AI Art Seething General

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
I actually don't mind AI art but I 100% agree on this. I was trying to get some steampunk inspiration the other day and it felt like every fourth image was some poorly rendered AI shit. It was the same deal when I googled Medusa- tons of AI schlock in the images. I only want to see that stuff when I specifically look for it and not any other time. It's deeply annoying.

found a hack where inputting -ai or before 2022 works wonders. Don't have the post, but it works 100% i think.
 
An SFM animator with over 13 years of career used AI in one of his animations... All hell broke loose.
Tweet | Archive | Youtube Link
Edit: Noticed something curious, people are commenting about the animation normally on Youtube.
But, on Twitter? People are overfocusing on AI usage while completely ignoring the animation itself. Curious.
2ih3JULaGX.png

1706455735347.png

1706455767118.png

1706455800979.png

1706455821498.png

1706455842602.png

1706456037702.png

1706456094090.png

1706456227425.png

1706456254535.png
 
An SFM animator with over 13 years of career used AI in one of his animations... All hell broke loose.
Tweet | Archive | Youtube Link
Edit: Noticed something curious, people are commenting about the animation normally on Youtube.
But, on Twitter? People are overfocusing on AI usage while completely ignoring the animation itself. Curious.
View attachment 5673983
KaiserNeko speaking on behalf of Grant Goodeve is just hilarious to me. All that "we," like he's been voted the lead spokesman of every VA on Earth.
 
An SFM animator with over 13 years of career used AI in one of his animations... All hell broke loose.
Tweet | Archive | Youtube Link
Edit: Noticed something curious, people are commenting about the animation normally on Youtube.
But, on Twitter? People are overfocusing on AI usage while completely ignoring the animation itself. Curious.
View attachment 5673983
These actors do not want you doing this.
Like I said. Most actors do not give a shit if you take existing clips and remix them for fun.

What is the fucking difference between remixing and AI voice synthesis, other than that one sounds much better than the other. Either way your not giving money to the original voice actor, you're making them say things they didn't say before, both can be used for fun. Of course its coming from a no name "voice actor," who probably hasn't ever had a good paying gig because he charges too much and sounds like shit.

All this AI seething, from Palworld to voice acting, has made me realize that the only people who don't actually like its application are two-bit wannabe artists that hate that the common Joe can make whatever he wants now at the click of a button, instead of paying for their exorbitant rates.
 
Using models, characters, textures, and software by Valve that doesn't belong to you? Perfectly fine.

Using a voice that doesn't belong to you? How DARE YOU. You CANNOT do that. You need to PAY that person to say a few words into a microphone.

Like the PaleMoonLight user said, nobody cared when they were taking the actual voice lines and splicing them together. That was directly taking that person's exact voice and altering it to say other things, and it was considered fine. But now, it used "AI" and so it is bad, that's it, no other thought required. Total knee jerk reaction.
 
All this AI seething, from Palworld to voice acting, has made me realize that the only people who don't actually like its application are two-bit wannabe artists that hate that the common Joe can make whatever he wants now at the click of a button, instead of paying for their exorbitant rates.
Frankly, I hope fields like concept art and animation get flooded with pajeets and beaners, and when the same anti-AI tards start bitching about being unable to find work or being forced to work for pennies they get accused of being racist.
 
Why do you think people are generally pro Large Language model but anti AI art ?

Let's keep it a buck as pointed out here a lot of the backlash to AI art is not logical at all, for example people supported straight up taking voice lines and splicing them together, yet using an AI to make new voice lines from scratch using the same voice is wrong somehow ?

Do you think AI art will become a partisan issue ? Some argue it already has but I am not so sure it has yet.
 
Why do you think people are generally pro Large Language model but anti AI art ?
The answer is pretty simple: they are self-centred and don't give a fuck about anything that doesn't directly affect them. Solidarity with novelists? No, no, that involves me... reading words, and using my head. I just want to draw cock and tits, leave me alone.
 
Using models, characters, textures, and software by Valve that doesn't belong to you? Perfectly fine.

Using a voice that doesn't belong to you? How DARE YOU. You CANNOT do that. You need to PAY that person to say a few words into a microphone.

Like the PaleMoonLight user said, nobody cared when they were taking the actual voice lines and splicing them together. That was directly taking that person's exact voice and altering it to say other things, and it was considered fine. But now, it used "AI" and so it is bad, that's it, no other thought required. Total knee jerk reaction.
When AI voices were popularized TF2 SFM animations were the first to embrace it
One of the most popular SFM animation as of recently uses AI voice cover of Joji's Tick Tock, and it's universally loved.
 
Why do you think people are generally pro Large Language model but anti AI art ?
I think it comes down to three aspects: novelty, quality, and threat.

People are used to being able to put text into a box and get text back. Searching the web for info isn't new. So having an "AI" that takes in text and spits it out isn't as surprising as suddenly having access to the ability to put in text and have a machine spit out a full-fledged image. At least not at today's current quality levels.
LLMs have improved greatly, but they're still obviously limited. AI image gen is still limited and funky, but has come a very long way since the days of things like Google Deep-Dream. So part of the pushback is how sudden this increase in quality is, whereas most people writing stories aren't going to be threatened by a program today that only remembers a couple pages at a time.
And that threat is what motivates the large anti-AI art movement. You have a large number of people whose identities are wrapped up in the idea that they're capable of a skill that most people do not possess, suddenly confronted by the fact that a new and decent-quality alternative has arrived and is threatening their position, both in making money through commissions and their self-image as a possessor of talent.

In the next five years, if we have GPT-4 successors becoming high-powered enough to successfully imitate high-quality writing, while still being able to be run on consumer hardware available to the masses, then we might have a second wave of screaming as "professional authors" are suddenly at risk of being overshadowed by John Public typing in a couple paragraphs and having a machine spit out 200k tokens worth of mediocre prose.
 
Why do you think people are generally pro Large Language model but anti AI art ?
In my experience with artists, they don't care about writing because everyone "can" write. It's something basic that's taught in grade school. Graduating even middle school requires you to be able to write a sentence and an essay, so it's considered trivial. (Of course, writing good fiction is completely different from writing a good grade school essay, but they don't care about that.)

But drawing, that's special. Not everyone can do it well, and it takes a lot of practice, and so they consider it sacred. Making writing easier doesn't matter, because it is already "easy", but making drawing easy? Now you are affecting them. Now their special skill is able to be accessed by anyone, and that's a threat to their social status. They don't want the average person being able to make a picture, because it takes away from them being the special ones able to do it.

This is why we saw the enormous glut of absolutely horribly written webcomics in the 2000s. Every artist thought they could also write a competent story. The art may have been passable, but the actual plot and dialogue was horrendous.

Also, as Mr. Miyagi said above, generating text is relatively common. Chatbots and the like have existed for decades, so it isn't that novel of an idea. Obviously none of those were on the same level as current things like GPT-4, but the concept has been around for many years. Generating images, though, that is new.
 

Someone made a fan pretend cinematic trailer for the Imperium of Man, the human faction in Warhammer 40,000. Someone in the comments has pointed out the mild humour in using AI to generate this. Anyway, personally I was shocked by how impressive it is in many respects, despite having many obvious issues. This type of stuff would be a pipeline dream just a year ago. The progress of AI is staggering.

1706533269084.png
1706533311547.png
1706533374881.png

It does make me wonder if we will get to the point where average Joe can generate an entire film on their RX 7000-series AMD card in a few hours of tinkering. With how things are going, it doesn't seem impossible. We just need someone to release the closed source models, whether through legal or illegal means. Then we will see the same type of jump towards the impossible we saw with Stable Diffusion and LLMs.
 
Last edited:

Someone made a fan pretend cinematic trailer for the Imperium of Man, the human faction in Warhammer 40,000. Someone in the comments has pointed out the mild humour in using AI to generate this. Anyway, personally I was shocked by how impressive it is in many respects, despite having many obvious issues. This type of stuff would be a pipeline dream just a year ago. The progress of AI is staggering.

View attachment 5675936
View attachment 5675938
View attachment 5675940

It does make me wonder if we will get to the point where average Joe can generate an entire film on their RX 7000-series AMD card in a few hours of tinkering. With how things are going, it doesn't seem impossible. We just need someone to release the closed source models, whether through legal or illegal means. Then we will see the same type of jump towards the impossible we saw with Stable Diffusion and LLMs.
There it is boys.

It probably won't replace vfx teams completely, at least not within the next few years, but I can see this becoming an assistant tool that goes through a lot of the grunt work and make teams significantly smaller. OR they find ways to make it create higher quality work and somehow make teams even bigger. Who knows.

Using models, characters, textures, and software by Valve that doesn't belong to you? Perfectly fine.

Using a voice that doesn't belong to you? How DARE YOU. You CANNOT do that. You need to PAY that person to say a few words into a microphone.

Like the PaleMoonLight user said, nobody cared when they were taking the actual voice lines and splicing them together. That was directly taking that person's exact voice and altering it to say other things, and it was considered fine. But now, it used "AI" and so it is bad, that's it, no other thought required. Total knee jerk reaction.
I think I mentioned this before, but photobashing is widely accepted as an image creation process, especially in the professional arts world. Lots of them sperged about AI and copyright. Funny how that works huh?
 
View attachment 5683289
Tweet
TIME.
average normal person gets around 8 hours of sleep and 8 hours of work. thats roughly 8 hours of free time, which is very generous. even if we cut it down to 4, that's still a decent chunk of time to do whatever. You could totally dedicate 30 minutes to an hour of learning how to draw and get good at it
 
I've gained even more contempt for anti-AI artfags with all their incessant gabbing and hysterical nonsense, but you guys are seriously retarded about learning to draw. People like to say they don't have time, but a million to one times it's just them being unwilling to put in the effort and making excuses. What are you? Some sort of single mother negress supporting 5 children and taking three jobs? Are you working in a North Korean workcamp and given no breaks or you'll get the bullet? No? Then you have more than enough opportunities to learn. Schedule time, doodle during break time, substitute activities, whatever. Though, RyaiArt is a faggot and I've never heard of any AI users complain about gatekeeping and a lack of resources(lol).
 
Back