Alt-BEAM Archive
Message #05239
To: beam@sgiblab.sgi.com
From: Steven Bolt sbolt@xs4all.nl
Date: Sat, 10 Jul 1999 22:08:23 +0200 (CEST)
Subject: [alt-beam] Re: Chaotic controllers (was Something funny with the 1382)
On Sat, 10 Jul 1999, Terry Newton wrote:
> >Emulated Nv's can do just that, if I understood Terry's postings
> >correctly.
>
> Actually, Nv's (emulated or not) don't do it, it's the -wiring-
> between the Nv's that must change in response to the environment.
Is what I meant. Sorry about the imprecise wording.
> Gigabytes have been written on this subject, search for evolve robot
> or for a real kick try evolve hardware - goes way beyond anything
> I'm messing with, but interesting things happen when chaotic nets
> are coupled to a suitable horse and the connections are coupled to
> a score based on the environment. It can't do all the far-out
> fancy stuff outlined in the GA papers but hey it's got attitude...
Don't play down your work. Real life often does just that:
Hesitate when no learned, known to be correct action is available
from memory - then do something at random - and record the success
or failure somewhere in the brain bits. Quite complex creatures
have no more (or even less) brain power than our little uCs. There
*has* to be a better way to use the tech we got. Though I'm willing
to bet that we won't find out until we have more, better, smarter
sensors...
Best,
Steve
----------------------------------------------------------------------
# sbolt@xs4all.nl # Steven Bolt # popular science monthly KIJK #
----------------------------------------------------------------------
------------------------------------------------------------------------
eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com
- Simplifying group communications
Home