Two dozen Fields Medal winners have signed a public declaration arguing that the rush to apply AI to mathematics is severely misaligned with the goals of the field. Published at mathandai.org, the declaration acknowledges that LLMs can now solve major outstanding problems across many areas of mathematics, but warns that treating those problems as benchmarks for AI companies damages both the science and the community behind it. Signatories include Terence Tao, Peter Scholze, Cedric Villani, Artur Avila, Manjul Bhargava and Maryna Viazovska.
The declaration argues that solving problems is only a proxy for the real aim of mathematics: conceptual understanding and insight. Mass-producing true or false statements at an ever faster pace, it warns, could destroy the fertile ground that gives rise to new ideas. The authors point to rushed announcements that leave no time for proper writeups, for isolating the new methods and ideas, or for citing the previous work of others, raising severe attribution and plagiarism questions. Without willing mathematicians to develop and integrate AI-conceived ideas into the mathematical canon, the human transmission chain between researchers would be lost.
The signatories frame the dispute as part of a broader alignment problem. AI systems trained on a vast body of human work can now produce the outputs of that work directly, so the goals of the work and the goals of the technology stop aligning. The declaration says the issue is urgent for mathematics, for the companies building these systems, and for a society that will confront the same problem across many other forms of intellectual work. Anyone who supports it can add their name using ORCID.