Academic AI is very much about tomfoolery such as Eliza & Perry, ie simple text parsers that picked up individual words and phrase constructs and replied to them with no intelligence whatsoever.
A lot of the time, i'm assuming you'll be working with languages such as Prolog and Lisp (I hated them) and I dont think they are particularly relevent to games programming.
However it's not all gloom and doom
Apart fom the aforementioned linguistics based gubbins, your course will cover such things as path finding (aka the travelling salesman) & tree searching which are heavily used in games design. If your lucky they'll also give you info on A-Life imho a far more useful, and probably successful approach to AI, look at stuff like the opengl paper planes / flock of birds.
Finally you may get to cover such esoterica such as vision processing.
After much rambling, i'd suggest looking at the ground up approach that produces complex emergent behaviours from simple approaches rather than emulation of higher order mental functions. Neural nets are probably of more value than Knowledge Bases