So, brought on by this thread:
https://hexbear.net/post/8986189?scrollToComments=false
And @EdlritchEconomics@hexbear.net pointing out it’s a AI image; here is a whole thing.
So I think the general consensus is AI slop bad and we dont want it in this town.
AI slop is now a bit tricky and I wouldnt blame someone for being tricked and posting slop as if it were real because they were themselves mislead. I dont blame em, scrutinizing every image is an exhausting ordeal.
So my proposal, is that any AI slop presented as true should be removed. Posting AI slop to comment on AI slop seems fine and honest. I dont think anyone posting AI slop and presenting it as real should face consequences unless there is evidence of doing it on purpose but we cant have robot pictures pretending to be real pictures. We need Blade Runners.
I dunno if a task force or an if you see something say something method is best, but it’s just gonna get harder to spot for a while until the whole thing turns into a fun house mirror, so it seems worth getting in front of. You shouldn’t have to scroll the comments to find out a posted image is AI and I would prefer no image posted here be AI unless it’s meant to comment on that image as an AI image, illustrating a point about a slop image is one thing but if an image is AI I think it should just be removed but ill accept a tag being needed.
I picked Chat cause I dunno if anyone uses Feedback nowadays.
This also just made me think of photoshopping pictures to make them look like AI for plausible deniability/for a bit
You must log in or # to comment.
I think we should do your proposal because it would ultimately cause a massive amount of drama and hexbear hasn’t had a moderation struggle session lately. It’ll be like when hexbear renamed the_dunk_dank for the 4th time. Honestly it’s a very very bad idea and hexbear should do it because of that. It’s part of the hexbear character, making up moderation rules that don’t exist on any other website and enforcing those rules haphazardly.
Requiring all AI images to be removed from the website would create a mechanism for any post to be removed on a whim if accused of being AI. The chance of false positives is too high.
In 1997, IBM Deep Blue was the first computer to defeat a world chess champion at chess. The modern chess engine called ‘Stockfish’ is virtually unbeatable by any human, since like 2014. This same principle applies to AI image generation. 2 years ago, it was very easy to spot an AI generated image. 2 years from now, it will be virtually impossible, if not already.
Hexbear already has rule 13 which prevents spreading misinformation. Any AI image which spreads misinformation can already be removed by that rule.
There are also many instances where it could be important to post an AI image. For example, Trump posts AI images on truthsocial all day, every day. What if Trump posts an image of an explosion in Iran and says “This is a photo of the US bombing Iran”. Is hexbear not allowed to document the President of the imperial core using fake images to promote endless killing? This is not a hypothetical. This is something that Trump has done many times before. Trump posted AI images of the Iran women’s soccer team and said that Iran would kill the women for not singing Iran’s national anthem.
Forming an AI detection taskforce is an absurd proposition. Can you name a single website in the history of the internet which has created an AI image detection team? It doesn’t exist because it is irrational. It would be less irrational for hexbear to ban all image posts than it would be for hexbear to try to determine which images are AI. At least if hexbear banned all images, then it would be consistent and not based on misinformed guessing.
It would be more reasonable to require that non-OC images include a link to the original source of that image and then determine whether then say whether the source is a credible source or a non-credible source.
The primary purpose of hexbear should be documenting real things that happen. If a powerful figure purports an AI generated image to be real, that should be documented. Hexbear is a link aggregator website. The website is written to collect URLs of other websites. Reddit didn’t even allow image uploading until like over 10 years after it was formed.
Creating a punitive system rather than a preventative system is in itself reactionary thinking and anti-materialist. “We need Blade Runners.” is cop brain. It’s like pretending that you’re The Punisher and wearing Punisher Skull.
What would be more helpful is forming a team that teaches people how to post correctly. There’s not really anything that teaches people how to make posts and the rules have often been poorly documented.
Let’s look at a real world example of punitive thinking vs preventative thinking. Low income areas have higher rates of youth crime. Reactionaries will say something like that low income areas have more crime because low income people are just worse. Punitive thinking would say that this problem would be solved by having harsher punishments for children, maybe even trying them as an adult, because reactionaries believe that harsh punishments will deter crime.
Let’s look at the same situation with preventative thinking. If children in low income areas are given more life opportunities like free meals, afterschool programs, and access to higher education, then they are less likely to do crimes. If some teen is playing sports with a coach after school, then they will be less inclined to be shoplifting beers from the corner store.
If you don’t want people to make bad posts, then you should teach people how to make good posts. There should be places where people discuss the best posting methods and can ask for tips on how to post.
Read my post. Ylu are putting a lot of words in my mouth
Hexbear already has rule 13 which prevents spreading misinformation.
wut if it’s good misinformation
- Doctored images, fake social media posts, generated content, and satire/parody that appears real should be clearly marked as fake and/or posted in the designated communities.
rule 13 says that posts containing misinformation have to be clearly labeled as misinformation.
so it seems worth getting in front of
I want to get in front of a train
if somebody goes to chatgpt and say “make an image of a horse riding Stalin” and posts that to /c/art they should probably get banned. But AFAICT that kind of thing almost never happens?
pro genAI communists want to put genAI in our spaces because they know that other pro genAI spaces are deeply hostile to our ideology and the mainstream genAI companies have explicitly said they want us all dead. many will point to chinas use of AI but look me in the eye and tell me Xi Jinping is promoting AI anime girl chatbots.
I don’t like AI at all and agree with the sentiment.
But uh the Blade Runners weren’t the good guys in Blade Runner. They were escaped slave hunters. Deckard shoots a fleeing woman in the back in broad daylight and SA’s another in his apartment.
Yes. I understand movies. As well as jokes
Well I don’t understand either and I hate fun!
It turns out my silly metaphor that I didn’t mean to be a 1:1 parallel has confused a few people. I was just trying to be a bit silly.
I hate fun
how bout no one uses ai here. it has no place in person to person communication
This post doesn’t advocate for using AI, but for commenting on (usually bad) AI posts from reactionaries
i was drunk then; im drunk right now. i must have just saw this part of the post last night: “Posting AI slop to comment on AI slop seems fine and honest.” i see what they are saying now with a clearer mind
If it’s misrepresented as real, it absolutely should be removed and users who have a substantial history of making such posts should probably be assisted in re-evaluating their posting approach.
Agreed. I’ve ended out elaborating more extreme views in other replies im this thread that I may edit into my main post tomorrow but I gotta make supper and stuff rn
This would be a good thread if the original image was actually AI but it’s just a real photo taken with an iphone, they use AI upscaling now, which makes shit look like AI when it’s actually real. This really fucks with identifying AI
I’ll admit, that is a big wrinkles that i didnt know abkut at the time.
I now propose we mail each other Polaroids and hexbear become text post only bexause I fucking hate this world.
Yeah I’m little bit tipsy so please don’t ask me to elaborate but the AI upscaling on phone cameras makes me also feel like it’s a continuation (deliberate or otherwise) of poisoning the well
Probably. They probably see multiple benefits from it, for example adding more things to “stuff AI does” increases their stock growth, while poisoning the well is an added benefit.
We’re kinda fucked with identifying AI if laws mandating watermarks don’t happen.
WeChat kinda does it, short videos made with AI are required to mark themselves as such, and content suspected to be made with AI are also flagged. From my feed:
The former has been declared by the creator, the latter has been flagged as suspected AI content. I don’t know what programs these videos were made with, but both have added watermarks (含AI生成内容 ‘Contains material created with AI’ and 内容仅为AI创意 ‘This content is solely an AI creation’)
There are plenty of false positives and negatives, but it’s something.
We’re kinda fucked with identifying AI if laws mandating watermarks don’t happen.
The problem with that is it’s literally impossible, like mandating watermarking of all images touched by an image editor or of rendered CGI images. Open source local models running on open source hardware can never be watermarked because people can just make it not do that, and even slop from a compliant slop machine can just have the watermark denoised out by a non-compliant one or just be cropped for less intrusive ones.
Nah it’s not. The AI companies can be forced to do it and open source local models can be forced to do it too.
There will be a fringe of people who bypass these restriction, like there is a fringe of people who shoplift. It would massively improve the problem though and it would completely eliminate AI use from creative works in entertainment, like movies or videogames.
I don’t think you appreciate the global scope of AI grifter companies, like the endless parade of “unrestricted” image generation services that are just rebranding after payment processors shut them down and operating out of random countries to escape liability, renting servers and selling the outputs of whatever open source model and frontend server they feel like. Nor the degree to which the hobbyist scene will hammer away at any locked down model they get their hands on to rip out guardrails and make it spit out more gooner slop, which then just gets endlessly copied and passed around.
For that matter, we’re already so far past the point of extremely strong, scary, and trivial to run models proliferating that there really isn’t any chance for “well at least going forwards future things can be told to lock it down somehow and maybe someone will figure it out or something” to actually work. Things like Z-Image and Krea 2 are just horrifying, and then there’s Lightricks for video which is, uh, like watching a pod person try to imitate infomercial acting but it’s also still absolutely horrifying for all that it’s still offputting, useless dogshit.
Death penalty for non-compliance. Drone strike them.
I am 10% joking
The more im catching up the more over 100% im feeling.
no more half measures walter
If it’s purely for a shitpost, I personally don’t think it’s worth getting up in arms about
Edit: Also there have been updates in that thread: The exact pic may or may not be ai. but that specific meme car is indeed real
I think that needs some restrictions even. It should be tagged as AI, it cant be created by op and it needs to be mocking the image. No one should repost ai memes or jokes that they enjoyed. Any post containing an ai image or text should only be posted in mockery and be tagged as well.
The only thing more annoying than the absolute dorks who spam the weird dream their computer had everywhere like anyone cares are people who think they’re on some kind of righteous crusade against that sort of annoying and cringe behavior. The core debate should start and end at “no, we really don’t want to see the weird dream your computer had. Yes I’m sure you think it’s very funny but it’s really not,” to preclude a flood of low quality content akin to the way godawful CGI flooded everything back when Poser and Daz3d became widely accessible in the mid 00s, and then just not really giving a shit about things caught in the wild unless it’s explicitly misinformation (and a traditional photoshop used in the same way should be prohibited too).
I don’t think the average hexbear uses ai image generators, much less posts the results here. If ai stuff is ever posted here it’s almost always because the person didn’t know/didn’t care enough to investigate beforehand for a meme post. Nor should they have to imo. If it’s for a serious topic, sure, but like I said I’m singling out joke posts specifically.
Could you give a hypothetical? It can be pretty broad,.I just want to get a better understanding of where youre coming from before responding.
I guess what I’m getting at is: Unless it’s for a serious topic people shouldn’t be expected to check beforehand if an image was made by ai or not, especially if in the end it’s only for the sake of a gag. I’ve seen some people here get on other’s cases for inadvertently posting an ai image for a shitpost of theirs, and it felt like an overreaction to me.
I think I clarified a lot of this elsewhwre on the thread. Gonna make an edit tomorrow to the post with those in regsrd but its supper snd sleep time now.
And I edited my first post to reflect it, but there seems to have been updates in the post you referenced.
I got a lot of replies at the same time
I like the proposal. Not sure how to materially advance it but I agree something should be done to catch and remove this stuff.
I’m so scared of reaching the simulacrum comrades, I really don’t know how to manage that eventuality. I’m still holding out hope it all goes down in flames but I’m losing confidence in that option.
Transparency is what matters. If a poster is conned as well though, they need to edit as soon as it’s made known.
Trying to pass off slop as original content you created should be bannable imo.
But why allow any AI slop at all? Aside from reposting to laugh at? I get posting AI slop found elsewhere to make fun of and think thst is fine. But posters tricked by AI should have the post removed, with no hard feelings. Spotting it is also gonna be more difficult until the bubble bursts so I think it might be good to get a a Nerds Who Are Good At Spotting AI task force to nip things at the bud. I dont wanna punish anyone who is posting AI cause they were duped. It’s built to trick you. If intentionality is evident kill em. Any AI image post that isnt badically there to shit talk the ai image shouldnt be here and even then should get a tag
I think removal is fine for non-malicous posting of slop without a tag. Either tag it (and yes, agreed that reposting for mockery is fine) or having it removed or, if the attempt was malicious (“look at what I made”), ban of some amount.
I would prefer No Ai unless its a screenshot and in mockery policy and even mockery posts should be tagged. I don’t any AI on hexbear that isnt in that specific zone, personally. I don’t think even reposting ai content that is treated as positive or neutral should be allowed even if tagged. No slop zone.
dang it. I was hadMe too, until I checked the thread. That’s why I made this post! This can’t continue
Might want to check it again. EldritchEconomics was wrong and said so, it was not an AI image.
But yes, as for your proposal, it sounds good to me.
(@AntiOutsideAktion@hexbear.net you can rest easy, it seems you were not in fact had.)
Okay, we have already lost. We live in a hall if mirror and can’t trust anything ever and no one can stop it. Im not being mean or joking. This sucks
It is shocking to me too how much the world gets increasingly more ridiculous all the time, making it almost impossible to navigate the sickening absurdity, and even at different scales, from local interpersonal interactions on up. It does indeed suck.
I picked Chat cause I dunno if anyone uses Feedback nowadays.
Chat seems fine to me personally, especially if you want non-hexbears to see it, but !feedback@hexbear.net is theoretically for technical site issues (and is pretty low traffic, yeah), while !hexbear@hexbear.net is for meta / moderation discussion (and also relatively low traffic, yeah)
I’m in favor of your proposal.
There is no reliable way for the average person to determine what is or is not LLM or Generative Image Model content.
Yeah but there’s a difference between “method that includes all examples and excludes all non-examples” versus “did the user even look at it kind of closely for common tells”.
Once again. Dont want anyone facing consequences for not looking closely for common tells, just maybe we should have people looklng for common tells and removing the posts. A removal is not a ban or anything. It isnt a punishment, it’s designed to trick you. I’m just bringing up it may be at a point where it’s hard enough to tell, turns out the image in linked wasnt fully ai but a real image that was Ai upscaled and therefore ended out looking fake. I’m just hoping there are nerds who csn and are willing to try to keep.these streets clean as best they can.
Yeah, sure, though if someone posts a lot of AI shit, we should consider that they need to revise their approach.
