Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!oitnews.harvard.edu!purdue!lerc.nasa.gov!magnus.acs.ohio-state.edu!csn!gw1.att.com!nntpa!bigtop!kimpek!jmarkus
From: jmarkus@kimpek.dr.att.com (John W. Markus)
Subject: Re: RFD: My Learning/Thinking Neural Network
Message-ID: <DD413J.E3M@bigtop.dr.att.com>
Sender: news@bigtop.dr.att.com (Netnews Administration Login)
Nntp-Posting-Host: kimpek
Organization: AT&T Bell Laboratories, Denver
References: <40bge2$6ib@kaleka.seanet.com>
Date: Thu, 10 Aug 1995 19:35:43 GMT
Lines: 22

<40bge2$6ib@kaleka.seanet.com> inficom@inficom.seanet.com writes:
>>   Joseph Bridgewater <bridgwtr@vanzandt.amd.com> writes:
>> ...  We have developed a learning, thinking neural network.  
>>  It thinks.

>What do you mean by "think" in this case? IMHO, no software program, no matter 
>how sophisticated, truly thinks. I think software can EMULATE thinking to an 
>almost total extent, but I still wouldn't call even a near-perfect emulation 
>"true" thinking.

I would say that if the computer could think about problems that it was not
currently being asked to solve, then I would consider that thinking.  For
example, one day you feed a problem into your computer, and it gives you
an answer.  The next day you go in and give it a different problem.  It
gives you an answer, and then says (or outputs in some way), "By the way,
I believe that the answer I gave you yesturday may not be right.  Here
are my latest calculations."  I would consider that to be "true" thinking.

John-
jmarkus@drmail.dr.att.com


