From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!spool.mu.edu!tulane!darwin.sura.net!europa.asd.contel.com!uunet!orca!javelin.sim.es.com!biesel Thu Apr 30 15:22:57 EDT 1992
Article 5283 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!spool.mu.edu!tulane!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: Re: Intelligence, awareness, and esthetics
Message-ID: <1992Apr27.160232.9268@javelin.sim.es.com>
Date: 27 Apr 92 16:02:32 GMT
Article-I.D.: javelin.1992Apr27.160232.9268
References: <1992Apr20.191345.27706@javelin.sim.es.com>   <1992Apr24.182714.17683@javelin.sim.es.com> <1992Apr26.111908.2640@otago.ac.nz>
Organization: Evans & Sutherland Computer Corporation
Lines: 49

barryp@otago.ac.nz writes:

>My understanding of "the" Turing test is as a means of distancing the subject
>from the prejudices of the observer.  A computer passes the test if its
>responses are such as to convince the observer that it is intelligent (compared
>to a human under the same conditions).  When the test was proposed the
>human- machine interfaces were pretty primitive and the teletype was the
>obvious means of communication between subjects and observer.  In recent
>competitions computers have done pretty well at this test.

That is how I see it as well.

>We could postulate other tests (eg the ability to guide a robot through a maze,
>or the ability to recognize objects in a photgraph etc) but the essential
>point is that we only know somthing (someone) is intelligent by observing their
>behaviour.  We are prejudiced towards thinking that humans are intelligent and
>so some test needs to be devised that separates the observer from the subject.

>Creativity (as in composing music or painting works of art) is a very special
>facet of intelligence.  I don't see any reason why a concerto composed by a
>machine couldn't be performed before an audience and them be unaware that it
>was not a human creation.  There have been displays of computer art that (to
>me anyway) seem just as interesting as many human creations.  These would
>all be valid tests in addition to the Turing test.

"Interesting" is one way of putting it. I chose a musical composition as
an example of an alternative test primarily because music has a profound
emotional impact and speaks directly to my heart. My intent is to explore
what it means to "pass" the Turing test. It is not merely sufficient
to persuade some number of "experts" - much less a panel of amateurs - 
to gain acceptance as an intelligent entity.

A hundred and fifty years ago learned men were making "rational" arguments
proposing to prove that blacks were not truly human, not fully aware and
intelligent, in spite of the fact that these people were certainly capable
of passing the Turing test. Full acceptance comes only with respect, with
the unavoidable realization that there's 'someone in there'. For me, the
most convincing demonstration is via a work of art; for others the Turing
test may suffice.

>At some point you have to make up your mind whether I am an intelligent being
>or just a clever imitation.  you can't expect to wait an infinite time or
>test every posible aspect of an intelligent being's repertoire.

Quite so.

Regards,
       Heiner biesel@thrall.sim.es.com
       who thought of M. Zeleny as leavening.


