AI materials startup offers $1 million prize, presents 5 challenging problems

AI materials discovery startup Discovered Materials has launched a $1 million prize pool for five open problems in materials science, offering $200,000 per problem. The challenges include a theory to explain superconductivity in cuprate materials, a numerical method to predict silicon–silicon dioxide interface structures, and a way to clarify the relationship between grain structure and thermal properties. TechCrunch reported that the company intends the five problems to serve as benchmarks for measuring AI's progress in materials science research. Detailed terms for the prizes have not been separately published on the company's official website. It matters as a signal that AI-for-science startups are trying to define the field's yardsticks and draw outside researchers into their orbit; watch whether the bounties attract credible solutions and how the benchmarks shape evaluation of AI materials models.