Terence Tao: Math 2.0 [pdf]

Commenters debated Terence Tao’s “Math 2.0” argument that powerful AI finding proofs and solutions could “sterilize” fields, displace traditional career paths, and harm community incentives. math_dandy and Avicebron echoed worries that institutional reward structures and mathematician psychology will struggle to adapt, while theturtletalks emphasized loss of junior researchers’ routes to recognition and collaboration. samuelknight pointed to OpenAI’s solutions and an unexpected sub‑O(n log n) DFT result as evidence AI can upend assumptions, and gste argued that even if AI produces only proofs, it may still reveal new principles worth studying. Several commenters (qup, dekhn, squidmaster, chaosmanage) pushed back hard on Tao’s medical example, saying they would accept an effective AI‑discovered cancer treatment without a human‑understood mechanism.

Disagreement centered on trust, verification, and adaptation. dzink and others worried about hallucinations, verifiability, and intellectual conformity, while 19298857 feared loss of inspiration and over‑formalism. alecst and luckydata argued that models will improve and mathematicians should embrace new tools or risk irrelevance. vluft and others found some of Tao’s thought experiments unconvincing. Overall the split is between pragmatists who prioritize outcomes and tool adoption, and those who prioritize process, explainability, and reforming incentives before surrendering core parts of mathematical practice.