Google DeepMind's AI co-mathematician hits 48% on FrontierMath Tier 4

Google DeepMind introduced an AI co-mathematician — a Gemini-based agentic workbench that lets multiple AI agents manage parallel proof workstreams, track uncertainty, and review each other's work. It scored 48% on FrontierMath Tier 4 (a new high) and has already helped early users including Fields Medalist Timothy Gowers solve open problems.