What do you think of AI art?

And yeah, spare us the “it’s ruining life for us real artists” crap here. It’s mostly normies doing clean art and the most mainstream adult art who complain anyway. Us into more extreme stuff can only gain from it, and i think most of “us” seem to understand that. I just wanna know what you think about it, artistically and overall.

Personally i think a lot of it looks really boring and unattractive due to the specific style it tends to use, but there are ways to get away from that, and i find it just hilarious to make, even if the results aren’t really worth showing to others often.

I don’t have any means to make X-rated art though, i don’t know if there are any tools for that available through the web, or only ones that have to be run locally. Some web versions of Stable Diffusion, especially SDXL Turbo were completely uncensored for a while but it seems they fixed that “mistake” about a month after it was released, the quality was pretty crap anyway, so it’s nothing i would want to show anyone, but it sure was hilarious for a while.

Which brings us to the other thing i wanted to bring up: what do you think about the subject of realistic, you know what, i mean a certain kind, yeah i think you know. I mean, realistic cub furries are one thing, because they can’t really be real, since furries don’t exist, but when it’s another thing, and it’s so realistic you can’t tell if it’s real…. yeah, just wondering what you think about it.

personally i’m pretty indifferent towards it

i’m both a digital artist and experiment a bit with AI and both have pros and cons
AI is absolutely horrible from a usability standpoint, i can never reliably generate what i want and it’s literally impossible to generate new and specific concepts without going through tons of hoops like training loras

as a digital artist i’m extremely picky and specific about what goes in my art and AI has absolutely nothing on that, but that could also be a pro depending on what you’re looking for as well! it completely removes the “self” from whatever art you generate, which makes it really good for making abstract stuff that you’d never think of.
with paper and pencil you decide exactly where the lines start and end, for better or worse.

but of course the big barrier of entry is having a decent enough GPU (maybe 1050ti and up) to run stable diffusion locally to get the most out of it, and needing to sink the time into refining your prompts to not suck and actually make something half-decent

which isn’t a very involved process and is quite boring

though the main reason i use AI is because i like to spend my time drawing things that i want to make, which actually isn’t r34 and porn
AI can pick up the slack making things that i kinda want to make (more like feel horny enough to want to make) but don’t see the value in dedicating large amounts of time to

also as for your latter point…
there’s an ethics conversation to be had about whether generating something like that with AI completely locally and not even distributing it does any harm, but i’m never touching it regardless

i do have deeper thoughts on it than that, but im not really comfortable talking about it in general to be honest, sorry!

I don’t really care one way or the other.

It seems to be the big hot new thing to scream at.

But I just don’t really care enough about it to join the pitchfork mob.

I’ve seen people use it to make some pretty halfway decent cub art though…

It is inevitable though that technology will “make such and such activity easier.”

I mean, people had freak outs against Samplers and Synthesizers and Digital Audio Workstations.

“It will make music too easy and people will stop learning real instruments.”

Whatever… sorry I was never great at playing notes in 100% correct scale and pitch.

……………

……………

I guess the big “moral issue” is style stealing, people copying your style.

Biz Markie (a rapper from the late 80s) got sued because he sampled some song.

Until that landmark court case, people just sampled anything they wanted to.

https://en.wikipedia.org/wiki/Grand_Upright_Music,_Ltd._v._Warner_Bros._Records_Inc.

Maybe the “art scene” is gunna hafta start adapting similar laws?

The guy who made the famous “Amen Break” died never seeing a cent of royalties,

despite the fact that the “Amen Break” was one of the most sampled drum loops of all time.

If Copyright Law had worked differently and in his favor, he could have been rich.

………………

………………

I care not about it.

I like the niche concepts of Worms being sexualised and/or anthrofied, most people don’t, and even if they do, they tend to put big boobs on them and give them facial surgery. I see it as a replacement for likeminded people I WISH existed.

AI image generation is for shitposting, now impossible (I miss the early engines that got really funny or abstract if you were to feed them nonsense prompts, everything’s boring and slick now) or getting a quick-n-dirty visual that is easy to describe but hard to search for.

AI text generation is for menial stuff and RP, but I wish it didn’t have some stupid word quirk (“kiss-swollen lips” makes me want to jump out the fucking window). At work, we use it for menial text generation that is otherwise pointless anyway^(1)^ or stuff like “list every pairings between these two word lists”⁽²⁾. Privately, I’ve used it a lot to fix reference formatting. Zotero is great fro generating them, but only useful if all the info gets downloaded with perfect 1-to-1 into the right boxes, and I usually waste so much time fixing that. With ChatGPT you can just paste in a bunch of shit in whatever order and tell it “generate the APA citation for this”.

(1) I.E. the clients like it, but no one will actually ever read it. For context, I work in Google Maps profile SEO.
(2) This is not in English, so yes, the AI generating the list of 40-some combinations with no erroneous elisions/plurals/agreement/typos is genuinely 100% more practical!