Fields Medalists Warn AI Labs Threaten Mathematics as OpenAI Dispute Escalates
Twenty-five Fields Medal winners signed an open letter saying AI labs racing to solve famous math problems threaten mathematicians' work, as OpenAI's dispute with researchers escalates.
According to a TechCrunch AI report, NYU professor Tristan Buckmaster this week accused OpenAI of pressuring him not to credit a collaborator who works for Anthropic for solving an important math problem, and questioned whether the company had used their work with Codex to produce its own ground-breaking proof over a marathon weekend of inference. OpenAI's proof remains unverified.
On Thursday, OpenAI withdrew its sponsorship of a math event at Caltech after the company was criticized by researchers at the university.
The signatories of the letter argue that AI models' ability to solve outstanding mathematical challenges could benefit humanity, but only if the solutions can be understood and communicated by the math community and, ultimately, the rest of the world. "Often these solutions are announced in a rush, leaving no time for a proper writeup, the isolation of new methods and ideas, and citing relevant previous work of others," they wrote. "As in all creative professions, this raises severe attribution and plagiarism questions. Moreover, without the willing mathematicians who must take care of their development and integration into the mathematical canon, AI-conceived ideas would never become fully alive and the crucial human transmission chain between mathematicians would be lost."
Other mathematicians have grown paranoid about whether their own use of Codex was in turn fed into OpenAI's new models, and researchers fear the culture of open research will be threatened. As things stand, a frontier lab that sees a useful path to a discovery can spend tens of millions of dollars using large language models to beat the original researchers to a proof, a dynamic the letter's signatories say will incentivize secrecy.
The letter follows the Leiden Declaration, released in June by a working group of mathematicians. That document also addresses how LLM-generated proofs will change their work and offers recommendations for mathematicians, institutions and policymakers.
As in software engineering and other fields where AI tools are changing workflows, mathematicians locate value in the work around the work: the worth of mathematics lies not only in proofs and who receives credit, but in the intellectual super-structure that nourishes students, generates new questions and ideas, and integrates them into broader human civilization. The signatories warned that other disciplines face similar pressures. "The issues the mathematical community faces now are similar to issues that other scientific and creative professions are facing, and indicate issues that all of humanity might face: how to make sure that, as AI changes the way work is done, we do not lose sight of what that work was meant to achieve in the first place," they wrote.