Anthropic's Claude Makes Major Progress on Riemann Hypothesis

Anthropic announced that its research version of Claude has made significant progress on the Riemann Hypothesis. In an autonomous test, Claude improved the lower bound of the proportion of zeros on the critical line from 41.6% to 67.2%. The research was conducted in Claude Code, generating 31 million output tokens, and after 650 failed attempts, it mobilized 60 sub-agents to execute 2,400 shell commands and write hundreds of Python scripts.