AI Just Hit a Milestone. It Cracked Hundreds of Math Problems. --- OpenAI released findings on more than 300. It may be 'the most significant moment in mathematical history.'

 

“Since you last thought about math, the people who think about math every day have been freaking out.

 

A month ago, when OpenAI announced that one of its powerful internal models had cracked a notoriously difficult Millennium Prize Problem, it looked like the most spectacular demonstration yet of artificial intelligence's progress in mathematics.

 

Then mathematicians revolted.

 

The backlash from the math world coincided with a global reckoning about Silicon Valley labs racing to build technology that could spiral beyond their control. On the same day that OpenAI published its landmark proof, a researcher quit Anthropic over safety concerns. Within hours, the AI breakthrough was overshadowed by fears that AI could destroy humanity.

 

Meanwhile, mathematicians were dealing with existential questions of their own.

 

But as they publicly debated the future of their ancient field, OpenAI's model kept churning out new proofs behind closed doors. This left the company with a problem for a bunch of humans to solve: what to do with all that AI-generated math.

 

On Tuesday night, OpenAI published hundreds of proofs for mathematicians to wrap their minds around.

 

An unprecedented number of answers that had eluded mathematicians for decades arrived in a single night. The deluge included AI solutions to more than 300 problems across number theory, algebraic geometry and other fields with names that could give anyone haunted by high-school math a migraine. Among the major findings in the collection of roughly 720 papers were advances related to three of the five remaining Millennium Prize Problems, including the Riemann hypothesis.

 

"If a human did this, it would be an instant Fields Medal, no questions asked," Alex Kontorovich, a Rutgers University distinguished math professor, wrote on X about the "quasi-Riemann hypothesis" proof.

 

The understated rollout of such monumental results marked a shift in OpenAI's approach to publishing scientific discoveries, part of an effort to win back mathematicians who have become allergic to AI hype in recent months. This past week, many of them reported feeling a sense of awe.

 

As stunned math experts scrolled through the results and began to absorb them, even a researcher inside Anthropic called the release "obviously the most significant moment in mathematical history."

 

It was also the latest head-spinning twist in a historic year for AI and math itself.

 

This year, OpenAI and Anthropic have turned complex mathematics into an unusual form of marketing for their smartest models, trumpeting solutions to problems that had stumped humans for nearly a century.

 

But what once seemed like a safe way to flaunt AI's capabilities started to backfire as Americans have become increasingly skeptical of AI.

 

The labs have also lost the trust of some people who have devoted their lives to math and once embraced AI tools as outrageously powerful calculators but now feel like collateral damage in a battle of trillion-dollar companies. Their rivalry is so fierce that OpenAI says it pointed its frontier model at the Millennium Prize Problems -- and poured millions of dollars in computing power into the effort -- only after hearing that Anthropic had already solved at least one.

 

The criticism intensified when more than two dozen winners of the prestigious Fields Medal published a furious open letter, warning that the industry's rush to solve math problems threatened to undermine the very purpose of math.

 

Under pressure from the scientific community that OpenAI says it's trying to empower, the company recently helped assemble an independent advisory group of leading mathematicians.

 

"There are difficult questions ahead about how AI can support mathematical understanding," OpenAI said last month. "We want mathematicians to be at the center of shaping the answers."

 

Last week, that advisory committee laid out guidelines for how AI labs should publish their findings based on feedback from hundreds of mathematicians.

 

OpenAI says it drew on those recommendations as it unleashed this avalanche of math.

 

Instead of dripping out findings one by one, OpenAI dropped them all at once.

 

The company released 372 groups of results and said the model attempted roughly 4,000 problems in total. It also provided summaries of the model's "chain of thought" and estimated the average result required only a fraction of the computing power that went into the Millennium Prize triumph.

 

Earlier this week, as rumors swirled that OpenAI would soon drop a barrage of proofs, mathematicians scrambled to publish their own AI-assisted work before they could be scooped.

 

OpenAI says the release reflected the math council's advice and concerns -- with one exception.

 

"We ask them to stop testing advanced mathematical problems on proprietary models," top mathematicians wrote last week.

 

OpenAI responded: "It is important to continue to evaluate our internal frontier models on mathematics and other sciences, so we can accelerate developing the tools to advance those fields."” [1]

 

1. EXCHANGE --- Science of Success: AI Just Hit a Milestone. It Cracked Hundreds of Math Problems. --- OpenAI released findings on more than 300. It may be 'the most significant moment in mathematical history.' Cohen, Ben.  Wall Street Journal, Eastern edition; New York, N.Y.. 10 Oct 2026: B4.