Discovered Materials, a Y Combinator startup, announced a $1M prize pool for AI models to solve five open materials science problems. The startup uses frontier AI models to find new materials for advanced chips and partners with chip giants. An advisory group of academics from leading universities helped define the five open problems in the field.
Co-founder Advaith Sridhar stated that materials science problems are harder to specify than mathematics because they lack clear definitions. Professor Felipe Jornada noted that solving these problems would improve physics understanding and shape electronics and energy storage manufacturing. Frontier lab models were tested on these problems but struggled to make significant progress autonomously.
Sridhar emphasized the goal is to use the challenges as goalposts to track AI progress in the materials domain. The company aims to find industrially useful materials within a year, accelerating its initial timeline. This creates a concrete benchmark for developers to test their models against real-world constraints.
