From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!europa.asd.contel.com!uunet!orca!javelin.sim.es.com!biesel Wed Apr 22 12:04:09 EDT 1992
Article 5156 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!europa.asd.contel.com!uunet!orca!javelin.sim.es.com!biesel
>From: biesel@javelin.sim.es.com (Heiner Biesel)
Newsgroups: comp.ai.philosophy
Subject: Intelligence, awareness, and esthetics
Message-ID: <1992Apr20.191345.27706@javelin.sim.es.com>
Date: 20 Apr 92 19:13:45 GMT
Organization: Evans & Sutherland Computer Corporation
Lines: 54

Much of the discussion lately has focused upon the assertion that 
intelligence, as detected by the Turing test, necessarily implies
awareness, in the sense in which human beings are aware. This
assertion, it appears to me, is based on a basic confusion between
two distinct human capabilities for which the Turing test is
not a particularly good discriminator.

The test, as usually loosely formulated, does not distinguish between
being able to speak and having something to say.

Until very recently only human beings were capable of either, excepting
the talk of parrots and the recordings of human speech for the moment.
However, we now have machines capable of producing spoken or written, 
syntactically
adequate sentences for hours on end, and the prospects are bright - or
perilous, depending upon one's orientation - that such automatic speech,
suitably enriched with clever borrowings from a human interlocutor, can
fool some innocent or another into thinking he is talking to an
unusually coy and inebriated fellow mortal, when in fact he is only
interacting with the equivalent of an elaborate jukebox under computer 
control.

We speak haughtily of never being fooled ourselves: *we* are so much more
clever and sophisticated.

This, however, ignores the real issue: our perception of ourselves, which
includes a good bit of ego, and the conviction that what we feel - and
feel ourselves to be - is more important that what we can say, for we
must first feel something before we can express it. To have something to 
say means to have an unusually profound or clever insight into the world,
but one does not not have to express this via speech: music, art, sculpture,
any creative artistic enterprise suffices.

The arguments pro and con the Turing test are moot for me, as I know that
it would take an exposure to a truly moving piece of art produced by a 
computer - a symphony equal to one of Borodin's, for example - before I could
fully accept the awareness of a machine. Such acceptance would come as
an overwhelmingly pleasant shock of recognition, not as the grudging
acceptance of the concensus judgement of a panel of computer scientists.
I have enjoyed earning my keep among their ranks for some years, and I
know our strengths and limits too well. My heart requires rather more
evidence than my head, but it responds much more strongly. I suspect that
the same is true of many others as well.

This suggests an alternative to the Turing test: a computer can be
said to exhibit human intelligence and awareness if it is capable of
producing a work of art which finds both wide acceptance among art
lovers, and is indistinguishable from similar pieces or art produced
by human artists. The limitations of I/O devices being what they are,
a piece of music in conventional notation would appear to be the
most promising route.

Regards, 
       Heiner biesel@thrall.sim.es.com


