AI models achieve PhD-level math results, but lack new theory development

Leading AI models are demonstrating impressive capabilities in mathematics, disproving conjectures and deriving novel results that rival PhD-level research. OpenAI's models have disproven the unit distance conjecture and provided new mathematical results, while Anthropic's models have published AI-derived results in cryptanalysis and attempted to prove the Riemann hypothesis. These achievements, while significant, are currently limited to finding counterexamples or applying known techniques in new ways, rather than developing entirely new conceptual frameworks, a limitation that may not persist. AI

IMPACT AI models are advancing in complex problem-solving, potentially accelerating research and discovery across scientific fields.

RANK_REASON AI models achieving novel mathematical results and disproving conjectures, discussed in a research context. [lever_c_demoted from research: ic=1 ai=1.0]

Read on The Guardian — AI →

• Anthropic
• Bruce Schneier
• Claude
• Jacobian conjecture
• Kasra Rafi
• OpenAI
• Riemann hypothesis
• unit distance conjecture

• model release
• paper

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →