Lex Fridman
What's left for mathematicians now that AI proves famous theorems
with Ankur Moitra, Steven Brunton and Alex Kontorovich, 3 h 34 min
Published
More info
- Guest
- Ankur Moitra, Steven Brunton and Alex Kontorovich
- Published
- Length
- 3 h 34 min
- Views when added
- 1,373,898
- Channel subscribers
- 5,070,000
Chapters from YouTube
- Lex calls this an emergency episode, after AI systems claimed proofs of some of math's hardest open problems.
- Steven Brunton studies fluids, Ankur Moitra theoretical computer science, and Alex Kontorovich number theory.
- Navier-Stokes describes how fluids move. Riemann is about prime numbers, and quasi-Riemann is a partial step toward it.
- Unique games is a guess about how closely hard problems can be solved. Lean is software that checks proofs.
- Math has long rewarded solving famous problems. The guests say the point was always understanding why.
What they say
Tap a line to open it, tap a bullet to play.
AI reached results experts thought were far off, and Lean helped check them.
Quasi-Riemann chained known moves, and such combinations can still reshape a field.
On Riemann, AI cleared a whole strip of zeros but not the full hypothesis.
On Navier-Stokes, AI showed the math can hit infinity, but engineers' daily work barely changes.
Unique games fell too, but P versus NP still looks very hard.
Why deep learning works is still unexplained, and that matters for safety.
The guests argue math is about understanding why, and theorems are only part of it.
Students have lost the old playbook, so the guests tell them to stay curious.
If you only watch 14 min
Best if you like math or want to know how far AI has really got.
Books, people and tools they mention
Books
People
Tools and companies
Written by AI from YouTube's captions and checked against them, so mistakes are possible. Captions don't name speakers. Report a mistake
Watch the whole video on YouTube (opens youtube.com)