From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!unido!uniol!tpki.toppoint.de!elrond!freitag Tue Nov 26 12:30:50 EST 1991
Article 1429 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!unido!uniol!tpki.toppoint.de!elrond!freitag
>From: freitag@elrond.toppoint.de (Claus Schoenleber)
Newsgroups: comp.ai.philosophy
Subject: Re: AI
Keywords: ai,organic,inorganic,philosophy,artificial intelligence
Message-ID: <aBRoBB1w164w@elrond.toppoint.de>
Date: 20 Nov 91 01:00:21 GMT
References: <1991Nov18.190910.27713@uxa.ecn.bgu.edu>
Organization: Claus Schoenleber, Kiel, Germany (3-926986)
Lines: 62

cii4795@uxa.ecn.bgu.edu (Rob Koester) writes:

> Being fairly new to the concept of artificial intelligence I have a
> question regarding the philosophies that you have been discussing here.
> You have written about a number of concepts relating to ai including
> language, sight, thought process, etc..  What, specifically, would you
> consider a sucess in this line of research?  Would persons be satisfied
> with a "being" that only could sucessfully perform one of these
> functions, or would it need to be able to have complete "freedom of
> thought" (for lack of a better term)?  
> Just curious!
> 
> `````````````````````````````````````````````````````````````````````````````
> |Speech Communications                   the morning.  ...it smells 
> |Eastern Illinois University             like victory.
> |Charleston, IL.                           -Apocylypse Now
> -----------------------------------------------------------------------------

I try to answer in 'brief', Rob. [ a very relative term, I know :-) ]

A "beeing" (that would be a computer, today) that performs succussfully
one of those mentioned functions might be a great success in research
(and I think there are a few concepts that are successful in some way).

But "beeing" is (in my eyes) the wrong term. Research in AI and related
sciences do not (or don't have to) try to create Golems or homunculi;
it is much more interesting to learn about the _way_ we do our "thinking".
But "thinking" is very complex and has to do with interacting in an
environment. To realize environment you have to use a sensory system.
And that is very difficult to create. The other point is to find that
special way of storing, organizing and retrieve data our brain uses to do.

And final: How does brain work to get new data out of the stored?

The single functions would be a great success, but only the working system
of all functions might satisfy curiosity of the researchers.

You can imagine yourself the difficulties: How to create a "nose" for smelling,
a human-like eye? What is 'initiative'? (Beside: a very important subject in AI)
Is it possible to explain such a system what 'love' is, when "it" never has seen
and "touched" a woman/a man (don't know the sex of that system <grin!>)?
How to handle that huge mass of data eyes, ears, nose, tongue and skin
get as input every second in that short time?

So my answer is: Only the "complete freedom of thought" would satisfy.

BTW, "freedom of thought" is a very delicate term, and it is possible that
this freedom is just an illusion, for everybody is caught in his very own
universe of thoughts (and it's a great deed, in fact, to escape!).

Hope I did satisfy your curiosity,

regards, Claus.




Claus Schoenleber       freitag@elrond.toppoint.de
2300 Kiel 1             ...!uniol!tpki!elrond!freitag
Germany			+49 431 18863 (voice)
=================================================================
        "Never give a sucker an even break" (W.C.Fields)


