RE: Is generalisation a limit to intelligence?

From: Ben Goertzel (ben@intelligenesis.net)
Date: Sun Dec 03 2000 - 20:33:05 MST


thanks for the long and well-thought-out post, which I don't have time to
respond to in kind
right now ;(

I suspect that our differences on theoretical points are mostly points of
language and emphasis,
which would be resolvable if we were talking in person, or exchanging
carefully written papers rather than
hasty e-mails

> I glossed over a lot, but as you can see, info theory largely governs
> organizational behaviors to allow high predictive efficiency in a
> computationally reasonable manner and to weed out garbage. Other than
> that, it mostly looks like yet another giant network of motivated
> super-neurons, albeit optimized for silicon.

Your AI architecture sounds very interesting. It's clear to me that you
have a nice
approach to perception and adaptive self-organizing memory. From your brief
overview it seems
that you've left out some key aspects of mind (action & goal-orientation,
for example) but I
understand it was just a brief overview.

If you have any detailed write-ups of this stuff I'd be fascinated to read
them.

The inferential component of Webmind is based on probability theory and
could be equivalently
cast in terms of information theory, if one wished to do so. However, there
are other aspects of
Webmind that aren't so directly castable in information-theoretic terms.

ben



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:35 MDT