He's the perfect profile for having the start of a P/NP proof: a full professor in computer science, but not really in the subfield of CS theory.