Accelerating the Queen of Sciences

Mathematics is the foundation on which all hard science is built, and progress has depended on the rare combination of human creativity, intuition, and rigor that only the best mathematicians possess. This project asks whether AI can begin to develop those same qualities: not just solving known problems, but learning to reason, wonder, conjecture, and discover the way a mathematician does. By training AI systems within formal proof environments where claims are automatically verified, this team aims to build models that are provably right, and that can serve as true research partners across the quantitative sciences. Success here would not just advance mathematics; it would unlock progress across every field that depends on it.