Why haven't we applied the agentic workflow philosophy to art yet? We could have multiple agents simultaneously, with computer use and maybe a decision model or two thrown in work to create art, and remove errors (e.g. bad shading, anatomical inaccuracies) with a multi-step sub-workflow. Why has no one tried this?
> no "errors"if you want hyperrealism by bother with "art" just get a camera
because no one wants slop watching it? yesactually liking and paying it? lol no
>>110206617You could still have a style with this sort of workflow.
>>110206575It already exists for 3D modeling/animation.