl_tz, and @mia_glaese for turning our reasoning models’ raw intelligence into real-world coding performance - and uncovering new research along the way.

GPT-5-Codex is here: a version of GPT-5 better at agentic coding. It is faster, smarter, and has new capabilities. Let us know what you think! The team has been absolutely cooking, very fun to watch.