[happytoexplain]

Intellectually, and as a creator, I love AI image generation. As a casual user of the internet, I'm growing more and more annoyed with it, simply because of all the times I lean in to examine an image that my visual cortex is stumbling on some part of, only to realize after a couple seconds, "oh, it's AI", and sit back up straight.

[MintsJohn]

This is why ,as useful as they are, I also loathe llms, so much textual content is endless drivel, either fully created by AI or helpfully rewritten. Of course it's not new, but what is new (to me) is users using llms in discussion topics to either troll or to make their point, or receiving (real, work related) e-mails fixed up by chat-gpt; madness.

[mdp2021]

> only to realize

All of it? Some people would immediately recognize the typical StableDiffusion output but not a Midjourney.

[DrSiemer]

Midjourney is a limited set of models and (to me at least) much easier to spot than Stable Diffusion output, with it's giant array of custom models, LoRA's, textual embeddings and tricks like HiRes fix and upscale script.

[iinnPP]

I used SD1.5 when it came out and left it alone until a few days ago. The improvements are incredible and hard to articulate.

If anyone put SD down for a good period of time, they should really give it another go. Use the full suite of options and make sure you are using a model fitted for your purpose.

[mdp2021]

I just checked their website and still see monstruosities, misunderstanding ontologies or hallucinating them. See this image in their own showcase:

https://images.squarespace-cdn.com/content/v1/6213c340453c3f...

[DrSiemer]

That's probably because they use "clean" output from their own base models. What you get from those is a far cry from what a custom model like DreamShaper together with negative embeddings can do.