OpenAI says its internal model solved over 100 long-standing math problems after just a month of training
decoder ·
OpenAI announced that its new internal model solved over 100 long-standing mathematical problems, including the Hodge conjecture and a second Millennium Prize Problem, after only one month of training. This rapid achievement has sparked debate within the scientific community, particularly regarding whether AI-generated solutions undermine the conceptual understanding central to mathematics. In response to criticism from mathematicians, OpenAI established an independent advisory group at the Institute for Advanced Study to guide how results are shared with researchers and the public. However, the company explicitly stated that this group will not have authority over the pace of its internal research progress. Fields Medalist Timothy Gowers joined the advisory board but declined to sign an open letter criticizing AI's impact on the field. Gowers argued that while problem-solving is valuable, the primary risk lies in the potential collapse of social structures supporting human mathematicians and their careers. He expressed concern that policymakers might view human experts as redundant if AI can solve theorems efficiently. The announcement highlights the tension between AI's computational capabilities and the traditional goals of mathematical inquiry.
decoder