Levesque works in knowledge representation, and all the sample questions hinge on some shallow inference from default knowledge. It's understandable to think yo

Levesque works in knowledge representation, and all the sample questions hinge on some shallow inference from default knowledge. It's understandable to think your own field of research is central -- and it was, in the 80s. What makes it especially productive to focus on now? That's what I'd like to have seen covered here. The Turing test seems a red herring, since afaik it's not a big part of research evaluation currently.