Anthropic's Claude, mathematicians show 67.2% Riemann zeros on critical line
AI and human mathematicians just teamed up to make advances on two major puzzles about prime numbers.
Anthropic's AI, Claude, took on the famous Riemann zeta function and found that at least 67.2% of its tricky non-trivial zeros sit right where they're supposed to: way better than the previously known figure of 41.6%.
This matters for understanding how primes work (and even for keeping digital information safe).
Youness Lamzouri provides simpler proof
Anthropic's team verified Claude's results, but mathematician Youness Lamzouri also jumped in with a simpler proof on September 2, showing how humans and AI can really boost each other's work.
Meanwhile, Julia Stadlmann published a paper on August 31 reducing the bound from 246 to 240, before Axiom Math pushed it even further to 212.
It's a cool example of what happens when tech and math brains join forces!