Newsgroups: comp.ai.games
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!gatech!howland.reston.ans.net!EU.net!sun4nl!hermes.bouw.tno.nl!usenet
From: sst@bouw.tno.nl (Tako Schotanus)
Subject: Re: Reply to Loebner
Message-ID: <1995May4.135215.25936@hermes.bouw.tno.nl>
Sender: usenet@hermes.bouw.tno.nl (USEnet Postmaster id)
Nntp-Posting-Host: takont
Organization: TNO Bouw
X-Newsreader: WinVN 0.92.6+
References: <3ncug7$igt@nunic.nu.edu> <3o31vj$s5q@theopolis.orl.mmc.com> <1995May2.014329.19470@hobbes.kzoo.edu> <D807qM.HD0@wave.scar.utoronto.ca>
Date: Thu, 4 May 1995 13:52:15 GMT
Lines: 47

In article <D807qM.HD0@wave.scar.utoronto.ca>, 93funkst@wave.scar.utoronto.ca (FUNK STEVEN LESLIE) says:
>
>Hi Folks,
>
>        As a Cog Sci student, graduated and on my way home today.  I can
>only say that they Turing test competition is nothing more then a
>publicity stunt.  I believe it was the Chinese Room thought experiment
>which showed that acting intellegent does not mean that you are
>intellegent.  So, in the end I can only think of one possible, real,
>advancement that could come out of this.  That would be a very nice
>natural language interface.  Apart from that it serves no real purpose.
>I can not see how any serious researcher might take this seriously. 

Oh? I stated this in an earlier post already, but I'll do it again (but
shorter :) :

1) The Chinese room is impossible because you'd need a metaphysical book
   with all the answers to all the questions you could ever ask (AND tell
   the answers in a different way all the time, after a while even giving
   angry replies because you keep on asking (the same) stupid questions).
2) *IF* such a book *DID* exists the whole setup (room, person, book)
   *CAN* be considered intelligent! Why? Because you wouldn't know how to
   tell the difference, so what *is* the difference? None.

NB: I still think natural language isn't the only and by far not the best
    method to test for intelligence. I think natural language is too
    much dependant on the experiences of being human. Too much context
    is taken for granted, context computers don't have. Of course you
    could try to give computers that context, but I think this would
    mean raising computer like children (though it might be done faster
    compared to children and once it was done you could just make copies
    of the core "memory"). If I'm not mistaken, such a project (give
    computers the context they need) is currently busy, I don't know
    its name though, I've forgotten it ;)

Tako

>
>Just my two cents worth.
>
>Steve

_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
_/ Tako Schotanus                TNO Building and Construction Research _/
_/ Phone : +31 15 842393 Fax : +31 15 122182  E-mail : sst@bouw.tno.nl  _/
_/ My employer is required,by Dutch law,to disagree with whatever I say _/
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
