25 Fields Medalists Criticize AI Companies for Harming Mathematical Research

ME News report, September 12 (UTC+8), Beating AI Bulletin: Twenty-five Fields Medalists, including Terence Tao, Deng Yu, and Peter Scholze, have jointly issued a statement asserting that AI companies are treating the resolution of major mathematical problems as benchmarks, thereby harming mathematical research. The statement acknowledges that large models have made rapid progress in recent months and have already solved important unsolved problems across multiple fields, but mathematics places greater value on the new concepts, methods, and insights generated by such solutions. Their greatest concern is that AI companies are rushing to announce results prematurely. The statement notes that some solutions have been publicly disclosed before complete proofs could be formalized, new methods refined, or prior citations properly included, leading to serious issues of attribution and plagiarism. Without mathematicians to further process and organize these findings, the ideas discovered by AI struggle to be meaningfully integrated into the mathematical framework. Although the statement does not name OpenAI specifically, a companion article in The Economist directly titled its piece “Top Mathematicians Are Furious at OpenAI’s Approach.” Just days earlier, OpenAI became embroiled in controversy over attribution and undisclosed research related to its purported proof of the Navier-Stokes equations. Terence Tao stated that this statement was rapidly formulated by the 25 Fields Medalists following discussions over the past week, as they believe the situation has become urgent. (Source: BlockBeats)