AI's Disruption of Mathematical Research: A Crisis in Foundations

This content examines the rapid acceleration of AI capabilities in solving and disproving long-standing mathematical conjectures, highlighting a shift from theoretical possibility to practical disruption. It details recent breakthroughs by major models that have challenged established problems, raising concerns about the future of mathematical practice and verification.

- Historical Context of AI in Math
- Early successes in competition mathematics
- The turning point with Erdős conjecture disproof
- Recent Major Conjecture Disprovals
- Levent Alpoge and Fabel disproving the Jacobian conjecture
- Dimitri Rybin using GPT-5.6 for the dense Garg-Gommans conjecture
- Institutional Model Capabilities
- OpenAI's internal model solving 10 open problems with formal proofs
- Anthropic's progress on the Riemann hypothesis via user collaboration
- Implications for the Mathematical Community
- The Leiden Declaration and calls for guardrails
- Terence Tao's warnings on foundational crises

This overview provides context for understanding how AI is reshaping mathematical discovery and what it means for researchers and students navigating this new landscape.