janeric@control.lth.se (Jan Eric Larsson) (02/03/90)
In article <4819@convex.convex.com> cash@convex.com (Peter Cash) writes: >I think that Turing must have meant for the test itself to *constitute* a >definition of intelligence. That is, he must have been thinking in terms >of strict behaviorism: it it acts smart, then it *is* smart. Turing explicitly stated that the question "Can a machine think?" was too complicated and vague to really deserve discussion. He thought the question of whether a machine could successfully play the imitation game was more useful to state, as it can be answered by designing some kind of machine or program. No definitions here of any kind, just a sound suggestion to avoid classical problems if you can! I also think it would be correct to assume that Turing would have thought almost all of the "Turing test - Thinking machines - Chinese room" debate on comp.ai meaningless; and as long as the involved parties notoriously avoid defining or explaining what they mean by "consciousness", "mind", and so on, (Actually, there are many philosophical theories about such things), I would tend to agree... Jan Eric Larsson JanEric@Control.LTH.Se +46 46 108795 Department of Automatic Control Lund Institute of Technology "We watched the thermocouples dance to the Box 118, S-221 00 LUND, Sweden spirited tunes of a high frequency band."