The Association for Human Mathematics called for a boycott of OpenAI after the company released over 700 files of AI-generated research. The group said, “Mathematicians did not ask for this work to be done.” They worry that using powerful models to solve complex problems ignores important scientific research norms.
The Association for Human Mathematics has called on mathematicians to boycott OpenAI after the company published findings on 377 maths problems and released a repository with more than 700 files. The group says mathematicians did not ask OpenAI to do this work and has criticised the company’s approach to AI-generated mathematical research. The call for boycott also comes after OpenAI said last month that one of its internal models had solved the Navier-Stokes Millennium Prize problem, one of mathematics’ best-known unsolved problems.
In a statement responding to OpenAI’s October 6 release, the Association for Human Mathematics said mathematicians “did not ask for this work to be done” and rejected the company’s claim that the release would help advance mathematics. The group also criticised OpenAI for releasing more than 700 files at once, calling it “not a demonstration of scholarship, but a demonstration of power”. It urged mathematicians and the public to be sceptical of the company’s approach to AI-generated mathematical research.
"Mathematicians did not ask for this work to be done. The Advisory Group on Mathematics and Artificial Intelligence, from whom OpenAI has claimed to derive its legitimacy, opened their initial advisory statement by saying that frontier AI corporations should not test advanced mathematical problems on internal models. In ignoring the central premise of the Advisory Group’s position, OpenAI has indicated total disregard for the norms of scientific research — norms that guarantee that mathematics remains trustworthy, ethically researched, and in the public interest," reads the official statement.
Why are mathematicians calling for a boycott of OpenAI?
One of the concerns raised by mathematicians is that AI companies are using their own models to work on difficult and unsolved maths problems. The Advisory Group on Mathematics and Artificial Intelligence, an independent group hosted by the Institute for Advanced Study in Princeton, has previously raised concerns about how these systems are being used in mathematics. In a September 29 statement, the group said AI can produce mathematical arguments that human researchers may not be able to understand, verify or take responsibility for.
The group also warned that limited access to powerful AI systems could create a “two-tier system” in mathematics. In simple terms, this means researchers at AI companies could have access to powerful models and computing resources that are not available to other mathematicians, giving them an advantage when working on difficult problems. The group has called for more equitable access to these systems.
Meanwhile, OpenAI assures that it has been consulting the advisory group and used its advice and public recommendations when deciding how to release its latest mathematical results. This week, OpenAI published the results in a GitHub repository and included formal versions of many proofs in Lean, a programming language that allows mathematical proofs to be checked by a computer. The company also shared 10 summaries of the model’s reasoning, details on the number of problems it attempted and estimates of the computing used. OpenAI said the average result required the equivalent of about three hours of ChatGPT Pro thinking.