Five Grand Challenges in Material Science for AI to Solve

Over the last few months, we at Discovered Materials have worked with 10 leading professors to identify the hardest open problems in the field. They are decades old, extensively studied, verifiable and yet unsolved. We believe a breakthrough in any would unlock major advances in chip manufacturing, superconductivity, clean energy, and other domains at once.

We'd love to hear from the HN community: 1. If you're building reasoning models, research agents or scientific AI, point them at these challenges and tell us what happens. We’re interested in even a partial result or an interesting failure. 2. If you're a physicist or chemist, tell us whether we picked the right five, and which problem you'd add.