Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: Notesfiles; site uicsl.UUCP
Path: utzoo!watmath!clyde!burl!mgnetp!ihnp4!inuxc!pur-ee!uiucdcs!uicsl!rmooney
From: rmooney@uicsl.UUCP
Newsgroups: net.philosophy
Subject: Re: Can a Computer Think - (nf)
Message-ID: <17200001@uicsl.UUCP>
Date: Wed, 16-May-84 16:57:00 EDT
Article-I.D.: uicsl.17200001
Posted: Wed May 16 16:57:00 1984
Date-Received: Fri, 18-May-84 00:40:53 EDT
Lines: 25
Nf-ID: #N:uicsl:17200001:000:1375
Nf-From: uicsl!rmooney    May 16 15:57:00 1984

#N:uicsl:17200001:000:1375
uicsl!rmooney    May 16 15:57:00 1984


   My usual answer to this question is that unless one believes in mind body
dualism (which I regard as needless mysticism which is quickly ruled out by
Occam's razor since it requires an increased ontology and resolving the 
historically difficult problem of mind-body interaction), then the answer to 
whether computers could ever be said to think must be yes.  Since a computer 
is effectively a Turing machine it can perform any computational task.  If
nothing else it could numerically simulate every neuron in a human brain.

    The question about it only "simulating" such activity seems to be a
a meaningless remark since a perfect simulation must be, for all practical
purposes, equivalent to the real thing.  The only perfect map is a
physical duplicate of the territory. (see Mind's Eye by Hofstadter & Dennett
, specifically essay #5 on the Turing test for more on this.)

    As for the question whether any computer "thinks" now, I would not commit
myself since I believe "thinking" ability is a continuum and present software
is rather low on this continuum; however, I believe the Turing test would be
a sufficient empirical test since everyone but solipsists grant that other
people "think" and their only knowledge of them is their input/output behavior
In fact Turing originally proposed the test as a substitute for the question
"Can a computer think?"