Terence Tao has published the slides for “Math 2.0,” a public lecture he gave at Caltech. He writes that the talk was originally planned under the title “Machine Assisted Proof” and was postponed from May to October; its content changed significantly in light of recent events.
Tao contrasts a mathematical culture centered on automated solving of open problems with a vision in which AI tools sustain the mathematical community and strengthen understanding of how mathematics advances real-world applications. He argues that solving problems is a proxy for broader scientific aims, including exploration and understanding. The lecture proposes using AI not only to produce proofs, but also to search for explanations, principles, and promising research directions.
The source records Tao’s argument and the lecture’s framing; it does not report an empirical evaluation of AI tools or demonstrate that a particular approach produces better mathematical outcomes.