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

Discovered Materials, a San Francisco startup still only months old, has launched a $1 million challenge calling on AI models to solve five open problems in materials science. The company, which uses AI models to hunt for new substances used in building advanced chips, says each of the five problems carries a prize worth $200,000.

The startup sits in Y Combinator's spring 2026 batch. Materials science has no analogue to the Millennium Prize Problems, so Discovered Materials assembled an advisory group of academics at leading global universities to identify and spell out five open questions worth posing. The result is a mixed list: two theoretical puzzles and three requests for numerical methods that would improve how scientists model and design real materials.

The five challenges cover a theory explaining cuprate superconductivity, where certain materials carry current with zero resistance at ultra-low temperatures; a numerical method to predict structures at the interface between silicon and silicon dioxide, a pairing found in almost all semiconductors; a numerical method mapping grain structure to thermal properties for better heat-dissipation materials; a theory predicting why shining light on materials changes their behavior, particularly around superconductivity; and a numerical method to improve Density Functional Theory, a core technique for modeling atomic and molecular structures.

Advaith Sridhar, a co-founder of Discovered Materials, framed the prize as an answer to a measurement problem. "It's more difficult to specify problems clearly in materials science because it's not like math, where you have a very specific definition and can get to a yes or no," he said, adding that the team asked what the field's grand challenges are and how anyone would know AI is making progress on them. The framing reflects a gap: frontier models have cleared outstanding mathematics problems and pushed into the life sciences, but have yet to truly plumb the complex physics inside advanced materials.

Felipe Jornada, a Stanford professor who helped develop the challenges, said the motivation is explanatory. "Across materials science, phenomena we observe but cannot explain" remain, he said in a statement, and resolving them would improve understanding of physics while directly shaping how electronics are built, energy is stored and future materials are manufactured.

What to watch is whether a verifiable answer can emerge at all. Unlike mathematics, the prize problems invite partial progress and judgment calls about what counts as a solution, so the interesting test is how the judging standard holds up and whether submitted methods produce results that materials scientists actually adopt.

keyFacts:[