From newshub.ccs.yorku.ca!torn!cs.utexas.edu!sun-barr!rutgers!uwvax!meteor!tobis Thu Oct  8 10:11:03 EDT 1992
Article 7099 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!torn!cs.utexas.edu!sun-barr!rutgers!uwvax!meteor!tobis
>From: tobis@meteor.wisc.edu (Michael Tobis)
Newsgroups: comp.ai.philosophy
Subject: Re: AI rights :-)
Message-ID: <1992Oct2.195224.3903@meteor.wisc.edu>
Date: 2 Oct 92 19:52:24 GMT
References: <1992Oct1.232114.1593@murdoch.acc.Virginia.EDU> <1992Oct2.021220.18977@meteor.wisc.edu> <1992Oct2.122504.22245@murdoch.acc.Virginia.EDU>
Organization: University of Wisconsin, Meteorology and Space Science
Lines: 70

In article <1992Oct2.122504.22245@murdoch.acc.Virginia.EDU> lfoard@Turing.ORG (Lawrence C. Foard) writes:
>In article <1992Oct2.021220.18977@meteor.wisc.edu> tobis@meteor.wisc.edu (Michael Tobis) writes:
>>Not being a person who believes that the universe was created only 10,000
>>years ago or any such nonsense, it is still my working hypothesis that
>>information processing is neither necessary nor sufficient for sentience.
>                                   ^^^^^^^^^

>Maybe not sufficient, but how can one be sentient without having a brain?
>Don't resort to the supernatural, it still requires information processing.

Hmm, did I say that? Maybe I went too far: I meant to say intelligence
is unnecessary for sentience: consider the example of a cat. Also note that
the moments when experience is most profound, i.e, ecstatic experiences,
seem to be completely disjoint from the moments when intelligence is most
engaged. In any case, despite my inadvertent overstatement, that statement
is not objectively falsifiable.

>>Before I will willingly grant rights to your constructs, you will have
>>to convince me otherwise, or provide some other evidence for their 
>>sentience. What evidence do you have that artificial sentience is
>>equivalent to artificial intelligence? 

>Not all artificial intelligence would be, but there is absolutely no reason
>to believe that a machine cannot be sentient, after all we are machine and
>we are sentient....

Again, by assumption only. Unless and until it can be shown without
handwaving that experience can "emerge" from physical processes, the
statement that humans are machines is unverified.

>>What evidence do you have that
>>artificial sentience is possible? 

>The only thing that would prevent this is the existance of a supernatural,
>something which there is absolutely no evidence for. Otherwise we know
>that AI is possible since the brain works.

This is argument by contempt, not reason. There is clear subjective
evidence of a phenomenon that is unaccountable by purely objective
physical reasoning. The phenomenon is awareness. Given that this is
the phenomenon in question, it is highly premature to say we know
how the mind works, regardless of how much we know about the brain.

>>How do you propose to distinguish between
>>sentient "true AI" constructs and mere non-sentient though complex tools?

>Whats the reason to distinguish between a person and a complex tool?
>What distinguishs an AI passing the turing test from a human being?

Maybe nothing important, but maybe a lot. My point is we cannot know.

Awareness is not detectable objectively, and the "Turing test" is just
a formalization of our intuitions. We can design constructs like optical
illusions to fool our intuitions. Even if unanimous agreement is reached
that line A is longer than line B, it still is not always the case, e.g.

In that instance, we are fortunate to have objective tests. In the case
of AI, you propose to design a system that can pass people's intuitive
test for consciousness, and since such a test is the best test that is
possible, declare that the intuition is an infallible measure of the
phenomenon. 

But it ain't. Only private subjective knowledge knows whether private
subjective knowledge exists, and a model designed to pass the Turing test will
make the assertion that it does, whether or not the assertion is true.

My ocean model isn't wet. What makes you think your model of consciousness
is conscious?

mt


