Re: Artifical Intelligence can not be Artificial !

From: Stephen Harris (cyberguard1048-usenet_at_yahoo.com)
Date: 02/15/05 

Date: Tue, 15 Feb 2005 11:32:54 GMT

"Tim_Miltz" <tmiltz@etaion.com> wrote in message
news:1108452027.223196.199100@c13g2000cwb.googlegroups.com...
> Perhaps the ALL of AI boils down to an underlying issue to intelligence
> That is, if you don't have a living host, or criteria to define that
> which is 'living', this whole AI pursuit may be full circle back to
> Aristotle's 5 interdependent criteria for some thing to be considered
> 'alive' before you can entertain a definition of intelligence.
>
>

"Aristotle speculated that all life forms showed four activities:
nutrition or how food was obtained; growth; movement; and
"feelings or soul". Movement to Aristotle not only meant mobility
but it also mean changes an organism undergoes during its lifetime.
The last activity, feelings or soul, is interpreted today as sensing
and responding to the external environment.

The activities of life today are similar to those of Aristotle
but have been expanded as our knowledge about organisms grew. The
activities of life include metabolism, growth, replication and
reproduction, movement, and response to external stimuli."

SH: Nowadays, artificial intelligence is defined as behavior that would
be called intelligent if a human were to do it. This doesn't pay much heed
to what creates the behavior. The Turing test also used behavior as a
criteria; it was similar to if a program could hold an email conversation
with you and convince you it was human, then the behavior was intelligent.

It is a bit hard to distinguish just what type of thinking a human being
can do that a machine can't. Loosely, one might call it creative, but that
is hard to define and veryify an instance of it. They have chess programs
that can play grandmaster level chess now. That used to be regarded as
a measure of intelligence. But the programmers of the chess program(s)
says the program doesn't play chess like a human which takes into account
the source of the behavior. The programs succeed by brute force
calculation not thought. I think it will be necessary to require sentience
in
order to allow the verb thought, otherwise it is calculation. The claim that
the right program will instantiate a mind is a philosophical claim. The now
less cherished claim of strong AI, or GOFAI has receded, and like the
Turing test didn't require physical situatedness/sensory perception which
would be used to sense and respond to an environment plus build up a
common sense understanding of how the world worked, necessary to
interpret meaning within physical contexts (harkening back to Aristotle).

Regards,
Stephen