In this example, I've been playing around with new rendering techniques, and using more compositing. I've been experimenting with fractal depth and synthetic photography. These images are all rendered with models that are based on historic dadaist art, as well as my own 3d work. It's interesting how the ai training process has handled all of this. Combining styles from the dataset in new ways, and even creating entirely new 3d characters, in addition to the named ones I put in. It's also tamped down my natural cartooniness a little, and given all the humanoid characters better skin, so bonus! Another thing I didn't expect, but should have, is because of all the collage art I fed it, this model just happens to be excellent at vintage photography. It's interesting.