X-Message-Number: 8587
Date: Fri, 12 Sep 1997 10:48:00 -0400
From: "John P. Pietrzak" <>
Subject: Re: Intelligence
References: <>

Michael Riskin wrote:
> I am new to these electronic discussions of intelligence, human or
> artificial...which by the way makes me wonder why it is assumed that
> human intelligence is not also standardly referred to as " artificial"
> also.

I myself have probably spent far too much time on electronic discussions
of intelligence (by now, most readers of Cryonet would agree with this,
I'm sure).  But anyway, the term "Artificial Intelligence" as it is
used today originated directly from the attempt begun in the early
60's to construct, using digital computers, automatons which were able
to perform tasks considered to require "intelligence".  Generally,
the idea was to emphasize the man-made aspect of the automaton, I think.

> So...i have assumed from these discussions , that computer
> intelligence assumes the need to go through a specific series of
> analyses to come to a conclusion.  [...]

Certainly, there are well-defined constraints upon what a computer
can do.  We haven't yet defined very well exactly what humans can do,
but I would assume that there are also constraints on us; it may not
be easy to tell how much, but finding the value of 12 x 12 does take
some amount of effort.

> Now ....chess is a finite activity and unless kasparove can figure
> out a way to cheat and not get caught, he will ultimately have to
> lose.  [...]

Well, that's not quite true, in that although it is a finite activity,
there is as yet not enough storage or processing ability in either the
human or the machine to entirely circumscribe the set of possible board
configurations.  Therefore, for now, neither one can tell exactly what
is the "best" move in a given situation; they have to try and figure
out which moves are better than others through an imperfect predictive
process.  So, for now at least, Kasparov can still win.  (Besides, it's
no fun if you cheat.)

> Well...can intelligence be measured by the degree of subjective
> pleasure it can generate for its owner?

Hmm, skip the Turing Test, try the Orgy test?  Now *that's* a theory
that bears running some trials. ;)


John

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8587