Constraints, Control, and Creative Authorship Across Mediums

Manually created animation is more dissimilar to diffusion-derived content than live-action videos or footage shot with a camera. The reason is that you don’t get to choose the laws of physics. For example, if you burn sand and turn it into glass, it can only be translucent or opaque in certain ways.

This means that when you point a camera and shoot something, the canvas is not completely malleable. You do not have full control over how someone’s hair looks, how their skin appears, or other physical details. Those are largely given to you. You then use those constraints to create something more interesting or expressive.

In that sense, diffusion, when you consider the limits it imposes on creativity and fine detail, is actually more similar to live action than to animation. An animator has the ability to choose every single element they depict, such as the exact curve of a character’s head, like Stewie Griffin. While you can modify aspects of live action, you are still not starting from a blank canvas, so there will always be details you did not choose.

This suggests that we are already familiar with the idea that certain aspects of diffusion-derived content can be considered creative in much the same way as live-action films.