Steven Strogatz on How AI Is Transforming Mathematics

Summary

In a WIRED interview, Cornell mathematician Steven Strogatz reflects on the rapid changes AI is bringing to pure mathematics. OpenAI says it used tens of thousands of agents to solve the 90-year-old Navier-Stokes existence and smoothness problem, which carries a $1 million prize, although the proposed solution still requires independent verification. The announcement also triggered a dispute over credit: mathematician Tristan Buckmaster says he and Anthropic researcher Levent Alpöge had posted solutions to three closely related, slightly easier problems shortly beforehand, while the OpenAI result built on a strategy developed by Diego Córdoba and Luis Martínez-Zoroa. Strogatz argues that the problem has little practical engineering significance and is being used mainly to demonstrate the capabilities of corporate AI labs competing for attention. He also points to Anthropic’s claim that Claude proved 29,500 small theorems while formalizing an existing proof of Fermat’s Last Theorem, and to OpenAI’s reported progress on 10 other long-standing problems. In his view, breakthrough mathematics will eventually require AI, while human mathematicians may remain responsible for explaining machine-generated proofs for a time. His collaborator Alex Townsend says ChatGPT helped with a decades-old numerical linear algebra problem whose workload and cost had previously seemed impractical. The interview presents both sides of the shift: AI could democratize participation and accelerate discovery, but it may also remove the motivation of being first to solve the hardest problems, weaken demand for human researchers, and make mathematicians feel displaced at their professional peak. Strogatz sees mathematics as an early test of what humanity may experience when machines produce knowledge that people can no longer fully understand.