Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!att!occrsh!uokmax!nova!krobt
From: krobt@nova.UUCP (Robert Klotz)
Newsgroups: comp.ai
Subject: free will
Message-ID: <17.2493730D@nova.UUCP>
Date: 12 Jun 89 06:42:37 GMT
Organization: Nova Research Institute
Lines: 47

In an article of <5 Jun 89 15:24:49 GMT>, falkg@vaxa.UCalgary.CA (Geoffrey Falk) writes:

 " Message-ID: <1478@cs-spool.calgary.UUCP>
 " 
 " 
 " Although it will always be impossible to strictly prove that any entity 
 " has a consciousness, it is evidently possible for consciousness to take 
 " place. The Turing test reduces to another moronic demonstration of the 
 " Other Minds problem.  The evidence for consciousness will be in the 
 " demonstration of behaviour that was not envisioned by the creators of the 
 " system.
 " 
hi,
 
   you were doing so good, and then your conclusion left us with something 
more vague than the turing test.  under your definition, most of my programs, 
in their early test runs, have consciousness.  i can certainly clearly 
demonstrate that their behaviour is in no way what i "expected" when i wrote 
them.

   i am begining to wonder if attempts at artifical intelligence are a 
waste of time.  i am not trying to say that it is unatainable, i am just 
wondering why we would even want it.  computers are fast becomming tools 
which are fulfilling all purposes which i originally thought would require 
"artifical intelligence".  expert systems are becomming very well developed, 
and soon there will be a good expert expert.  sence everyone has decided that 
expert technology isn't really true "intelligence" then that eliminates one 
big need for "ai".

   i have just written several pgms which allow users to query a database in 
plain and simple english.  i am sure most of you have seen such algorythms.  
everyone agrees that these manipulations are proformed without consciousness, 
cognition, or intelligence, even though the user always seems to get the 
answer to his/her questions.

   what i am trying to get around to is that when ever a problem arises which 
requires "ai", soon an alternate algorythm is found which is obvious to all 
is not truly "intelligent".  perhaps intelligence, artificial or not, can be 
defined as a significantly massive collection of algorythms useful in problem 
solving.

...robert
--  
------
{att!occrsh|dasys1|killer|uokmax}!mom!krobt | argue for your limitations
  or   --------------                       | and soon you will find that
    krobt%mom@uokmax.ecn.uoknor.edu         | they are yours.