Twenty-five leading mathematicians have signed an open letter claiming their intellectual work is under threat as AI labs aim to outdo each other in solving famous math problems. Each signer has been awarded the Fields Medal, considered the most prestigious award in mathematics.
New York University professor Tristan Buckmaster this week accused OpenAI of pressuring people to discredit Anthropic collaborators who solved important math problems, and questioned whether the company used its collaboration with Codex to create its own groundbreaking proof during a long weekend of reasoning.
OpenAI on Thursday withdrew its sponsorship of a math event at the California Institute of Technology after the company was criticized by researchers at the university.
The ability of AI models to solve the world’s unsolved mathematical challenges could be a boon for humanity, but the signatories of the new letter argue that this is only true if the solutions are understood and communicated to the mathematical community and ultimately to the rest of the world.
“Often these solutions are published in a hurry, leaving no time for proper articles, isolating new methods and ideas, and citing relevant past work of others,” they write, but OpenAI’s proof remains unverified. “As with all creative professions, this raises serious questions about attribution and plagiarism. Moreover, without motivated mathematicians who must take care of their development and integration into the mathematical canon, the ideas conceived by AI will never fully come to life, and the important human chain of communication between mathematicians will be lost.”
There is a real fear that other mathematicians will become paranoid and wonder if their use of the Codex will be reflected in OpenAI’s new models, threatening the culture of open research. Today, if a frontier research institution finds a useful path to discovery, it can use LLM to beat the original researchers and spend tens of millions of dollars to prove it. This is a dynamic that encourages secrecy.
The letter follows the Leiden Declaration published by a working group of mathematicians in June. The paper also addresses how LLM proofs change work and provides a set of recommendations for mathematicians, institutions, and policy makers.
As in software engineering and other fields where AI tools are changing workflows, mathematicians are finding legitimacy in their work. The value of mathematics lies not just in the proofs and who gets the credit, but in the intellectual superstructure that develops students, discovers new questions and ideas, and integrates them into broader human civilization.
And if you’re not particularly interested in the harsh world of high-stakes mathematical proofs, don’t forget that the next field is yours.
“The challenges currently facing the mathematical community are similar to those faced by other scientific and creative professionals, and illustrate the challenges that all of humanity may face: As AI transforms the way we do work, how can we ensure that we don’t lose sight of what that work was intended to accomplish in the first place?” they wrote.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
