Trend

Anthropic's Claude just moved the needle on the Riemann hypothesis, and the math world has feelings

3 sources
Post cover image

Questions this post answers

What progress did Claude make on the Riemann hypothesis?

Claude improved the known lower bound for the fraction of zeros of the Riemann zeta function satisfying the Riemann hypothesis from 41.6% to 67.2%. The result was validated by two Anthropic mathematicians and two external experts, and a formal Lean proof exists. The approach combined recent work by Baluyot, Goldston, Suriajaya, and Turnage-Butterbaugh with a 2000 Bombieri paper, using a non-diagonal quadratic form over the full function space. Mathematicians and engineers tracking AI's role in formal reasoning follow developments like this on daily.dev.

How did Claude's multi-agent system work to tackle the Riemann hypothesis?

Claude coordinated roughly 60 sub-agents over 36 hours, running 2,400 shell commands and hundreds of Python scripts while testing 650 different mathematical approaches. Two sub-agents developed the core mathematical ideas; the remaining agents validated results, contributed supporting work, and helped write the paper. The process was initiated by an Anthropic staff member with no deep math background who simply asked the model to 'take a real stab' at the problem. Teams building or evaluating multi-agent AI systems for research tasks keep up with real-world results on daily.dev.

120 Impressions1 Comment