Terence Tao compares the AI-mathematics wave to the early-20th-century foundational crisis
At the 2026 International Congress of Mathematicians, Terence Tao offered a cautiously optimistic reading of the AI-mathematics surge, comparing the moment to the foundational crisis of the early 20th century — a period of disorientation that ultimately reforged the discipline's foundations rather than destroying them.
The comparison is precise, not alarmist. The foundational crisis — the paradoxes that forced set theory, formal logic, and the reexamination of what a proof is — felt like a threat and became a rebuilding. Tao's framing places the AI wave in that lineage: unsettling, forcing the field to re-examine authorship, verification, and what mathematical understanding means, but plausibly generative rather than destructive.
Coming from Tao, the measured tone carries weight. He is among the most credible voices on where machine assistance genuinely helps and where it overreaches, and his willingness to name the disorientation while resisting doom is the signal most mathematicians will calibrate against. The discipline takes its temperature partly from him.
The practical reading is that mathematics is entering a period of methodological renegotiation, not obsolescence. The tools that produce proofs faster than they can be refereed force new norms — formalisation, provenance, division of labour between machine and human — and the field that emerges will have different foundations, as it did a century ago.
The Decoder — AI keeps cracking unsolved math problems, and mathematicians have mixed feelings → · Science News — An AI math breakthrough sparks calls for new guardrails →