OpenAI AI Solves the Navier–Stokes Millennium Prize Problem

OpenAI has solved the Navier–Stokes Millennium Prize Problem by proving that fluid dynamics can develop singularities in finite time. The proof was produced by a massive multi-agent system powered by a next-generation AI model and verified through Lean formalization. This milestone demonstrates a significant leap in AI capability, moving beyond text generation into high-level mathematical reasoning and scientific breakthrough.

Key Points

• OpenAI's internal AI system proved that Navier–Stokes equations can develop a singularity in finite time, resolving a 90-year-old mathematical mystery.
• The solution was generated by a multi-agent system involving 10,000 concurrent agents and was verified using the Lean formalization language.
• The AI system also successfully disproved the regularity of unforced Euler equations during its research process.
• The breakthrough was achieved using a model more advanced than GPT-6 Astra, utilizing 300 billion output tokens across the entire project.
• OpenAI is releasing the findings to highlight AI's potential to empower scientific research rather than to claim the Millennium Prize.

Sentiment

The community is deeply mixed: commenters are impressed by the possibility of a major AI-assisted mathematical breakthrough, but the dominant tone is cautious, skeptical of OpenAI's framing, and concerned about attribution and scientific incentives. Hacker News does not simply reject the article's core claim, but it also does not broadly accept it at face value.

In Agreement

• If the proof and formalization withstand expert review, the announcement marks a major leap in AI-assisted mathematical research and undermines claims that current AI approaches cannot contribute to frontier mathematics.
• The speed of model improvement in mathematical reasoning appears striking, especially given claims that internal systems rapidly surpassed recently public capabilities.
• Even if human mathematicians and prior work shaped the result, the ability to coordinate, formalize, and extend research at this level would still be a major AI achievement.
• The direct engineering value of the Navier-Stokes theorem is less important than the broader implication that AI systems may now contribute to extremely hard intellectual work.

Opposed

• The claim should not be accepted until independent experts review both the mathematics and the Lean formalization, since formal systems and benchmark-like setups can still be misused or gamed.
• OpenAI may have benefited from or raced ahead of human mathematicians' unpublished ideas, raising serious questions about attribution and research ethics.
• The announcement may function partly as public relations, and some commenters distrust OpenAI's incentives and framing around frontier capability claims.
• AI-powered competition around open problems could make researchers less willing to share partial progress, damaging norms of open scientific collaboration.