Following up on its historic Millennium Prize breakthrough, OpenAI has upended the scientific world once again by releasing more than 700 papers containing over 300 advanced mathematical proofs. The massive drop—which includes advances touching three of the five remaining Millennium Prize Problems, including the Riemann hypothesis—has left mathematicians stunned, with Rutgers University professor Alex Kontorovich noting that such a feat would instantly secure a Fields Medal for a human researcher.
However, the release highlights a deep rift between Silicon Valley labs and the academic community. The unprecedented AI output comes on the heels of fierce pushback, including an open letter from more than two dozen Fields Medalists warning that the industry’s rush for mathematical dominance threatens to undermine the core purpose of the discipline. Critics have raised concerns over corporate rivalry, transparency, and whether AI models may have inadvertently drawn on unpublished academic work.
Although OpenAI consulted an independent advisory group of mathematicians to shape the guidelines for this release, tension remains high. While the math advisory council urged labs to halt testing advanced problems on proprietary frontier models, OpenAI defended its approach, asserting that evaluating AI on complex mathematics is crucial for accelerating scientific progress.
