AI Crisis in Math: Navier-Stokes, Riemann Hypothesis & Complexity Theory | Lex Fridman Podcast #503
AI has just produced results on famously hard problems, and three mathematicians argue the right response is to refocus on understanding, good questions and human community rather than treating theorem-proving as the point of mathematics.
Watch on YouTube ↗Charlie's summary
- ▶ 3:42
Ankur feels pulled two ways: AI is a superb teacher, but the old blueprint for students has been thrown out
Ankur says the tools let him learn other fields as if he had “a world expert who's there to teach you about things in your own language.” But the playbook for advising students and postdocs, which was to take a problem people care about and improve it, no longer works. He feels intellectually lonely, because patient, humble questions get lost in a flood of AI results.
- ▶ 19:53
The doorman fallacy: mathematicians do more than prove theorems, so cheap problem-solving raises the value of everything else
Ankur borrows the doorman fallacy: reducing a job to its most visible task, like holding a door, misses the rest of the value. Proving theorems is the visible part of mathematics. A mathematician is “someone who thinks deeply,” which also means building theories and making conceptual connections. When well-defined problem-solving becomes cheap, the orthogonal work becomes “all that much more valuable.” Ankur says this is the moment to recenter on values, not hit the panic button.
- ▶ 11:00
Alex's solved axiom: even if AI could do anything humans plus AI can, pure math survives in human heads
Alex defines the solved axiom as: any problem solvable by a human working with AI could be solved by AI alone. Chess shows the danger, since centaur play collapsed into accepting the engine's top move. Even so, Alex argues pure mathematics is valuable because it lives “in the heads of human beings,” not as ones and zeros in data centers. He doesn't actually believe the axiom holds today. Ankur thinks it is at best half true.
- ▶ 1:21:06
Mathematics is the why, not the fact: knowing a result is true without understanding it adds nothing
The panel retells a play in which von Neumann asks God if the Riemann hypothesis is true. God says yes, but Neumann is unsatisfied until he learns why. Alex says math “is not the fact, it's the why,” and a true statement he doesn't understand has “contributed zero to my knowledge.” Ankur adds that mathematics is the fight with a problem: “when you fight with a good problem, it fights back at you,” and you learn from that. Proofs should come with the story of what was learned.
Give your email to unlock the rest of the key points — for every episode.
Do more with Charlie
Free — sign in with your email, no password.