To ask a somewhat nuanced question, what do you feel the modern relevance of Lisp and Prolog are in AI? After writing a great deal about both language families, your first "go-to" language these days seems to be Python. Have major features for exploratory programming historically associated with Lisp been incorporated into dynamic/scripting languages such as Python, Ruby, and Lua?
I think that when I was in grad school, Lisp was unique in the power it brought for the type of exploratory programming that was necessary for AI. I think that today Lisp is still a great choice, but there are other choices that are also good---as you say, other languages have incorporated many (but not all) of the good parts of Lisp, so that today the choice of language can be made based on other factors.: for example, what language do you already know, do your friends know, etc.
There is a lot of content in an AI course, and I didn't think it made sense for an instructor to take a week or two out of the semester to teach Lisp, so we added Java and Python support. Java because it is widely-known, and Python because it is fairly widely-known and because, of all the languages I know, it happens to be closest to the pseudocode we invented in the book.
I never programmed at a serious level in Prolog, so I'll let other people comment on that.