TheNextWeb has a writeup on an interest app called WordsEye. Basically you describe the scene you want to create and it’s created. “The technology behind it works by speech tagging and analyzing the context of the words used. Statistical parsing is used to bring the sentences into a format the computer understands and uses to create the 3D image.” I wanted to play around with it a bit but it’s in closed beta; I’m waiting on an invite. If you’re friends with me on Facebook you know I post my weird dreams occasionally; now I have an illustration tool!