I think he’s trying to replace his grad students so that he can solve even more interesting problems.

Terence is fascinating in how different his thinking is even from most other mathematicians.

Which, admittedly, seems like a much harder problem.

A world in which AI churns out amazing proofs would be pretty radical though.