From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!sol.ctr.columbia.edu!destroyer!gumby!yale!cs.yale.edu!mcdermott-drew Mon May 25 14:05:04 EDT 1992
Article 5616 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!sol.ctr.columbia.edu!destroyer!gumby!yale!cs.yale.edu!mcdermott-drew
>From: mcdermott-drew@CS.YALE.EDU (Drew McDermott)
Subject: Turing Test
Message-ID: <1992May13.145303.29242@cs.yale.edu>
Keywords: Turing
Sender: news@cs.yale.edu (Usenet News)
Nntp-Posting-Host: atlantis.ai.cs.yale.edu
Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158
Date: Wed, 13 May 1992 14:53:03 GMT
Lines: 30

[I'm taking so long to respond to these messages that the originals
are unavailable by the time I get around to it.  Sorry if that's
causing any confusion.]

At some time in the remote past, Thomas Clarke (clarke@acme.ucf.edu)
wrote:

  Subject: Test for Intelligence

  Much discussion of late has dealt with the good and the bad of the Turing test.   
  It seems evident that some sort of test for intelligence needs to be  
  established so decisions can be made concerning the intelligence of artificial  
  systems.

I disagree strongly.  Artificial intelligence is concerned with how
much of thought is computation.  Its method is to take particular
skills and develop computational theories of them.  For any given
skill, one requires tests for evaluating how well the system is doing,
but there is no need to spell out these tests in advance, and
certainly no need for a single all-encompassing test.  Should vision
theorists test their models by seeing how well their programs can
converse with them about scenes?

I think it is amazing that so many people are obsessed with Turing's
little scenario.  Remember that he was trying to shake people up based
on no empirical data at all about what parts of thinking computers
could do duplicate.  If Turing were alive to see what computers can
actually do, I'm sure he would be the first to say that his Test is
completely irrelevant.



