Summary
OpenAI has published a blog post describing scientific results in math, physics and computer science that it says were generated by its AI models. Wes Roth walks through the announcement and the reaction to it.
The headline claim is that an OpenAI model disproved the exponential time hypothesis (ETH), a widely believed conjecture closely tied to the P versus NP problem. Hundreds of papers rely on ETH to establish conditional hardness results in fields such as graph theory, computational biology and databases. The video explains P versus NP and why it matters for cryptography, with commentary from Scott Aaronson and a note from OpenAI president Greg Brockman that P probably still does not equal NP. The post also reportedly covers work related to the Yang-Mills existence problem, one of the Clay Millennium Prize problems.
The video also covers how OpenAI is handling credit. The company is not taking credit itself and has researchers Russell Impagliazzo, Ramamohan Paturi and Ryan Williams shepherd the results. It closes with the debate among mathematicians about AI-generated proofs and who gets to work on open problems.
📺 Source: Wes Roth · Published October 27, 2026
🏷️ Format: News Analysis
