数学家称 OpenAI 不理解自己的 Navier-Stokes 证明
原标题：Mathematicians say OpenAI doesn't understand its own Navier-Stokes proof
AI 摘要
数学家质疑 OpenAI 对其 Navier-Stokes 问题潜在证明的理解，指出该证明与西班牙数学家 Diego Córdoba 和 Luis Martínez-Zoroa 的方法仅表面相似，实际路径不同。该证明用 Lean 形式化语言写成，自发布以来一直由 Clay Mathematics Institute 审查。OpenAI 还因在 GitHub 上发布约 400 项结果引发数学界不满，并被指在未公开模型上使用了他人工作。OpenAI 研究员 Noam Brown 表示，约 1 万个模型实例、88 小时、1300 亿 token 的多智能体部署对结果的贡献尚不清楚，他认为成功不到 10% 归功于多智能体设置，真正关键的是基础模型本身。
正文节选
Mathematicians say OpenAI doesn't understand its own Navier-Stokes proof In early September, the company presented a 166-page potential solution to the Navier-Stokes problem. OpenAI claimed the proof followed the approach of Spanish mathematicians Diego Córdoba and Luis Martínez-Zoroa. Several groups of experts now disagree, Scientific American reports. "When you go to the details, it is a different approach," says mathematician Gonzalo Cao-Labora of EPFL. The similarities are only surface-level