Estás usando un navegador desactualizado. Es posible que no muestre este u otros sitios web correctamente. Deberías actualizar o usar un navegador alternativo.
AI Derangement Syndrome / Anti-AI artists / Pro-AI technocultists / AI "debate" communities - The Natural Retardation in the Artificial Intelligence communities
It's probably because these people see that the "peaceful" ways of resistance, i.e. the campaigns for people to vote for Kamala Harris, and the Hong Kong Freedom protests, were stomped out, and thus people feel like they have to use the extreme ways out of "necessity".
Which is kind of inherently funny, since taking extreme measures won't accomplish anything either. Never 4get man who burned himself up over Palestine or something.
I don't actually hope for extreme measures to be taken, but that is where the lolz are found.
Like somebody firebombs what turns out to be the custodial shed where they keep the lawnmowers at OpenAI, is sent to prison for decades, ruining his life, and all the freaks start going "UH, UH, WE DISAVOW! NO VIOLENCE, EVEN THOUGH THAT'S WHAT WE'VE BEEN OPENLY ADVOCATING FOR YEARS!"
Letting the bar-to-entry lower and easier to access for anything will follow suit a bunch of low-effort creations, no shit there's a ton of AI slop because its easier for everyone to make images now. Much like how it was now way fucken easier for people to "draw" and make images with a bunch of assistants and tools like stabilizers, transforming, the paint bucket, etc. Or how once game engines like Unity and Unreal became the "goto" for many aspiring game devs, you'll be left with a bunch of shovelware and mobile apps filling the majority of Steam and appstore pages.
Is this a bad thing? Not inherently, as it's a neutral thing and what should be expected once the tools to help streamline the creative process becomes more accessible. But it's real funny these wholesome progressive "#AllArtIsEqual" types tell you not to use AI and that you should "pick up a pencil", implying that the bar-to-entry for creations should be higher and if you can't reach it based on their highly arbitrary expectations, then you don't deserve to make images.
"Oh you can't draw a cat so you have to resort to AI? Either pick up a pencil and learn or kill yourself AI bro "
If I told someone their drawing sucks ass and they don't deserve to make drawings, these people would call me an ableist fucking chudcel and that all art is about "self-expression" no matter how good or bad. But once someone uses AI, suddenly there's a threshold that must be met in order for these groups to deem if it's acceptable for you to make pictures or not. Yikes, how exceedingly gatekeep-y and unprogressive!
But it's real funny these wholesome progressive "#AllArtIsEqual" types tell you not to use AI and that you should "pick up a pencil", implying that the bar-to-entry for creations should be higher and if you can't reach it based on their highly arbitrary expectations, then you don't deserve to make images.
"Oh you can't draw a cat so you have to resort to AI? Either pick up a pencil and learn or kill yourself AI bro "
I think they're suffering from an inferiority complex, because they see someonee using a more advanced tool as a threat to their income\social position.
Thread tax:
Liberal niggo explains to you why you should only be in cohorts with random people who draw online instead of earning money. What did they expect, that she'd jive with them just for the "idea"?
Are people really this incensed about their beliefs that AI destroys the environment? Is it just that AI is so uniquely evil compared to everything else to merit this response? Have they been desensitized to years of believing that it's morally acceptable to murder people who disagree with you, and this just happened to be the cause de jour that cropped up after that belief reached critical mass?
my theory is that they've done it to themselves
news articles about ai cut out lots of important parts. the lefties look at these articles, and let their imagination run wild. that—along with supplementary fear mongering—solidifies their view of ai. according to them, ai is making people dumber. the data centers are also drinking up all the water, making the world hotter, killing people, and will probably end the world faster. protesting and peaceful action won't do anything, and there is no way out. you need to become the hero, and kill people. if you think that needless, senseless violence is not the way to go, then you're just a liberal.
this sense of urgency has reached it's peak right after kirk. more and more people are starting to believe that only violent revolutions bring change, no matter if there are more peaceful ways of dealing with it. this, combined with erroneous ai fears, is what we are seeing today.
this sense of urgency has reached it's peak right after kirk. more and more people are starting to believe that only violent revolutions bring change, no matter if there are more peaceful ways of dealing with it.
That is exactly the reason why I feel tha this violent revolution is only a matter of time. Ever since Donald Trump went down those stairs, the general population has been fed a daily dose of 'Trump is Hitler'. This has culminated into several assassination attempts. The spreading of this dose meant making a Slaweel Ryam of anyone who does not follow the woke, which included Charlie Kirk. His death ultimately proved that '''''resistance''''' ''''''''worked''''''''.
Then again, they could use the gigaloading technique or the subpoena technique.
this sense of urgency has reached it's peak right after kirk. more and more people are starting to believe that only violent revolutions bring change, no matter if there are more peaceful ways of dealing with it. this, combined with erroneous ai fears, is what we are seeing today.
That is exactly the reason why I feel tha this violent revolution is only a matter of time. Ever since Donald Trump went down those stairs, the general population has been fed a daily dose of 'Trump is Hitler'. This has culminated into several assassination attempts. The spreading of this dose meant making a Slaweel Ryam of anyone who does not follow the woke, which included Charlie Kirk. His death ultimately proved that '''''resistance''''' ''''''''worked''''''''.
Then again, they could use the gigaloading technique or the subpoena technique.
They must have also been inspired from the Indonesian and Nepali protests, and used the success of the Nepal ones as further proof that violence is the only way.
There's absolutely no realistic way to stop the trajectory of the environmental damage human civilization is causing in a truly meaningful way. Things like clean energy and general rules against pollution are good but more akin to band-aid solutions to slow the progress of the problem, so to speak. More avenues of research and scientific solutions are needed and that is something AI can and will help with. Technological evolution is not only for betterment, but important for the survival of our species and continued existence of human civilization.
I remember the same kinds of people being very hostile in a similar open way about nuclear energy, until it settled in that nuclear energy is a lot better for the environment than burning fossil fuels and that critique just kinda quietly stopped (after the damage was done).
It's also funny that this kind of critique always comes from the same kind of people who just gotta have the newest iThing every year. Why are there no campaigns from these people to make smartphones more sustainable and repairable and last for ten to twenty years? The technology is certainly there. Why aren't they protesting Apple and other companies who through their policies cause immense mountains of toxic trash and waste valuable resources? There aren't, because the new iThing is cool and that's just why I can't take these people seriously.
They must have also been inspired from the Indonesian and Nepali protests, and used the success of the Nepal ones as further proof that violence is the only way.
I remember the same kinds of people being very hostile in a similar open way about nuclear energy, until it settled in that nuclear energy is a lot better for the environment than burning fossil fuels and that critique just kinda quietly stopped (after the damage was done).
It's also funny that this kind of critique always comes from the same kind of people who just gotta have the newest iThing every year. Why are there no campaigns from these people to make smartphones more sustainable and repairable and last for ten to twenty years? The technology is certainly there. Why aren't they protesting Apple and other companies who through their policies cause immense mountains of toxic trash and waste valuable resources? There aren't, because the new iThing is cool and that's just why I can't take these people seriously.
It's also funny that this kind of critique always comes from the same kind of people who just gotta have the newest iThing every year. Why are there no campaigns from these people to make smartphones more sustainable and repairable and last for ten to twenty years? The technology is certainly there.
You can apply the same argument to any anti by pointing out that their streaming video and online gaming uses, oh no, data centers and a GPU. But those uses are okay because they like doing those things.
It's why they made up the retarded "consuming water" argument. Now the "good" data centers and GPUs (the ones they use for gaming) don't use water, but the "bad" data centers and GPUs (the ones used for AI) do.
Other than the complete nuclear annihilation of the third world of course. That and all the hyper green retards killing themselves and donating all their possessions to be sold off to make more nuclear powerplants.
I have to confess I love all the paranoia and witch hunting regarding ai art online. Normies and twitter art trannies are rapidly reaching the (correct) conclusion that digital art has almost zero artistic value
>your art is ai, provide a timelapse
>your timelapse is from procreate, you used ai
>provide a video of yourself drawing right now.
>the video doesn’t prove anything, you’re using AR to trace!
>your art is bad, you used ai
>your art is good, you used ai
>i personally don’t like you, you used ai
It never ends, it’s never enough, and it couldn’t happen to a more deserving group
I have to confess I love all the paranoia and witch hunting regarding ai art online. Normies and twitter art trannies are rapidly reaching the (correct) conclusion that digital art has almost zero artistic value
>your art is ai, provide a timelapse
>your timelapse is from procreate, you used ai
>provide a video of yourself drawing right now.
>the video doesn’t prove anything, you’re using AR to trace!
>your art is bad, you used ai
>your art is good, you used ai
>i personally don’t like you, you used ai
It never ends, it’s never enough, and it couldn’t happen to a more deserving group
>your art is slop, you used ai
>your art doesn't look drawn (i.e. 3d), you used ai
>your art was done too fast, you used ai
>your art has one detail wrong, you used ai
Truly endless. I mentioned it a while back, but it's really being reinforced now: People, in their quest to remain ignorant of the tech, have developed pretty severe blindspots. There are times where I just ask myself why? Why the fuck did you latch onto that as proof of AI, and dismiss the rest of it in the process?
Nevermind how much of the fanart on the internet is done by tourists just looking for clicks, of course they'll get things wrong.
You're being too logical. Fucking accuse anyone of using ai. Don't elaborate or explain why something is ai. It just is. You don't need to point out reasons why it's ai because it just is. Saying something was done too fast still leaves the possibility of it being real. Don't even pretend that the person is claiming to be posting human art, treat them as if they are posting ai art with 100% certainty.
Less 'that's too quick to be human' or 'there's a minor error' more 'wow I'm always amazed at how far technology has come' and 'cool what model did you use'.
>your art is slop, you used ai
>your art doesn't look drawn (i.e. 3d), you used ai
>your art was done too fast, you used ai
>your art has one detail wrong, you used ai
Truly endless. I mentioned it a while back, but it's really being reinforced now: People, in their quest to remain ignorant of the tech, have developed pretty severe blindspots. There are times where I just ask myself why? Why the fuck did you latch onto that as proof of AI, and dismiss the rest of it in the process?
Nevermind how much of the fanart on the internet is done by tourists just looking for clicks, of course they'll get things wrong.
One thing I love seeing that I forgot to mention is the very blatant questions meant to suss out ai use. Like “oh wow I love how you rendered the hands, can you show a timelapse of your process”
Nigga it’s more honorable to just outright accuse than to be a pussy about it
I always was interested in doing pixel art via AI, sadly diffusion models inherently don't really understand pixel grids or limited color palettes. I figured out a very involved workflow that goes way back 30ish years where I figured how to reduce colors to be able to display pictures on my then already old Amiga without losing to much information.
The original chipset Amiga has a 12-bit color palette (4096 colors) and can only display at maximum 32 colors at once in a normal way (there's also HAM, half brite and chipset tricks to do more but let's not get into that). This palette and depth limitation usually leads to a rather muted, earthen color tone palette I always really liked, very different from the bonbon and pastel colored games on the game consoles of the time that were more popular in the US.
Krea 2 is one of the first model that understands using a palette like this for pixel art. I can then do some magical mathematics on top to reduce the colors without losing information and put the image into an actual pixel grid. The picture below was Krea generated and has only 17 colors (feel free to check, its a png)
(the original image is 104x209 pixels, I blew it up by a factor of 6x so it'll be viewable on modern computers, if its blurry its your browser)
This image would actually easily be viewable on my Amiga, even the color palette is optimized for it.It's not real pixel art, but we are getting closer and closer.
I don't really have contact to that scene anymore but I wonder how the demo scene feels about AI art. if they hate it because of he usual untrue argument of "plagiarism machine" that'd be especially ironic considering plagiarism back in the day was part of the average work flow. (many a classic pixel art is just drawn images of actual artists carefully reproduced on the computer pixel by pixel, changing details here and there, credit usually never given)
Krea 2 is a very good model. I hadn't had much luck with loras yet though, either they come out over or undercooked. Not sure what the problem is but I feel very close to a solution.
This extension is really old by now (and unoptimized + full of bugs,) but the underlying model is good and will intelligently turn any image into pixel art in an information-aware way (unlike eg. nearest neighbor scaling which is info-blind.) It's sort of like how image codecs work. You can get your LLM of choice to implement it into whatever you're using. There may be a newer version of the model by now, haven't looked into it
yeah, that's the biggest problem in my solution. I can make good mathematical guesses about "pixel heaps" and regional clustering so it's not completely information unaware, but of course the algorithm doesn't really see the picture and will gladly reduce colors in areas that are the main focus of the image and probably shouldn't be reduced. Krea is the first model my solution actually works decently with because Krea 2 understands "limited color palette" in it's prompting and will actually keep the colors to somewhat of a minimum.
I'll take a look when I get around to it, thanks for the hint. I really want to train Krea 2 on some select pixel art I have and see if I can make it more consistent in color selection/general style. The model seems to have sort of a "wall" in learning that's hard to overcome and when you push too hard, it'll just cook the model. I guess it is the sheer size of it. Wonder if ideogram would fare better, but as cool as the json prompting is, it's hard to get aesthetically pleasing results out of that one. It really needs an LLM finetuned on generating prompts for it, IMO.
"
