OpenAI Solves Navier-Stokes Problem, Sparking Existential Crisis in Mathematics Community

OpenAI Cracks 200-Year-Old Mathematical Challenge

On September 8, OpenAI sent shockwaves through the mathematical world by announcing it had solved the Navier-Stokes problem, one of seven prestigious Millennium Prize Problems that have challenged mathematicians for decades1

1

2

2

. Using an internal large language model, an army of 10,000 autonomous AI agents tackled the problem in an 88-hour marathon that burned through millions of dollars' worth of computing power1

1

. The Navier-Stokes equations describe the physics of flowing fluids—from blood in a beating heart to swirling winds on meteorological maps—and the central question was whether these equations always provide a consistent picture or can "blow up" to produce physically impossible infinite values1

1

.

A Field Divided Over AI's Role

The OpenAI breakthrough has triggered what Brown University mathematician Javier Gómez-Serrano calls an "existential crisis" in mathematics1

1

. Speaking at Harvard University just three days after the announcement, with over 1,000 people tuned in on Zoom, Gómez-Serrano—who has partnered with Google DeepMind on AI applications—said he needed to update his slide depicting the five stages of grief mathematicians experience when confronting AI in mathematics1

1

. "By any standard, proving this kind of world-class theorem for a problem that has been open for 200-plus years—it's a genuine accomplishment," he acknowledged1

1

. Sergei Gukov of the California Institute of Technology described the moment more dramatically: "As a mathematician working on AI tools, I am knee-deep and fully embedded in this crisis. It's an earthquake"1

1

.

Controversy Over Research Ethics and Timing

The announcement sparked immediate controversy when NYU mathematician Tristan Buckmaster alleged that OpenAI may have scooped his work3

3

. Hours before OpenAI's public release, Buckmaster posted preliminary write-ups on the related Euler equations, work he had completed with collaborator Levent Alpöge of Anthropic using AI tools including OpenAI's Codex1

1

. Buckmaster implied OpenAI might have accessed his Codex logs and incorporated them into training data1

1

. OpenAI told The New York Times on September 10 that the timeline made this "categorically" impossible, claiming it had focused on Navier-Stokes after hearing rumors that Anthropic had solved two Millennium Prize Problems3

3

. UCLA professor Terence Tao captured the broader concern on Mastodon: "We have now seen that even the rumor of someone working on a problem can trigger a massive amount of AI-powered effort to flatten it"1

1

.

Corporate Power Reshapes Mathematical Research

The episode exposed deep tensions about corporate involvement in research. When Caltech undergraduates organized a "mathathon" with $2 million in computing credit donated by OpenAI and Anthropic, drawing over 1,000 sign-ups, the mathematical community pushed back hard1

1

. Two days after the Navier-Stokes announcement, more than 1,000 mathematicians signed an open letter objecting to what they viewed as corporate intrusion and exploitation of young mathematicians1

1

. "To put it bluntly, AI companies are engaging in research misconduct. This event exploits mathematicians at the very beginning of their careers," the letter stated1

1

. OpenAI subsequently withdrew from the event. On Friday, 25 Fields Medal winners released "A Severe Misalignment of AI in Mathematics," declaring that "the push by AI companies to solve mathematical problems as a benchmark is detrimental to the science of mathematics, and to the mathematical community"3

3

.

What This Means for Mathematics' Future

The Navier-Stokes solution represents a watershed moment for AI in mathematics. Just a year ago, AI advancements in math were limited to Mathematics Olympiad problems, with researchers often laughing at embarrassing errors2

2

. Now, with only five Millennium Prize Problems remaining—including the Birch and Swinnerton-Dyer Conjecture and the Hodge Conjecture—major AI companies are racing to solve them2

2

. OpenAI told the New York Times it has already made significant progress on another problem, though it didn't specify which one2

2

. The mathematical community now grapples with fundamental questions about what the next generation could lose: the idea of math as human understanding built through collaborative work and shared scrutiny1

1

. As Max Weinreich, a math professor at CUNY Baruch, observed: "We're seeing how swiftly an entire field can be ravaged by AI"3

3

. Watch for how AI's growing influence in mathematics reshapes not just problem-solving, but the very nature of mathematical research and discovery.