(Disney Meets Darwin)
11 Discussion
Virtual Realityc Mustn't be a Lonely Place
I am interested in a kind of virtual reality which is less
about being immersed in a virtual Euclidean space than it is about interacting
with virtual dynamic entities, and being a member of a complex adaptive system.
Real reality has a lot to do with interacting with other living things.
When explorers of virtual reality systems enter into their artificial worlds,
they must adapt to the "nature" of these worlds, to some degree. Upon entering,
users are often met with a collection of inert, Euclidean objects. Complex
computational entities can exist there too, and adapt to the virtual world,
and to the human participant as well.
In video games, behavioral entities -
moving sprites, warships, race cars, pong balls, cartoon characters - which
were already designed to live in a microworld, can be made to adapt to their
worlds, through some automatic optimization schemes. They can be successful
in their niche, able to deal with gravitation or any other physical laws
(whether Newtonian or "Cartoon"), and to respond to the actions of humans
and other entities in meaningful ways.
But they could also adapt
similarly to the ways in which performers - dancers, actors, comedians - have
adapted through feedback with their audiences, by becoming increasingly
expressive, entertaining, or funny. Often there is no computable formula
for what makes a skit funny or a dance movement provocative. The performer
must simply see if it gets a laugh or a good review. It is an optimizing
process.
The creatures described in this paper are,
of course, not conscious of the fact that they may be getting a laugh, any
more than a flower is conscious of the fact that it is attracting a bee.
They also have nothing in the way of stimulus/response mechanisms, allowing
them to learn as individuals. The population is what learns - it genetically
adapts according to the evaluations of the animator, and the settings of the
multiple fitness criteria.