Existential risk from superintelligent AI, naturally. Even if it turns out to not be a risk I'd sleep a lot easier at night with an ironclad mathematical proof

Existential risk from superintelligent AI, naturally. Even if it turns out to not be a risk I'd sleep a lot easier at night with an ironclad mathematical proof of why exactly it isn't on the presses. And if it is a risk, it's almost certainly the most important thing anyone could work on right now. I'm interpreting "can't fail" as "guaranteed to succeed one way or another", here.