AI Solved a Millennium Math Puzzle. Then the Human Drama Began.

For centuries, cracking a grand mathematical conjecture was the ultimate solitary human triumph. But when OpenAI recently announced that its artificial intelligence had solved one of the most notoriously difficult puzzles in physics and math, the celebration was immediately hijacked by a modern corporate drama.

The puzzle in question is the Navier-Stokes existence and smoothness problem, one of the seven Millennium Prize Problems established by the Clay Mathematics Institute in 2000. It asks whether the complex equations governing fluid dynamics can mathematically break down and predict impossible scenarios, such as infinite velocity. OpenAI claims its unreleased, next-generation model proved that the full equations do indeed break down. The company added that it does not plan to claim the $1 million prize.

Instead of popping champagne, however, the math community is debating a bitter priority dispute. NYU mathematician Tristan Buckmaster and Levent Alpöge, an employee at OpenAI’s chief rival Anthropic, had spent nearly a year using public AI models to prove a simplified version of the exact same problem.

According to Buckmaster, when he reached out to OpenAI after hearing rumors of their parallel work, he was met with a harsh ultimatum. OpenAI allegedly told him he could either publish his simplified proof and watch OpenAI release the full solution the very next day, or he could collaborate on a joint paper—on the strict condition that Alpöge be dropped from the authorship due to his affiliation with a competitor.

The overlap between the two breakthroughs is striking. Both solutions rely on a specific mathematical approach pioneered by Diego Córdoba and Luis Martínez-Zoroa. While Brown University mathematician Javier Gómez-Serrano notes this was a known promising route, Buckmaster publicly questioned whether OpenAI’s models were trained on or accessed their private chat transcripts. OpenAI executives firmly deny accessing transcripts, though they admit their project was "inspired" by rumors of the NYU-Anthropic team's progress.

This controversy highlights a sobering paradigm shift in how science is done. If cracking the universe's biggest mysteries now requires frontier AI resources held by a few fiercely competitive tech giants, traditional open academic collaboration is in jeopardy.

Yet, there is a silver lining for human intellect. Experts have long noted that AI lacks "research taste"—the intuitive ability to select promising directions. If OpenAI's agents succeeded primarily because they followed the specific mathematical trail blazed by Buckmaster and Alpöge, it proves that human intuition remains the irreplaceable compass in the age of algorithmic discovery. The future of math isn't just about raw computing power; it's about who gets to hold the steering wheel.

Key Points

• OpenAI's AI successfully proved that the Navier-Stokes equations can break down, solving a Millennium Prize Problem.
• Independent researchers from NYU and Anthropic had already made significant progress on a simplified version of the problem.
• OpenAI is accused of issuing an ultimatum to exclude an Anthropic researcher from co-authorship due to corporate rivalry.
• The dispute raises ethical questions about AI research methods and the survival of traditional academic collaboration in the AI era.

Why It Matters

As frontier AI becomes essential for major scientific breakthroughs, the clash between closed corporate AI labs and open academic research threatens to rewrite the rules of scientific discovery and intellectual credit.

Sources:

• What OpenAI’s latest controversy tells us about the future of math — MIT Technology Review - AI