Terence Tao on the OpenAI's Navier-Stokes solution

Terence Tao on what OpenAI's answer to the Navier-Stokes problem says about AI & science: "there is now an increasingly realistic scenario in which a primarily AI-generated solution to the problem ... contaminates the problem as a source of further advances".

Why? With maths problems like this, the solution is not where the value lies: "Computational fluid dynamics is already a mature subject", so solving the problem "would not radically transform the way we ... model weather prediction or climate change".

The value lies instead in the effort to find a solution, which "have historically led to fundamental insights and influential theorems", as exploring "alternate routes ... that are superficially "dead ends"... end up being highly instructive in the nature of their failure".

It's the journey, not the destination, which is of value. Today, however, we have a solution from an "AI company running the harness [which] keeps the process ... almost completely out of public view ... there would be almost no value added to mathematics as a consequence... Prematurely solving the problem by purely AI-powered methods - particularly without full transparency into the solution process - can contaminate this process to the point where it actually becomes a net negative for the progress of mathematics as a whole."

For those of a mathematical frame of mind he sets out an alternative approach of AI-assisted mathematicians working together, arriving at a solution and learning a lot along the way.

See also: https://www.nytimes.com/2026/09/22/opinion/artificial-intelligence-ai-danger.html