The statement that he's critiquing does reflect the wider-spread, overly simplistic view of AI. Contrary to hype, recent events represent only partial development/peeling of the top layer from the AI onion, which has more known unknowns and unknown unknowns than known knowns.