2 ms·

> Yet neither does Alpha go. Finding solutions really is an NP-class problem, while proving them is much easier. That's not really true. Proving that an alread

> Yet neither does Alpha go. Finding solutions really is an NP-class problem, while proving them is much easier. That's not really true. Proving that an already found solution is correct might be easy for certain solutions (that's the whole idea of proof certificates). But finding a proof for an arbitrary theorem is certainly not easy. And that is what mathematicians do. Here, finding the proof IS finding the solution.

MayeulC 7y ago
Of course, but when you have found a proof, verifying that it is correct is fairly trivial (if tedious by human standards, it's only verifying that each statement is correct). I meant proving in the sense of "checking that the solution works". In a mathematical problem, the solution would be a tentative proof for a theorem. Proof which needs to be verified. I should probably have used "verify" instead of "prove" in this context.