Has AI destroyed mathematics?

- Teacher version
- Student version
- Copy link
- X
- Facebook
- LinkedIn

On 6th October 2026 Open AI released 722 mathematics papers on an unsuspecting mathematical community. These concerned 372 problem areas where mathematicians had been working on questions to do with topology, number theory, combinatorics, and even huge unsolved questions such as the Riemann hypothesis. These papers were generated by a hitherto unreleased frontier model being developed by Open AI. Is there any point in human beings doing research mathematics anymore?

Aim: This activity is an extended investigation into the scope and methods of mathematics with attention to ethics. Students will need to think about what it is that human mathematicians can do that generative AI cannot. They will also consider the role of understanding and explanation in the process of mathematical proof. The first two activities are suitable for the whole class and take 50 mins. There are additional activities for interested students that could take an additional 60 minutes.

Time: 50 plus an optional 60 minutes.

Commentary: The strategy here is that the students are confronted with real mathematicians not just journalists. Students will read a short article by a populariser of mathematics who is herself an accomplished mathematician. They will then watch short video clips of real mathematicians and computer programmers discussing the implications of recent dramatic events. At the end of the activity they will have a clearer idea of what mathematics used to be like before AI and the nature of the crisis it is facing. They will be required to reflect on their own view and to link to other areas where AI seems to be making big changes to the way in which we make knowledge.

Hannah Fry: Has AI destroyed mathematics? (30 mins)

Read the text below and answer the questions in groups. Be prepared to share with the whole class.

Writing in New Scientist mathematician Hannah Fry described the collective shock of the mathematical establishment when it woke up to find that Open AI had solved 372 of their most cherished problems. Prof Fry notes that far from being exhilarated that these problems were now solved and that the sum total of mathematical knowledge had been increased, mathematicians were dismayed not only that the ‘solutions’ offered almost impenetrable and led to no new methods, insights, or mathematical intuitions, but that, in many cases, mathematicians who had been working on them suddenly faced a situation where the reason for their existence was gone.

Hannah had been working on a particular set of equations for her PhD - the Navier-Stokes equations - way back in . These are super important in physics and engineering because they describe the motion of fluids, where they are a water droplet on a leaf, the movement of air around an aerofoil, or the rotation of whole galaxies. The big question was whether these equations could break apart and yield a situation known as a singularity where the fluid could whip itself up to infinite velocities. Hannah writes: “This was a such a big deal to the mathematical community that at the turn of the millennium, the Clay Mathematics Institute promised they’d give $1 million to whoever managed to solve it.”

The Navier-Stokes equations are basically Isaac Newton’s second law of motion: F=ma, but applied to fluids. AS Fry puts it: "They are, in my not entirely unbiased opinion, staggeringly beautiful, able to articulate everything from the drip of a water droplet on a leaf and the churn of an entire atmosphere to the gases swirling around in a galaxy in the sky.

"The thing is, these equations have a tricky little bit that pops out of the acceleration part, which contains a feedback loop within it. The faster fluid flows, the faster the gradient can get between two adjacent streams, and so the faster fluid flows. It’s a bit like the fluid version of putting a microphone next to a speaker. The feedback loop can get louder and louder until everything ends up breaking. Or, in the case of the Navier-Stokes equations, the speed of the flow heads off to infinity.

"To be clear: no one was ever worried about fluid dynamics tearing open a black hole when you stir your tea or anything. It was just that the weird behaviour of this little term left the possibility that the equations themselves would break.

“On the flip side, there’s another term in the equation – the viscosity – that acts as a brake and calms everything down, stopping the flow from getting completely out of control. So, the question was: does the brake always kick in?”

Various groups of mathematicians had been using AI to work on this problem, and by all accounts they were making progress. But apparently once OpenAI heard that others were getting close, in the words of Fry, “they threw everything at it: 10,000 AI agents, sending each other 3 million messages over 88 hours, costing millions of dollars in tokens, to find a single example where everything blows up."

Tristan Buckmaster from N.Y.U. continues the story. “At that time, you have to understand that the Navier-Stokes equation became this tech bro obsession,” Most of those tech bros didn’t even understand the equations, he said, but they’d heard mathematicians point to them as a place artificial intelligence might help. As Buckmaster put it, solving Navier-Stokes was becoming “the yardstick to say, ‘We’re better than that other AI company.’”

The Navier-Stokes equations did indeed fall, and an AI proved it, though at the same time unleashing a firestorm of accusations of plagiarism and the use of underhand methods. Fry laments this breakthrough in her article: “This should have been a glorious moment of scientific exposition, but instead, mathematicians have spent the last few weeks having a massive collective wobble.” Part of the reason is that nobody yet understands the answer. Terence Tao, arguably the world's most well-known mathematician, had devoted a lot of time to it, and is as puzzled as anyone. “A mathematical result is supposed to yield understanding rather than just being the exercise of a formal logic.” Yet this ‘solution’ is 166 pages of dense and terse text backed up by thousands of lines of formal code. There are no insights, intuitions, or beautiful methods to found here - no mathematical aesthetic. As Fry puts it: “they came for the glory without giving much back. So, we are left with a proof that almost no one wants to read, a $1 million prize that no one will want to collect, and no one who can stand at a blackboard and explain why this particular vortex makes everything break, when nothing else we tried ever did."

Fry observes that in the weeks since the release, mathematicians are having their own existential crisis about their role and purpose, like Lee Sedol did after AlphaGo, Gary Kasparov with Deep Blue, and probably every literature graduate, musician and artist since ChatGPT was launched.

For Fry, at its worst, AI is an engine of extraction, it extracts from the countless nameless individuals that have contributed, without their permission, to its training. “The real sting in the tail is that OpenAI set its sights on mathematics because the prestige, the respect and the glory was just another thing that could be mined. It didn’t need to be that way. They so easily could have been open and collaborative. But the moment when one of humanity’s greatest intellectual traditions became vulnerable, they chose conquest over stewardship.”

References:

Fry, H. (2026/14/October) Hannah Fry: How OPenAI has caused an existential crisis in mathematics. New Scientist.

Howlett, J. (2026/22/October) The inside story behind OpenAI’s million dollar math proof. Scientific American.

Discussion questions

Why are mathematicians shocked?
Why are mathematicians not celebrating the release of the new mathematical results as they would if human authors had produced them?
Why might mathematicians fear their future?
What is the Navier-Stokes problem?
Does Fry consider the actions of OpenAI to be collaborative and in the spirit of the tradition of mathematics as an area of knowledge?