OpenAI says it's no longer sponsoring Caltech's math hackathon, where participants use AI to solve research problems, following criticism of "slop mathematics"

The default behavior for AI companies is now to claim credit for mathematical discoveries made by their models, requiring human mathematicians to perform unpaid verification.

OpenAI has withdrawn its sponsorship of Caltech's math hackathon following criticism from mathematicians who described the event's premise as "slop mathematics." The event encouraged participants to use large language models to solve research problems. Mathematicians argued that AI companies claim solutions, leaving human researchers to perform the uncompensated labor of verification and proof. This controversy intensified when OpenAI claimed its AI agents solved the Navier-Stokes equations, a problem also recently researched by academics who suspected OpenAI may have used their chatbot conversations. OpenAI acknowledged that de-identified usage data could have contributed to model improvements. The mathematicians also expressed concern that the hackathon would serve as advertising for AI companies, allowing them to claim credit for student efforts. The event's organizers stated their intention was to foster excitement about mathematics using modern tools and that they were open to feedback.

The dispute highlights a growing tension between AI developers and the academic research community regarding intellectual property and the attribution of scientific breakthroughs. Mathematicians are pushing back against the narrative that AI can independently solve complex problems, emphasizing the essential role of human expertise in validating and advancing knowledge. This conflict may lead to new standards for AI-generated research or increased scrutiny of AI claims within scientific fields.

OpenAI's withdrawal and acknowledgment of potential data contributions suggest a nascent awareness of the ethical implications of AI in research, potentially signaling a shift towards more collaborative or transparent engagement with human experts. The future of AI-driven scientific discovery will likely involve navigating these ethical and practical challenges, with a focus on fair attribution and the distinct contributions of both AI and human intelligence.