Keynote Address

Saturday, April 4, 2009 at 4:00pm in Psychology 100

AI Claims versus Cog-Sci Aims

Douglas Hofstadter 
Center for Research on Concepts and Cognition
Indiana University

      AI started out in the late 1940's as a profound scientific endeavor -- to create thinking machines.  Over its several decades of existence, however, it has metamorphosed (and perhaps predictably so, had there been a sufficiently foresightful predictor, back in those early days) into a random hodge-podge of military, industrial, and technological goals.  From initially being a serious and humble effort to lay bare the human mind's complex and subtle secrets, AI has rapidly degraded into an largely amoral dog-eat-dog fight for funding in which loud and preposterous claims of sudden giant revolutionary steps forward compete against realistic descriptions of modest and gradual progress consisting of small insights -- and guess which type of self-description tends to bring home the bacon in this deplorable tooth-and-nail battle for money?
      In my talk, I will talk about some of the louder voices in this bizarrely ruleless bazaar, and will briefly describe some of the projects that, though they may have grabbed great attention, have delivered essentially nothing of use, and I will also talk about some projects that have delivered some perhaps useful or interesting technology but that have nothing to do with understanding the human mind, and that perhaps never claimed to.  Finally I will talk about the goals of cognitive science and of cognitive scientists, and I will speculate about where we are going, and whether we really want to get there or not.