Animal Behavior Test

[Lee Hulbert]

Has anyone tried to do an accurate protrayal of animal behavior in an
AI? I'm talking about an AI that could perhaps succed at a
proto-Turing test, where the test is not to appear human, but to
attempt to behave like a natural animal. Boids gives some success,
but after watching awhile, you realise that these "birds" never eat,
never sleep, never even stop to rest. Is there any porject to make a
virtual animal that behaves in a natural way? Not an alife animal,
but a simulation of a real species' behavior. If not, does anyone
have a good idea on where to start?

Lee

[Maximal]

Obviously militars are doing it. I don't know why no others are doing
it. Maybe 'X files' is not so fantastic ...

lhul...@hotmail.com (Lee Hulbert) wrote in message news:<1cdcc92f.01072...@posting.google.com>...

[Carl Burke]

Bruce Blumberg's group at the MIT Media Lab has done a lot of work
along these lines. Google for that and related sites, and you should
find plenty to work with.

--
Carl Burke
cbu...@mitre.org

[Vitorino Ramos]

Lee Hulbert wrote:

Hi Lee:

My response will not satisfy you at all, mainly cause I had other aims in
simulating this specific animal behaviour (social insects: ants). Anyway,
here is the ref. and the abstract:

Vitorino Ramos, Filipe Almeida, Artificial Ant Colonies in Digital Image
Habitats - A Mass Behaviour Effect Study on Pattern Recognition,
Proceedings of ANTS'2000 - 2nd International Workshop on Ant Algorithms
(From Ant Colonies to Artificial Ants), Marco Dorigo, Martin Middendorf &
Thomas Stüzle (Eds.), pp. 113-116, Brussels, Belgium, 7-9 Sep. 2000.

Abstract: Some recent studies have pointed that, the self-organization of
neurons into brain-like structures, and the self-organization of ants
into a swarm are similar in many respects. If possible to implement,
these features could lead to important developments in pattern
recognition systems, where perceptive capabilities can emerge and evolve
from the interaction of many simple local rules. The principle of the
method is inspired by the work of Chialvo and Millonas who developed the
first numerical simulation in which swarm cognitive map formation could
be explained. From this point, an extended model is presented in order to
deal with digital image habitats, in which artificial ants could be able
to react to the environment and perceive it. Evolution of pheromone
fields point that artificial ant colonies could react and adapt
appropriately to any type of digital habitat.

What probably will be of interest to you is the model. I have used it in
order to map all the pheromone distribution (of the ant colony) evolving
within digital grey images. But you can use it to map (simulate) ant
positions (evolving on time) on any X,Y space, achieving the patterns you
find in real colonies (TRAILS: the mathematical model relies on real ant
behaviours). You can get a PDF copy (long version) at the following
web-site:

http://alfa.ist.utl.pt/~cvrm/staff/vramos/ref_29.html

For an e-PostCard ("A Strange Metamorphosis" [from Kafka 2 Red Ant], in
PPS format) showing an aesthetic interpolation possibility to the
following model check:
http://alfa.ist.utl.pt/~cvrm/staff/vramos/kafka2ant.html

and, last but not least, a general paper (a ART+SCIENCE project for an
ART BIENNIAL, Portugal + Netherlands) concerning other ideas (sinergy
among human users in the internet) within the same strategy:
http://alfa.ist.utl.pt/~cvrm/staff/vramos/ref_37.html

Unfornately, the figures in this last web-page can only be seen in
Netscape (I am working on it for IExplorer). Anyway, one of those figures
is available here:
http://www.lxxl.pt/bienal/niu/niu.html, or here
http://www.bienalutopia.com/ (the official UTOPIA ART BIENNIAL web-page).

Hope this helps or at least inspire some of your research, King regards,
Vitorino

------------------------------------------------------------------------
Eng. Vitorino J. Castelo Ramos
MSc / PhD Research Fellow on Artificial
Intelligence, Artificial Life & Image Analysis

CVRM / IST (Technical Univ. of Lisbon - PORTUGAL)

URL - http://alfa.ist.utl.pt/~cvrm/staff/vramos
MAIL - vitorin...@alfa.ist.utl.pt
------------------------------------------------------------------------
"When you only have a Hammer, you tend to see every
problem as a Nail." , Shopenhauer
------------------------------------------------------------------------

[Mike B.]

Check out this excellent article by Ernest Adams on Needs-Based AI.

http://www.gamasutra.com/features/designers_notebook/20001222/adams_01.htm

"Lee Hulbert" <lhul...@hotmail.com> wrote in message
news:1cdcc92f.01072...@posting.google.com...

[Didier Vaiser]

I knew about a toolbox able to help do this kind of things : swarm
(http://www.swarm.org)
If I remember well you can find some very interesting links about
realizations.
Or you can also read the Alife books (ALife I, II, III, ...) Some
experiments were very surprizing and even found applications in the real
(pest management e.g.).

But this is quite far from me now, so good luck and hope it helps you

Didier

[Keith]

There is always that catz and dogz (or whatever it was) by Microsoft which
simulates animals (only, of course, in a very cartoon like way). I am
however
curious what possible use such a program could be or lead to apart from just
general
interest or for entertainment. Good luck anyway, sound like fun.

Mike B. <micbona...@home.com> wrote in message
news:xFW87.5130$S86.1...@news1.rdc1.mb.home.com...

[Lee Hulbert]

Actually, I am interested in the entertainment aspect. I'm trying to
research the concept of a virtual safari. Wander through a virtual
Sarengeti(sp?)landscape filled with realistic animals going about
their daily lives. Both as an educational tool, and as an
entertainment, with the possibility of allowing virtual big game
hunting.

"Keith" <keith....@durham.ac.uk> wrote in message news:<tmjj1b4...@xo.supernews.co.uk>...

[Keith]

This sounds a bit like the game Sims which I was reading about on one of
these groups recently (comp.ai ?). And although perhaps not true AI, this is
how theoretically it could work. Your animals (inhabiting the world) have
desires/feelings (i.e. hunger/tiredness) that change over time. They make
their decisions on what to do next according to the relative weightings of
these desires and what objects they come across. So a pool of water (object)
will have a number of possible uses (functions) such as washing or drinking
that say an elephant can use depending on its current state (hot or
thirsty). You might have played Sims, I haven't, but if that is the sort of
thing you were thinking of but with animals then it might be worth taking a
look at that.

Lee Hulbert <lhul...@hotmail.com> wrote in message

news:1cdcc92f.01080...@posting.google.com...