Newsgroups: comp.ai.alife,comp.ai.philosophy,comp.ai,alt.consciousness
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!agate!sunsite.doc.ic.ac.uk!lyra.csx.cam.ac.uk!warwick!slxsys!pipex!pipex!uunet!psinntp!scylla!daryl
From: daryl@oracorp.com (Daryl McCullough)
Subject: Re: Thought Question
Message-ID: <1995Jan17.172724.9815@oracorp.com>
Organization: Odyssey Research Associates, Inc.
Date: Tue, 17 Jan 1995 17:27:24 GMT
Lines: 74
Xref: glinda.oz.cs.cmu.edu comp.ai.alife:1818 comp.ai.philosophy:24726 comp.ai:26561

prem@ix.netcom.com (Prem Sobel) writes:

>All who have taken the time and care to learn about consciousness, to
>study it through the means appropriate to it have discovered, or rather
>verified, that there are sources of knowing beyond the physical.
>Creativity itself is one example. Love is another example. Love is
>not a pattern of bits encoded in the neural network of the brain.
>Love is not a determinism of automata theory. Is feeling how another 
>is feeling, through empathy, through identity, even through physical
>observation and reasoning by analogs of our own experience still
>comes to down to an inner state of feeling. The feeling of well being,
>anger, humor, the joy in learning, and so much more is beyond the
>states of bits encoded in the brain.

Prem, I thought that the issue being discussed was not whether a
computer program can experience love, consciousness, etc., but about
whether it is possible to build a computer program that reproduces the
outward behavior of a human being. Earlier, you claimed that it was
not, saying:

>You have got to be kidding !!!! While it is possible to build a very
>accurate servo mechanism, perhaps with a computer controlling it,
>there is no way for that machine design to implement something that
>can anticipate and respond to any circumstance. Only living and
>especially thinking conscious animals manage to do this very well.
>Those that fail don't survive. Consciousness is of survival benifit
>to say the least.

One doesn't need to show how a computer can be capable of love to show
that it is capable of responding to all possible circumstances as well
as a human. While there may be speed and memory limitations to
man-made computers due to physical reasons, there is no reason in
principle that a computer cannot reproduce human behavior to arbitrary
accuracy. The reasons are the following:

(1) Because of limitations to the sensitivity of human senses, there
are only a finite number of distinguishable sensory experiences that
can occur in a single second. It's a huge number, but it is still
finite. There might be an infinite number of possible sights, sounds
and smells that we could be exposed to, but there can be only finitely
many *distinguishable* sights, sounds and smells.

(2) Because our life spans are finite, there are only a finite number
of seconds we can ever experience.

(3) Putting 1 and 2 together, we come to the conclusion that there are
only a finite number of possible total lifetime experiences that we
can ever have.

(4) Because of limitations in our ability to control our muscles, our
lips, our eyes, etc. there are only a finite number of different
actions that we can possibly choose to do during a single
second. There might be an infinite number of possible actions, but
there are only finitely many that are voluntarily distinguishable: We
can't choose to move our finger one trillionth of an inch, because we
just don't have that kind of control.

(5) Putting 2 and 4 together, we come to the conclusion that there are
only finitely many different patterns of action that we can choose to
do during a single lifetime.

(6) Putting 3 and 5 together, we come to the conclusion that the
behavior of a human being during a finite lifetime can be described as
a relationship between a finite number of possible input patterns and
a finite number of possible output patterns. Since all relations
between finite sets are computable, it follows that each human's
behavior pattern is computable.

So, regardless of whether consiousness is computable, we know that, in
principle, a human's external behavior is.

Daryl McCullough
ORA Corp.
Ithaca, NY
