From: Steve Witham (sw@tiac.net)
Date: Wed Dec 24 1997 - 07:28:51 MST
>Brian Atkins, <brian@posthuman.com>, writes:
>> They showed some cool video of what appeared to be evolving
>> GAs in a "real" 3D environment. This was the first time I'd
>> seen that, and unfortunately I didn't catch the person's name
>> who had done it.
Hal Finney writes:
>Very interesting! One problem with such simulations though is that
>sometimes they are "cooked" so that a particular outcome is more or less
>inevitable.
Sounds like some work by Karl Sims, while he was at Thinking Machines.
It was pretty general. A tree structure of segments where parents and
children were connected by joints. Each segment was a rectangular prism
and contained a neural net and maybe some sensors. The net's inputs
and outputs connected to neighboring segments' nets, local sensors
and joint "muscles." He made the physics fairly realistic.
One cool effect resulted from a bug in the physics where you could gain
momentum by clacking two parts together. Organisms evolved to exploit
this and that's how he found the bug!
He evolved things using selection criterea for: getting far
in water, getting far on land, getting high off the ground, and winning a
two-player grapple-the-hocky-puck game. For each of these a variety of
forms evolved.
--Steve
-- <sw@tiac.net>Steve Witham Don't dream it, su to it.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:45:16 MST