Mathematicians around the world are very angry with Sam Altman's OpenAI

OpenAI has terminated its sponsorship of a mathematics event organized by the California Institute of Technology (Caltech) following criticism from a group of current and former mathematicians. The dispute arises as AI companies, including OpenAI, increasingly leverage large language models (LLMs) to address complex research issues. OpenAI's decision follows an open letter from Caltech mathematicians, who expressed concerns that the event may have detrimental effects on the mathematical community.

The mathematicians' primary grievance is not merely the application of AI in mathematics but also how AI companies present their claimed breakthroughs and the subsequent verification process. They argue that AI companies often present their models as having solved difficult theoretical problems, leaving human mathematicians to verify the work, develop proofs, and ascertain the validity of these claims.

This practice, dubbed "sloppy mathematics," results in research mathematicians having to step in to properly review, disseminate, and sometimes debunk the claimed results, a task that goes unchecked, uncredited, and unacknowledged.

The tension escalated further after OpenAI claimed its AI agents had solved the Navier-Stokes equations, a set of mathematical equations studied for nearly 90 years to describe the movement of liquids and gases. This announcement drew significant criticism from mathematicians. In response to the backlash, OpenAI's research lead, Dan Roberts, acknowledged the disruptive impact of AI advancements in mathematics and expressed the company's intention to collaborate with the math community to determine the appropriate integration of this technology and its communication of implications.

The Caltech Math Hackathon, the event in question, aims to provide young mathematicians with an opportunity to utilize AI tools while tackling challenging research questions. Participants compete in a 40-hour round, using $20,000 in AI tokens to address open mathematical problems. Successful teams can advance to a second round, receiving six months and additional AI tokens to continue their research.