Episode #105
Terence Tao – Kepler, Newton, and the true nature of mathematical discovery
“And what those stories teach us about how AI will revolutionize math”
March 20, 20261h 23m9 chapters
People
Topics
What this episode covers
“And what those stories teach us about how AI will revolutionize math”
Where to start
We need a semi-formal language for the way that scientists actually talk to each other
We need a semi-formal language for the way that scientists actually talk to each other
Start at 59:20
How would we know if there’s a new unifying concept within heaps of AI slop?
How would we know if there’s a new unifying concept within heaps of AI slop?
Start at 11:44
If AI solves a problem, can humans get understanding out of it?
If AI solves a problem, can humans get understanding out of it?
Start at 53:00
People and topics
People
Topics
Key takeaways
- Kepler was a high temperature LLM
- How would we know if there’s a new unifying concept within heaps of AI slop?
- The deductive overhang
- Selection bias in reported AI discoveries
- AI makes papers richer and broader, but not deeper