Alberts, Gerard <g.alberts@uva.nl> writes:
The crucial book is Hubert L. Dreyfus, What computers still can't do. A critique of artificial reason (MIT, 1972. 1992). The original What computers can't do, is basically a philosophical argumentation. The revised edition has an ample introduction offering a most readible historical view of the debate as it evolved.
For more historical context on Dreyfus's role as AI critic, I like a recent retrospective paper he wrote. This one focuses on the constructive flipside of his famous critique of AI, his less-famous efforts to cultivate, through collaborations with AI researchers, a different kind of AI that would be responsive to his critiques. He doesn't think the efforts were successful, but in my opinion it's an interesting postmortem of that line of work: Hubert L. Dreyfus (2007). Why Heideggerian AI failed and how fixing it would require making it more Heideggerian. Artificial Intelligence 171(18): 1137-1160. http://doi.org/10.1016/j.artint.2007.10.012 More solidly on the original ethics point, one of the Dreyfus-influenced AI researchers, Phil Agre, has a nice article on how to reconceive of AI research as a "critical technical practice" (admittedly not quite the same way of framing things as the "ethics" frame, but related): Philip E. Agre (1997). Toward a Critical Technical Practice: Lessons Learned in Trying to Reform AI. In: Bridging the Great Divide: Social Science, Technical Systems, and Cooperative Work, Erlbaum, 1997. http://polaris.gseis.ucla.edu/pagre/critical.html -Mark -- Mark J. Nelson The MetaMakers Institute Falmouth University http://www.kmjn.org