Re: Creatures game (was re: AI)

From: Elizabeth Childs (echilds@linex.com)
Date: Sun Dec 19 1999 - 13:52:31 MST


> >From my understanding of the game it's a good sim for training and
> breeding but not for evolution. The number of creatures was too
> small for evolution to work properly in the first game. If it did
> work, the biology of the game permits creatures to need no food,
> live forever, regenerate, etc., with the proper mutations and so
> I'd expect evolution to concern the development of those things
> and not more "realistic" adaptations.

One of the good things about the game is that players can design new objects
for the game and have a lot of control over the creature's genome. Thus,
there is a large community of Creatures fans making new toys and new
creatures. You can download immortal norns, exceptionally aggressive norns,
norns that are unusually responsive to praise, and many other variations.

Also, you can get some objects that make it much easier to raise the stupid
things, such as a book that will teach it to talk. Even so, raising them is
amusing for a few hours, but it quickly grows dull.

Some people load up their world with the maximum number of creatures, then
let it run for twenty-four hours or so and declare the survivors to be the
best breeders. But the creatures aren't really bright enough for this
experiment, and most of them die before the time is up. There's usually
only one or two left.

I wish the same company would put out a game with some animals that are less
cute that have real evolution, preferably with predators and prey.

> It *is* interesting that the Norns get behavior from a simplistic
> neural-net-like brain structure. They do demonstrate adaptive
> behavior without explicit programming, and that's quite a nice trick.

I agree, this is well done. Also, if the Norns aren't stimulated, they
start to get brain damage - I'm not sure if this is done on purpose
(probably) or is an emergent property.

Also, some of the Norns are smarter than other Norns, as some sort of
emergent property of their genes.



This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:06:08 MST