OpenAI 攻克黎曼猜想等 90 个数学难题

原标题：[AINews] Quasi-Riemann-Hypothesis: OpenAI publishes 722 math papers solving 90 of the top 500 open math problems; “the most significant moment” in >100 years of mathematics

92AI 研判分

核心综述

OpenAI 发布内部 Navier-Stokes 数学模型，宣称解决了包括黎曼猜想在内的前 500 大开放数学问题中的 90 个。该成果被 Anthropic 研究员评价为“百年来数学界最重大时刻”，并伴随博客、代码库及论文同步开源。这一突破标志着 AI 在纯数学推理领域达到全新高度，远超 Mistral 同期发布的 Large 4 模型关注度。

报道全文原始报道全文

Pour one out for Mistral, who shipped a decent Large 4 “Le Chonk” model on the new 3800 GB300 cluster funded by their recent Series D.

But they were overshadowed by more mathematics results from OpenAI’s internal Navier-Stokes math model - published as a blogpost, repo, and tweet. The best compliment comes from their Navier-Stokes competitor from Anthropic, who despite his personal issues with OpenAI, does not mince words: “It’s obviously the most significant moment in mathematical history.”

This bears some qualification, but most experts seem to agree that it solves many of the top 500 open problems in math.

In particular, Result 003, the Quasi-Riemann Hypothesis, is somewhere between a Fields Medal result and “the biggest result in number theory in 200 years”.

The most astonishing is the how - while Navier-Stokes was done in 88 hours and 10,000 agents, these solutions were 3 hours of ChatGPT Pro on average.