Averil ✒️ DoomBloomArtIt's been a few years since I drew this. I still stand by it.
I don't repost this drawing as often as I'd like to because whenever I do, I attract insufferable bros.
If you are an insufferable bro arguing about AI images I will block you because I'm tired of your shit
Ashwin DixitHere's a project I am working to bring to life. I think you will find it relevant.
@purrperl Hi there! I think it's admirable that you're trying to build an alternative that addresses some of the known ethical problems with current models.
That being said, your post is pretty brief and I don't know what the goal is and who this is for, other than addressing the named issues on a technical level. I personally don't see a use for LLMs for anything but pure language problems, and for art specifrically, tools that don't aim to put out a "result" are 20000% more useful to me.
@purrperl What I'm trying to say is that the AI assistance for me as a visual artist anything of the kind (text|img) -> finished img because even if properly licensed and compensated, it's still cutting most of me out of the process. I don't want that and I'm sure many writers, musicians etc see it the same way.
On the other hand there are ML-driven tools such as: "scan the body/hand pose from an image and apply to a low poly 3D model for rough reference" which I find immensely helpful.