What’s Left for Mathematicians? | Daniel Litt

Daniel Litt discusses how he uses AI in mathematical research and why he changed his forecast about automated results. The conversation distinguishes technical proof work and literature coverage from choosing worthwhile questions and developing mathematical understanding. It examines checking machine-generated proofs, the guarantees and limits of Lean formalization, repeated discoveries by different models, and how research assessment might change. Further topics include publication incentives, competition between AI laboratories, teaching mathematics and questions about computational complexity. Statements about the future of mathematics are presented as the guest’s views.