OpenAI’s September 8 announcement of an AI-generated solution to the Navier-Stokes Millennium Prize Problem—produced via an internal frontier model and roughly 10,000 agents in 88 hours—has sharply raised trader expectations for further announcements. The company followed with an October 6 release of 722 manuscripts covering hundreds of open problems, including claimed progress toward other Millennium Prize challenges such as aspects of the Riemann hypothesis and Hodge conjecture, though several papers faced withdrawals or revisions after mathematician review. Competitive pressure from labs like Anthropic, combined with OpenAI’s emphasis on Lean formalizations and compute-intensive agent swarms, underscores accelerating capabilities in mathematical reasoning. Key near-term catalysts include ongoing independent verifications, potential additional model releases, and any follow-up company statements on unresolved problems before year-end deadlines.
如果 OpenAI 或 OpenAI 的官方代表在市场创建至指定日期 23:59 ET 之间宣布 OpenAI 已解决任何符合资格的千年大奖难题,本市场将判定为“是”。否则,本市场将以“否”结算。 符合条件的问题包括黎曼猜想、P 与 NP 问题、杨-米尔斯存在性与质量间隙问题、霍奇猜想,以及 Birch 和 Swinnerton-Dyer 猜想 (https://www.claymath.org/millennium-problems/)。关于纳维-斯托克斯存在性与光滑性问题的公告不符合条件。 符合条件的公告必须明确表示该问题已得到解决;关于部分结果或解决进展的公告不符合条件。公告必须将该解决方案表述为 OpenAI、其研究人员或其模型的成果,无论是独立完成,还是与外部研究人员合作完成。如果公告只是将功劳归于或祝贺外部研究人员,包括使用过 OpenAI 模型、算力或研究额度的研究人员,则不符合资格。无需 Clay Math Institute 或任何其他组织提供进一步确认。 本市场判定的主要依据将是 OpenAI 和/或其官方代表发布的官方信息;不过,也可以采用可信报道形成的共识。
机器翻译。如有争议,以原文为准。