AI can now do math. But can it ask good questions? - Ken Ono

Ken Ono is a number theorist at the University of Virginia and advisor to the NSA. Last year AI helped him discover new formulas that detect prime numbers, patterns he admits he wouldn't have thought to look for. Ken suggests that when mathematicians make breakthroughs, they're hallucinating too. They’re proposing patterns beyond established results, but with an important difference: they've learned to tell promising hallucinations from dead ends. AI can now tackle math problems that once seemed far beyond its reach. But it still can’t do the one thing that matters most: ask good questions. We also talk about why the Riemann hypothesis might get solved by finding something close enough, how he deliberately teaches false proofs to train students, and how AI tools are beginning to support rather than replace mathematicians. -- Episode links -- Transcript: https://epochai.substack.com/p/ai-can-now-do-math-but-can-it-ask -- Timestamps -- 00:00:00 – Why predicting problem difficulty is so hard 00:00:27 – How mathematicians and AI both “hallucinate” 00:02:15 – AI as a copilot for mathematical discovery 00:04:00 – How AI helped reveal new formulas for primes 00:09:02 – The promise and peril of AI 00:15:22 – What makes a great mathematical question 00:17:54 – Teaching by making deliberate mistakes 00:21:22 – Will AI reshape the world like the Industrial Revolution? Special thanks to interviewer Daria Ivanova: https://youtube.com/@inQEDible