‘Immature playground boasting’: Mathematicians uneasy at OpenAI’s latest scalp

The recent news that OpenAI has cracked a Millennium Prize Problem, the Navier-Stokes problem, has sent shockwaves through the mathematics community. The achievement is significant not only because of its prestige and $1 million reward but also because it raises fundamental questions about the role of artificial intelligence in mathematical research.

For decades, mathematicians have been grappling with these problems, often working tirelessly for years to make breakthroughs. Now, AI can do what was once thought impossible: solve complex problems at an unprecedented pace. OpenAI’s achievement is all the more remarkable given the scale of its effort, deploying 10,000 agents on the problem and investing a near-trillion dollars.

While some have hailed this as a revolutionary moment in mathematics, others see it as a worrying trend. “It’s frustrating to see these big tech companies burning this fuel up just so that they can show off about how great their latest model is,” said Prof James Robinson of the University of Warwick. The concern is not just about the pace of progress but also about the authenticity and value of mathematical research.

With AI capable of solving problems at an unprecedented rate, mathematicians are left wondering what will be left for them to do. “We’re all just waiting to see what happens,” said Prof Colva Roney-Dougal of the University of St Andrews. “It’s coming so fast.” This is not just a question of job security but also raises fundamental questions about the nature of mathematical research.

The problem, as mathematicians are quick to point out, is that AI-assisted maths can create a culture of dependency. With problems being solved at the press of a button, what incentive do mathematicians have to work on them themselves? “It’s more like becoming an accountant,” said Prof David Silvester of the University of Manchester. “You’re auditing, you’re checking.” This is not just a problem for individual researchers but also has implications for the teaching of mathematics.

Universities are already grappling with how to adapt their curricula to this new reality. Undergraduates are being told not to use AI for certain problems, while lecturers are struggling to strike a balance between encouraging students to use technology and ensuring that they have a genuine understanding of mathematical concepts. “There’s no point in doing it because we can’t vouch for its authenticity,” said Silvester.

The OpenAI breakthrough also raises questions about the value of human contribution to mathematical research. While AI may be able to build on existing work, the real challenge lies in understanding and appreciating the beauty of mathematics itself. “You’ll still want to be able to understand the mathematics yourself,” said Prof Alexander Paseau of the University of Oxford. “And we’ll still enjoy it and appreciate its beauty: the sheer enjoyment and the beauty of a mathematical proof will always be there.”

As mathematicians grapple with this new reality, they are also grappling with a deeper question: what does it mean to do mathematics in the age of AI? Is it about solving problems at an unprecedented rate, or is it about understanding and appreciating the underlying concepts? The answer, as ever, lies in the math itself.

The Navier-Stokes problem may have been solved by OpenAI’s latest model, but the real challenge for mathematicians lies ahead. As they navigate this new landscape, they must confront the possibility that their work may no longer be valued on its own terms.