AI Derangement Syndrome / Anti-AI artists / Pro-AI technocultists / AI "debate" communities - The Natural Retardation in the Artificial Intelligence communities

promptstealing.jpg

Doomhamster is also Antizionist, by the by.
 

It's laughable when it's done by AI prompters as much as it's laughable when it's done by general artists whining about people copying their style (as in just looking similar, not full on 1:1 tracing), posing, color palette, character designs, etc. Of course, only one of these is something they consider laughable.
 
It's laughable when it's done by AI prompters as much as it's laughable when it's done by general artists whining about people copying their style (as in just looking similar, not full on 1:1 tracing), posing, color palette, character designs, etc. Of course, only one of these is something they consider laughable.
I actually think it's cool when people resuse prompts but run them through a different checkpoint - it's just interesting how different the outputs can look.
 
People are seething NUCLEAR levels by Coca-Cola's new AI ad.
On one hand, I can understand, as this ad is pretty dull. It's pretty obvious AI did the heavy lifting here, given that each scene doesn't last more than 5 seconds. Very disappointing, but alas those are the limits of generated videos (:_(
That's what's weird, they aren't. Locally you can try for 6 to 10 seconds, though current models weren't trained for that and can get weird after 5. But Sora supports 15 seconds. Most sites like Kling support 10. I think they did shorter cuts just to cram more into the ad and go global with the Coke deliveries.
 
I actually think it's cool when people resuse prompts but run them through a different checkpoint - it's just interesting how different the outputs can look.

Oh, no. I like those, too. I was specifically referring to people who crash out over someone stealing their totally hard earned prompts. It's interesting seeing how different checkpoints handle certain poses or recognizes characters. Let's me know if this checkpoint's worth putting more <7GB files into my computer.
 
Why would conservatives support ai art, it doesn’t make any sense and it’s always off, shows they never cared about art anyway, yeah I know I sound like a libtranny but for some reason ai art is popular with right leanings for some reason
 
Why would conservatives support ai art, it doesn’t make any sense and it’s always off, shows they never cared about art anyway, yeah I know I sound like a libtranny but for some reason ai art is popular with right leanings for some reason
It is a tool, like a pencil or a camera. Pencils allows people to make bad drawings. Cameras allow people to take bad photos. Generative art allows people to create bad images. And social media allows all of these things to be shared. But like the pencil and the camera, generative art can also be used to create good images. These tools are out there, they're being used in industry, and if they're used well you won't even notice.

Using AI image generation as a serious artist with a vision in mind is not, at a high level, inherently different from using anything else. You have to learn how to prompt effectively (it helps immensely to have some art knowledge here so you know the names of the effects and styles you're trying to produce). You need to understand how things like seeds, CFG settings, models and even image resolution will affect your outputs. You have to curate the outputs and use inpainting and/or manual image editing skills to fix weird details and/or bring the image closer in line with your vision. You have to learn how the different controlnets work, develop an intuition for what different levels of denoising do, etc, etc. And you have to have a keen eye for detail so that you can nudge every little aspect of an image in the direction you want it to go.

Just because you're not applying each brush stroke by hand does not mean you "don't care about art." The final product can and should be subject to every kind of critique that would be given to any other work and deficiencies should egg the artist on to improve. I certainly don't believe in going easy on AI art, but I also don't believe in dismissing something because of the tools it was made with. Melty anime girl renaissance portraits with too many fingers are bad because they're bad, not because they're made with AI.
 
Eurogamer released their review of Arc Raiders, the new multiplayer shooter from Embark Studios, who also released The Final, titled "A Smartly Designed Extraction Shooter Marred by One Inexcusable Decision".

1762841563189.png
This inexcusable decision? It's AI-generated voices, which the article dedicates nearly half the article to discussing.
1762841687522.png
 
With the release of models like qwen edit you can basically take an AI generation, edit all mistakes and things you don't like roughly in any image editing program, feed it into qwen edit to have the edits smoothed out and changes applied and if done well you'll basically never know it was an AI image. That's not even delving into things like img2img, masking, style transfer, loras, controlnets, IP adapters etc. etc... Most people somehow think making stuff with image models is just low-effort writing a text prompt on a website and the model will generate something roughly in the ballpark of the prompt, but there's a lot more refinement you can do and control you can have over the process. I'm very confident there's a lot of AI art out there at this point that is not recognized as such.

You can tell people trying to make money out of it are using this because at the beginning of it all, there was quite a lot of sharing of workflows. This really, really stopped. Everyone's looking out for themselves now, trying to get ahead of others with the technology. I'd not be surprised if a lot of "anti-AI" artists are using these models themselves. It's a productivity booster.
 
Eurogamer released their review of Arc Raiders, the new multiplayer shooter from Embark Studios, who also released The Final, titled "A Smartly Designed Extraction Shooter Marred by One Inexcusable Decision".

View attachment 8153909
This inexcusable decision? It's AI-generated voices, which the article dedicates nearly half the article to discussing.
View attachment 8153917

I haven't played the game but AFAICT it doesn't even use "real" generative AI for the voices, just text-to-speech. Even that shitty article admits that it's "text-to-speech-based performances generated from the voices of real actors."

That's not stopping the rage on Bluesky, though! Mostly they seem to be crashing out over the realization that they cannot, in fact, dictate what other people are allowed to enjoy:

Screenshot 2025-11-10 at 5.12.05 PM.png

Screenshot 2025-11-10 at 5.11.52 PM.png

Screenshot 2025-11-10 at 5.11.46 PM.png

Screenshot 2025-11-10 at 5.11.58 PM.png
Does this Myka guy also chimp out when he hears people use Siri on their phones? What an exhausting way to live.
 
Eurogamer released their review of Arc Raiders, the new multiplayer shooter from Embark Studios, who also released The Final, titled "A Smartly Designed Extraction Shooter Marred by One Inexcusable Decision".

View attachment 8153909
This inexcusable decision? It's AI-generated voices, which the article dedicates nearly half the article to discussing.
View attachment 8153917
It's actually insidious and disgusting how these assholes try to control content and "punish" studios who don't act the way they want. You can just see them sitting back going "see, game devs? Don't want a 2/5 and to have your sales wrecked? Then don't use AI, chuds." :smug:

I'm picturing 5 years from now when the majority of games are using AI for development in some fashion, and they're still pulling faces and angrily giving every game a 2/5.
 
Very much different from "My special brushes UwU i trust my own skills but I wont share/sell them cause youll quickly overtake me UwU".

Ask yourself: What's the fucking point of art? Profile pictures and ERP profiles? You'll never compete with your consumers ability to produce what they actually want given the tools, and it ain't 4 months wait for 3 hours of work at $250.
 
Painfully generic AI art lookalike SamDoesArts released a vid on the coke ad:

The thumbnail is interesting, edited to have mangled hands because that’s still a thing they complain about I guess:
IMG_1969.png

Meanwhile, a frame from the actual Coke ad:
IMG_1985.png

It looks kinda fucked already, why not just make fun of that instead of misrepresenting it?
 
One thing artists are starting to do now is put "AI training prohibited" on the art they make.

Such an attempt at control is powerless -- at least in the USA -- where "training" with reference can be either Fair Use, de minimis, or not even able to be an infringement.

BTW, as I said in the AI art seething thread, AI art is not inherently "art theft": if no substantial and unique portions are used in a way that's lying about who made what, then it's not "plagiarism" -- and if there are substantial and unique portions used that can still be Fair Use under American copyright law (especially if "transformative").
 
One thing artists are starting to do now is put "AI training prohibited" on the art they make.

Such an attempt at control is powerless -- at least in the USA -- where "training" with reference can be either Fair Use, de minimis, or not even able to be an infringement.
Unless you're severely over fitting your Lora, it's still not going to be able to make a 1:1 reproduction unless you use other tools outside of just prompting. The way AI art is made is so similar in concept to how humans learn to do art so it's funny seeing artists cry about it. I wonder how places like Japan with their insane copyright laws will handle it, though nothing is stopping outsiders from scraping or yoinking stuff from pixiv or Twitter.
 
Back
Top Bottom