The new version of the image generation is so incredibly garbage it’s genuinely unusable. Instances of AI artifacting have multiplied exponentially, the system barely follows prompts, characters are jumbled messes of blurry features, it’s awful. If it’s an issue of expense, I would be willing to donate along with countless others if you took a collection. Please as a community we beg of you to change it back. I would be happy to pay for access to the better model, or a subscription for improved functionality, just anything but this.
“slop” is currently used as a catch-all term for AI generated content, so by all means, we are making slop. Own it.
Why would you build your entire posture around opposing updates altogether and then ask for a rollback on top of a fee, instead of providing concrete examples and references of how your input is being misinterpreted so that the Dev can maybe try and fix it?
¯\(ツ)/¯
Effort, perhaps. Artsy slop isn’t impossible to achieve, as you claim, you just have to tinker a bit. It is true that some looks are harder to achieve than others, but number one that’s always been the case, and two, truly impossible styles are few and far between. Pulp fiction covers is definitely not one of them.
So if I had to take a guess, I’d say you have the CFG turned all the way up and that’s what’s causing 90% of your problems, yet you’re out here blaming the tool. Take this:
(guidanceScale:::1)And put it somewhere on your prompt. For anything that’s supposed to look like a painting, using values between 1 and 5 will typically give you the more natural-looking results, whereas taking it higher will result in a more polished, artificial and plastic look, reminescent of the evil cartoons that haunt your dreams.
Try it. Play around. Assume your prompt is bad. Report your findings. And ask for help before launching another drama series about your beautiful collection of 10s of thousands of slop images in a serious, academic, representational art avec des croissants style. ( ~_~)