AI seems less exciting because it has matured, and progress is harder and slower. (2)
AI has specialized, so it has broken off into many diffrent subdisciplines (e.g. AI of medicine).
Good formulations of the big problems of AI must make sense in all of the social and intelligence fields, and verification of the importance of the problems must come from those fields. (4)
".. for much of its history, artificial intelligence has worked in a virtual intellecual vacuum, isolated ... from most work in related fields."(5)
You used to be able to publish things in AI that were already known in other fields.
Starting in 1980, the field went less academic and more applied.
"Today, novel ideas must be novel in all fields at once, not just in artificial intelligence, and treatments of these ideas must satisfy standards of competence established in each of these fields." (6)