AI Solves 80-Year-Old Math Conjectures; Terence Tao Warns of 'Proof Surplus' Era
新智元 · wechat · 2026-08-07
AI has recently solved multiple long-standing mathematical problems, from disproving the 80-year-old Erdős unit distance conjecture to finding counterexamples to the Jacobian conjecture. However, top mathematicians are divided on the implications.
- AI Capability Leaps: Internal models from OpenAI and Anthropic have made breakthroughs in discrete geometry and high-dimensional sphere packing, with public models successfully replicating some results. The independent First Proof evaluation shows AI can now produce publishable-level solutions to novel research problems.
- Divided Opinions: Mathematicians like Fields Medalist Deng Yu view AI as an assistant that will handle technical details rather than replace researchers. However, others like Tim Gowers worry that AI-generated proofs lacking human comprehension will lead to an explosion of undigested academic literature.
- Terence Tao's Warning: Tao highlighted that mathematics is entering an era of 'proof surplus.' While AI accelerates proof generation and verification, processes like interpretation and canonization remain deeply human. He warned of an impending 'indigestion' crisis where major results are proven correct but remain incomprehensible to anyone, urging the community to refocus on the core values of mathematical research.
Related event: AI Breaks Math Conjectures, Tao Warns of Proof Glut(13 posts)→
More from AGI Musings
- Economist argues model providers face hard limits on vertical integration into white-collar sectors — sethlazar · 2026-10-16
- METR outlines three blind spots in public knowledge of AI lab internal agent risks — CFGeek · 2026-10-16
- LMR-BENCH at EMNLP 2025 tests whether LLM agents can reproduce real LLM research — ZhiyuChen4 · 2026-10-16
- Deep Research becoming useful to experts will signal RSI arrival, researcher predicts — ZeeshanZiaML · 2026-10-16
- Physicist pushes back on ML community's acceptance that models may be beyond understanding — tszzl · 2026-10-16
- Ben Affleck calls AI 'just matrix multiplication'; Grady Booch echoes the skepticism — Grady_Booch · 2026-10-16