> If his models could predict useful room temperature superconductors, it seems we would have found one already. A model that can check a given solution is not

> If his models could predict useful room temperature superconductors, it seems we would have found one already. A model that can check a given solution is not one that can quickly check the entire solution space. For a pure-CS analogy, this is the very definition of P vs NP.