> The question is is the AfterDark model accurate enough to tell us
> something about the processes involved in fish in an aquarium? Like
When I was at the artificial life conference a couple years ago in
Boston I saw a paper given by a former classmate, now at the University
of Toronto, that described a system that combined VR and ALife techniques
to create a simulation of fish that was so realistic it was difficult
to distinguish from a real video of a shark attacking a school of
colourful tropical fish. IMHO, they did cheat a bit, mapping bitmaps
scanned from photos of real fish to the 3D models, but this is more
of a statement of my background studying in a dept. that was concerned
with the morphology of pigmentation modelled with differential equations
among other things, than a real criticism because the U of T group
was more concerned with the evolution of behaviour and they only
made the fish look more realistic for fun. The point is that the
discipline of alife is really not very far from the ability to run
behavioural experiments on virtual fish that will tell us something
about real fish.
> the Byrds algorithm, which takes three simple rules and, applied
> iteratively to a group of objects in a simulation, provides an
> accurate model of flocking behaviour and motion, a model can give
> insight into the mechanisms that might underlie the reality, even a
> simplified model.
This sounds like Craig Reynold's "boids". The same system was used
to animate the virtual bats in the title sequence of the second(?)
Batman movie. The model does indeed generate flocking behaviour
that *appears* accurate, but the question is does that indicate that
real birds somehow implement the same three simple rules in their
nervous system? Last I heard that was still an open question.
-- David McFadzean david@lucifer.com Memetic Engineer http://www.lucifer.com/~david/ Church of Virus http://www.lucifer.com/virus/