Some mathematicians call for OpenAI boycott after AI-generated proofs flood their field

The cost of using AI coding assistants drops dramatically, making them broadly accessible for basic tasks.

The Association for Human Mathematics (AHM), led by Fields Medalist Terence Tao, has called for a boycott of OpenAI following the company's release of hundreds of AI-generated mathematical manuscripts. Many of these proofs are dense and require AI assistance to decipher, raising questions about whether a problem is truly solved if its proof is not human-understandable. The AHM accuses OpenAI of violating research norms by publishing these proofs en masse and urges mathematicians to cease collaboration with the company. This action escalates a debate that began when OpenAI claimed its internal models solved over 100 open math problems in a month, including an approach to the Navier-Stokes Millennium Problem. OpenAI previously established an advisory group, AGMAI, to address concerns, but the AHM argues OpenAI has ignored its advice and that the release is a "demonstration of power" rather than scholarship, potentially stemming from the company's ongoing copyright disputes.

The core of the conflict lies in the nature of scientific discovery: whether AI-generated outputs that are incomprehensible to humans can be considered valid solutions. This challenges the traditional research process, where understanding, verification, and dissemination by human experts are paramount. The AHM's call for a boycott signifies a potential schism in the mathematical community, forcing a reevaluation of what constitutes mathematical progress and the role of AI in its pursuit. The situation highlights a growing tension between rapid AI advancement and established academic integrity, with significant implications for how future discoveries are validated and integrated into human knowledge.

This event signals a critical juncture for both AI development and academic research, as the pace of AI-generated discoveries outstrips human capacity for verification and comprehension. The AHM's stance suggests a market bifurcation: one where AI assists human researchers within established frameworks, and another where AI acts as an autonomous discoverer, demanding new verification paradigms. The implications extend to funding, publishing, and career progression in mathematics, potentially devaluing traditional research skills in favor of AI interpretation. Future developments will likely focus on establishing protocols for AI-human collaboration and ethical AI deployment in scientific fields.