🚀 go-pugleaf

In article <7657@skye.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writes:
>
>One of the things that completely surprised me when I came to
>the Edinburgh AI Dept was that some students would defend the
>idea that ELIZA wasn't just a "trick", and that (so far as
>I could tell) they really thought this was the case.  That is,
>they weren't just defending it for the sake of argument.
>What surprised me even more was that members of the faculty
>were prepared to defend the same position.
>
>I still don't know what to make of this.

      As I mentioned earlier to Chalmers, the cause of these claims of
consciousness in machines is just the flip side of the suggestion that
living systems are themselves machines -- a sort of nerdy scientific
creationism. Presumably in this philosophy, in other words, living things
are fashioned by 'physical laws' acting on 'matter'. A better way to look
at it, which clears up both issues, is to recognize that living structures
and material systems have organized themselves, built up themselves. We
have no trouble with this when we look at human society -- human beings
created the cities and technology which is integral to our lives. Evolution
created (through the energies of living things) more complex, larger and
more 'advanced' organisms. Thus within the scope of easy observation we
see that self-organization and development are a fact. Creationists, on
the other hand, want to assert that these structures were created by some
_external_ agency. In the case of modern physicists, this agency takes
the form of 'physical laws'.

      It is the equivalence of 'created by physical laws' and 'created by
humans' that gives the machine the status of the living thing, i.e. its
own creator. Both are simply mechanisms created by an external agency, so
why couldn't machines be conscious? However, looked at the other way, i.e.
in terms of self-organization, machines are profoundly different from
living systems. Living systems, in this view, created themselves. Therefore
they must possess an infinite or indefinate regression of some sort to
have accomplished this. Machines, on the other hand, are merely matter
which already exists in its own right, being pressed into the service of
self-organizing systems for the purpose of enabling them to further self
organize. Thus there is no equivalence at the basic level. And this, IMO,
eliminates any basis for suggesting machine consciousness, e.g. as a function
of program complexity.

>
>-- jd



      John E. Snodgrass [snodgras@cyber.net]