Artificial intelligence has given the world a set of programming tools and practices, a set of experiments and models, a set of experiences, and a set of intuitions about what intelligence, representation, and computation are about and are not about. What of this legacy is still relevant and crucial to a cognitive science undergraduate? I should point out right away that the topic of this essay is what is often called "traditional" or "symbolic" or "good-old-fashioned" artificial intelligence. At least from the vantage point of La Jolla, such artificial intelligence has fallen on hard times of late--certainly it is out of style--and it is difficult to support the claim that artificial intelligence can provide scientifically viable models of human or animal cognition. In this essay I first review our departments computation curriculum, discuss whether artificial intelligence could ever form the basis for theories of cognition, and present a set of specific artificial intelligence topics that I believe should be covered in the computation part of a cognitive science curriculum.