Raw LLM Responses

My main concern is that any intent behind the art will go away and "artists" like these just let the machine throw images at them until they find something salvageable. Then they pretty it up with inpainting because they never bothered to learn to draw even a little. But in that process there's no space for any sort of artistic intent beyond the screening process. I'm sure you can get the AI to generate a pretty ring on a zoom of a finger or something, but the overall result isn't made from decisions like that. And the tools aren't adapted for allowing that kind of guidance. And on top of that diffusers doesn't fit the workflow for artists. It doesn't make drawing a sketch more effective or help visualize many different small steps when making anything. It's all or nothing and the entire method is algorithmically oriented towards finding normal good cases. Not what people tend to enjoy art for which is something that's different about it. If the concern was about the quality of the work in terms of lighting consistency and proportions it can be expected to be solved at some point. I also think there's a minor point in that ai artists would be conditioned by the software to like what's being generated. All the incentives point in that direction and it doesn't need to be deliberate. I wonder how audience expectations will change.