You are not wrong. This is surface level and weak.
Reallusion AI studio is actually doing some really interesting stuff by connecting their output into generative AI. I’m not 100% sure, but it looks like they’re using some kind of comfyUI setup under the hood that outputs several different AOV‘s from their software and uses them with open source and high-end models to create some very realistic looking stuff.
Im just playing around on the surface level at the moment - but that kind of flow is what I’m going to push towards. Ultimately that’s where this stuff is going.