Is it a certain AI body art style? I unno
My older friends don't understand how I can tell, they think I'm just paranoid.
I can't really explain it other than it just has an AI look, like a artstyle or a typeface
anyone able to explain it better?
Is it a certain AI body art style? I unno
My older friends don't understand how I can tell, they think I'm just paranoid.
I can't really explain it other than it just has an AI look, like a artstyle or a typeface
anyone able to explain it better?
Because it looks like AIslop and not something a human made.
this is the type of face you get 100% of the time when you type in "perfect, masterpiece, beautiful, sexy" etc. The shading is a result of having the CFG too high so it appears burnt. The lines melt together across the whole picture. No you're not crazy. It is the lack of creativity of a subhuman
Because it's the type of AI shit a pajeet who doesn't know how to prompt makes
Basically this:
model is trained on art + photos, results in overly realistic lighting that most artists avoid in favor of stylized light
also the transparent part of the cloth is a dead giveaway
there are just a lot of typical easy to spot AI flaws in this picture, hair, eyes, mouth, fabric, etc. and a typical mix of anime and realism, even better AI pictures suffer from this
>despite the fact that there aren't any deformities
are you blind?
Yeah, I'm wondering why he thinks that nose and buck teeth are normal, or her collarbone.
if you don't get out much and mirrors you don't know how people look like
avoid mirrors
I didn't even look at it in full definition to notice, I was just going off the general composition/style
Bro just open the image to its original size. Artifacts and hallucinations everywhere (backpack strap running under her clothes, blurryness/distortion around the hair strands, ear lobe is fricked up beyond all recognition). Like "no deformities" dude her collarbones look like they're going to explode out the top of her chest. Anyone who can't immediately see these just doesn't know basic anatomy.
This shit is just AIslop and it's worth nothing, the artstyle is the same as any other AIslop in general and the deformities happens in the same way
AI is limited greatly by its input. So all the art AI you see is just a better looking version of whatever coomers and artists were drawing for centuries. Ai is not creative, nothing original will stem from AI. People who take AI seriously are people dont understand high school maths, ie women and leftists
most of life doesn't require creativity. I've used chatbots for cooking and planting vegetables. It handled such NPC activities with ease.
If you have to ask "AI" to do things you should already know how to do yourself, then YOU are the NPC.
I should magically know how to cultivate the three sisters of the indigenous people of north america and spend several 10s of hours learning the method and years learning it thru trial and error.
Or I can do it in minutes with detailed steps from a chatbot
No, frick off satan
>there aren't any deformities
Almost everything is off in this image (the pupils, ears, teeth, etc.), but the most obvious giveaway is that rubbery/plastic look so typical of AI slop. Perhaps due to crappy training data or some other factors, most t2i models converge to this look and people don't bother to "prompt away" from it, in fact they find it pleasing.
Because it has the Korean idol anime like sameface, a result of overtrained slop models.
Also the skin finish and all the textures are shit
Because your brain works on the same principles as AI does, thus after seeing hundreds of images created in the same statistically significant method you can distinguish AI models from each other.
AI always produces a statistically most probable answer to a query, based on its learning data, but cannot create anything on its own, thus always makes 'samey' results.
>AI always produces a statistically most probable answer to a query, based on its learning data
you can enter more obscure results and get more obscure answers
That's a deformity
Because of her face. I don't even have to see the image full size to make out that this is what 90% of all SD1.5 models produce for "beautiful woman, masterpiece"
This just under your post
I wonder if people who don't see it within seeing 5 images have some kind of face autism
>This just under your post
what did he mean by this
Chronologically.
Yes but why refer to it, what about that face?
It followed OP chronologically on 1.
What's so difficult to understand?
OP ere
yeah everyone in the thread is pointing stuff out and saying I'm blind but I didn't look at it in full size, I just saved it cuz my brain just clocked onto the model cuz it's the most dead giveaway that you can't explain had it not had deformities
I'm tired of seeing this hideous AIslop art everywhere, I'm considering blocking all images off the internet and going text-only.
soon enough your browser will be able to detect and block images based on content https://www.w3.org/TR/webnn/#usecases-application
At the very least the overly soft skin on the "photos" is giving it away, morons still haven't figured that out
because someone "accidentally" put in "child's face, young" in the prompt. Like for 80% of the pictures here.
yes, AI faces of men are harder to spot from a thumbnail.
but you could find "art" like this before the ai craze and it always pissed me off, it was mostly made by npc so still ai in some convoluted way
no you couldn't
>t. twitter "artist" npc
cool technique you have there buddy, too bad you forgot to learn how to have a soul
you could, indeed
it's the excessive amount of detail, the fairly uncommon style, and yes, there are deformities, just subtle
Most of those have deformities and janked/abstract backgrounds, but you can easily be fooled if you don't know what to look for.
>I can't really explain it other than it just has an AI look, like a artstyle or a typeface
Yes, it's the style (usually). If it's not photorealism but with cartoon shading and/or other anime characteristics, then it's the amalgamation of everything on pixiv, or anime but with western-style noses.
But sometimes your brain will pick up on deformities even if you don't consciously notice them. Lots of people focus on hands, but I find that eye deformities in AI art are much more common (you can hide hands, but eyes are pretty much non-negotiable.)
hands, eyes, net-like textures, lineart (if present), skin, body shape, abstract backgrounds are a common giveaway, other subtle deformities, weird lighting, deformed distant objects, and OP's pic is overcooked, as another anon said, it's tons of giveaways
it's been over a year of this stuff and prompters continue to be defensive babies about the limits of their gens
they could have genuinely learned to fix all the issues manually by now but they're just waiting for the next model to save them
AI cannot get colour and lighting right, and every new version they release that isn't changing
also the models are all way overfit on stock images so a lot of it has that look to it
Because AI art typically has a similar style?
>Because AI art typically has a similar style?
it does not, AI can literally do a thousand different artist styles
https://rikkar69.github.io/SDXL-artist-study/tags/
most users just fail to unlock this potential
Yes which is not contrary to what I said
Get a lora for a different style.
>aren't any deformities
There's plenty
>how
Because it uses a popular model and popular prompts. Don't kid yourself into thinking you can recognize all AI art at a glance, you're not the genius that you think you are, you just never tried using AI and don't realize how versatile it is
Because the AI art that is shared around for you to see is made with the same prompts and models. If you look at stable diffusion threads you can see that it can produce many different art styles
Because it's an overrendered neotenous ambiguously white/asian woman staring blankly into the middle distance in front of a vaguely futuristic background.
AI food images are almost impossible to tell apart from the real.
>anime+photo+digital painting+neon
SD sloppa
AI generated images have a distinct style 99% of the time but like I said many years ago when people were panicking about computers emulating instruments, "it's a computer, why does it need to sound like a violin?". Same deal here and it's probably a good thing. It will give everyone who hates it a chance to vent their ire, and most people don't and won't ever care.
teeth, ears, jaw, cheeks, chest
just stop torturing the fricking machine, jesus chirst, we're wasting electricity on dumb shit