Altman Responds to AI Math Discovery Controversy: Findings Should Be Public, but Publication Methods Need Improvement

OpenAI CEO Sam Altman said on the "Cheeky Pint" podcast released on October 29 that he understands the pressure mathematicians face when confronted with a large volume of AI-generated results, but believes the public has a right to know about important scientific discoveries. He acknowledged that OpenAI needs to find release methods that are more considerate of the research community, allowing results to be made public while also receiving sufficient explanation and scrutiny. OpenAI recently released hundreds of mathematical manuscripts in a concentrated manner. Some researchers criticized the lack of adequate explanation, citation, and verification, making it difficult to quickly assess the value of the results. When the host asked if the company had more results beyond the 700 proofs already released, Altman answered "yes," but did not specify the number or release timeline. He also stated that the company's research on mathematical problems has cost hundreds of millions of dollars; specifically, work related to the Navier-Stokes equations cost approximately $10 million, possibly slightly more. The controversy over the release cadence has extended to other disciplines. OpenAI recently announced three research advances in mathematics, computer science, and physics, adopting different disclosure methods: releasing relevant materials for the first two before publication, while temporarily withholding the AI-generated raw materials for the physics result, allowing an external research team to review them first. These research claims still require further scrutiny by the academic community and cannot be considered recognized breakthroughs based solely on the company's announcements.