Can AI Prove It? Terence Tao on “Big Math” and Our Theoretical Future

Tao argues that mathematics’ next breakthroughs will come less from the lone genius than from “big math”: large, open collaborations of humans and machines, where progress is shared early and failure is cheap. He separates what AI does well (computation at scale, experimental math) from where it still fails, and asks how to keep answers that merely sound right tethered to proofs that are actually right. Along the way he covers the Green–Tao theorem, rubber-ducking, the “mental diet” AI demands, teaching with AI, and the strain of funding cuts on universities. Recorded October 2025 for the Berggruen Institute’s Futurology podcast.