In the process we also confirmed something that was long suspected by many not to be true.
That AI could solve the RH within a few years. Tao just months ago was saying he only thought AI would be able to solve it but around 2050.
We’ve come a very long way from “LLMs won’t generalize” which was the predominant belief a few years ago.
So the proof reflects a lot more than just the rate of primes.