A YC startup is putting $1M behind five unsolved materials science problems for AI

Why it matters

Discovered Materials, a young San Francisco startup backed by Y Combinator, has launched a prize challenge asking AI models to solve major open problems in materials science. The company assembled an academic advisory group to define the challenges, which include explaining cuprate superconductivity, predicting silicon-silicon dioxide interfaces, mapping grain structure to thermal behavior, understanding how light alters material properties, and improving Density Functional Theory. The effort addresses a gap: while frontier models have advanced in mathematics and the life sciences, the complex physics of advanced materials has resisted them. Before announcing the challenges, the company let frontier lab agents work autonomously for days with internet access, and they made little headway. Co-founder Advaith Sridhar frames the problems as goalposts for measuring whether AI capability is climbing in this domain. The startup's main business is developing semiconductor materials with chip-industry partners, and Sridhar says he now expects useful materials sooner than the timeline he gave previously.