Terence Tao – Kepler, Newton, and the true nature of mathematical discovery
Summary
Terence Tao discusses AI’s role in science, using Kepler’s data-driven discovery of planetary laws as an analogy for how AI can generate empirical regularities. He argues that while AI lowers the cost of idea generation, the bottleneck has shifted to verification and evaluating partial progress. Tao highlights the complementary strengths of human depth and AI breadth, suggesting a future where AI maps broad scientific landscapes for humans to tackle complex, deep problems.
Topics discussed
Kepler's discovery of planetary motion laws
The shift from theory-first to data-driven science
AI abundance and the bottleneck of verification
Historical parallels: Newton, Darwin, and scientific progress
Signal extraction and the Jane Street puzzle solution
AI progress on the Polymath and Erdős problems
AI as a co-author and productivity multiplier
Formalizing proofs and the limits of AI reasoning
Statistical patterns in prime numbers and conjectures
Personal workflow, serendipity, and the future of math
Listen ad-free on Castria