X-Message-Number: 410
From att!uunet.UU.NET!ghsvax!hal Wed Aug 28 10:53:33 PDT 1991
Date: Wed, 28 Aug 91 10:53:33 PDT
From: ghsvax! (Hal Finney)
To: 
Subject: Uploading in principle?

I was very impressed by Alan Lovejoy's essay on whether uploading is
possible in principle.  He concludes that consciousness is essentially
a symbolic process and that therefore uploading should theoretically
be possible.

Philosopher John Searle has for many years argued exactly the
opposite.  He claims that no purely symbolic activity can ever be
conscious.  His arguments are directed against Artificial
Intelligence, but I think they would apply to uploaded intelligence as
well.

Searle calls his main argument the "Chinese room".  I will take some
liberties with it to consider the uploading case.  Imagine that a
Chinese speaking woman has been uploaded; a computer is simulating her
mind at some level, perhaps the neuronal level.  This computer is able,
among other things, to read and write Chinese in an intelligent
fashion just like the uploaded person had been able to.  However, it
is actually a computer running a program.

Now, Searle argues that anything the computer can do, he should be
able to do.  In particular, he should be able to hand-simulate the
actions of the computer - keeping track of the values in various
program variables, updating them according to the rules of the computer
program which might appear as a list of instructions he is to follow.
By following these instructions, he is able to take in, say, questions
written in Chinese, and eventually to produce answers written in
Chinese, answers which would appear to come from the uploaded Chinese
woman.  (Of course, he would be much slower than the computer, but
eventually the answers would appear.)

The question is whether there is a conscious Chinese woman in this
situation.  Searle argues that the answer is no.  He doesn't know any
Chinese at all, yet by mechanically following this list of
instructions, he is able to take Chinese questions and provide Chinese
answers.  Yet he certainly does not thereby become a Chinese woman.
He still doesn't know a word of Chinese.

Searle argues that this means that the kind of mechanical symbol
manipulation that a computer does when it runs a program isn't enough
for consciousness.  It can "simulate" consciousness, perhaps, but
there is no actual consciousness involved.  It's all a game played
with abstract symbols.

[ One might ask whether or not normal human consciousness is a "mere"
  simulation played out in our wetware.  If so, that may expand the range
  of possible scenarios for reanimation from cryonic suspension. - KQB ]

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=410