It generates random 2D pictures, and then people vote on whether the
pictures are good or not. Strikes me as quite an interesting paradigm
for content development ala "Spore." I've thought about implementing
all kinds of evolutionary interfaces for all kinds of programming and
art problems before, but I've never really gotten around to it. This
shows a certain level of doability and proof of concept. I think rather
than leaving images to public vote, one would just vote privately on all
one's stuff, using a sliding scale from 0 to 10 (fractions allowed). 2D
images or 3D objects would be generated in batch over some time chunk.
Then the "artist / designer" would sit down with a sheet of these things
and decide what's likeable or not. The problem is probably dominated by
how fast an artist can submit votes. I bet the time to generate images
or models is a lot less than the time it takes to analyze and vote.
Brandon Van Every Seattle, WA
"The pioneer is the one with the arrows in his back."
- anonymous entrepreneur