Abstract

This investigation of the feasibility of artificial intelligence begins with a consideration of pertinent arguments by experts including Chomsky, Dreyfus, Fodor, Newell and Simon, Searle, and Stich. Representationalism, the view that cognition is primarily a matter of rule-governed manipulation of symbolic structures in the brain, is examined at length and ultimately found implausible. Exploration of an unorthodox approach to the design of intelligent machines generates a model of intentionality which reveals how mental states such as desires and beliefs can arise from purely physical foundations. Construction of such models is advocated as a promising research strategy in philosophy of mind and philosophy of language. Turning back to mainstream efforts to produce artificial intelligence, the tempting notion that Connectionism is a panacea for the difficulties encountered by more traditional techniques is debunked. Finally, it is argued that despite immense obstacles to the production of artificial human-like intelligence, judicious application of AI technology can produce very useful results