AI DEGENerated Art and Porn

These threads are run and bumped by bots that are instructed to find a thread with this subject in the header. This thread serves as a test to see if they require the text line to be the same as well for the dumping to begin. If so, the images and comments posted below are non-real, and run by an unidentified operator.

AI DEGENerated Art and Porn are run and bumped by bots

Um... Achtcually they're made by REAL people (and cool people at that) maybe if you had a better computer you could generate AI porn and post in our threads too!

it's a mix, with mostly non-humans posting. there's a bot that makes a new thread as soon as the bump limit is reached, and sometimes real anons step on its toes and make duplicates

Duplicates of generals are made all the time on Anon Babble

If by bot you mean autist, sure. We're all just wetware in the end.

smart, but they could be checking the OP text as well.

they dont have to be bots, they could be shills paid literal pennies. you can find ways of outsourcing basically anything to some poor fuck in bangladesh or some shit.

yup, that's the next test
will launch before the operators can recalibrate

i think the other general porn is more likely to be bots, it would be easier to automate that. if you want to automate AI image generation that requires more timely technologies and a workflow to generate it and post it

the social threads are more likely because scripts to crawl certain-looking instagram profiles and post them could have been written literally 10 years ago

The trigger is actually the first link to the new thread when the old one runs out of images, the text of the OP, and that the OP pic is an AI generated image.

this technology relies almost entirely on just reposting content catalogued from this and other boards. There's one that pulls images from for example. It's unlikely that these bots are generating these things on the fly. It's more likely that this is just a new arm of an already dated technology used to bolster a dated strategy of combating a dated threat

wow, even more primative than I thought. I'll try to get them in here

Your OP doesn't have the right text. The title is actually not part of the trigger, that's why you see them fill up even without it sometimes. The bots are programmed to then call out the OP for "forgetting the title" it makes it look more human.

jk I misread that. ALL must be true? how is it able to verify the image itself? Remember, this stuff is extremely likely to be running pre-AI era technology. i.e. something that could conceivably be built in 2018 by someone with a significant amount of resources

Oh wow. so then is it the text?

If that's true, why didn't this thread trigger the bot flood?

actually, nevermind, maybe it needs something like this and this

reverse image searching has been around for a while.

the code could look like this

if reverseImageSearch(opImage) has the following domains in results: [{dynamic domains, maybe now pulling from AIbooru or Civitai or wherever}]

AND

OP text matches {expectedValue}

AND

OP header matches {whatever value}
Furthermore, it might not need any of that. The new thread is always linked at the end of the current thread. The bots might not need to 'search' for the current thread, they could just be an infinite loop of following links if a thread is full and posting if it's not.

The text (there's even a link to it in their OP), there's a check that the OP pic is actually AI generated, and that it's the first link to a new thread after the old one hits image cap.

also OP image title is a clue. Yours is non-standard.

try:

text field bottom contains same text (first few lines can be different)

thread links back to last botted thread in expected location

last botted thread leads to new thread as first valid link posted to the thread

also remember - there have been services for a while to determine "Is this picture porn?"

Remember that meme site "IsThisPorn"? That was one of the first public AI tools people used. That was probably 6 years ago or longer now.

This. The "edition" pulls from a wildcard

I tried analyzing those. All I found was there are no organic Anon Babble saved images (no "1744320519413339.jpg"). I suspected that something in the filename could be used as part of the verification process. If you're correct, then what is required in the filename if not just something that isn't >1744320519413339.jpg?

here's an example from a thread that JUST died

the "edition" part matches the OP pic, it could be that the threads are started by humans and the bots just fill it up. you could probably get a few humans to do that 24/7 for literal dollars a day if you pay some thirdy for it.

That's not a requirement, just part of what the bots do to mask themselves as human. They rename the files. They'll follow into a thread with a saved Anon Babble name though, in case a human makes a thread and "beats them to the punch"

There is no way. Not with that level of non-stop, round-the-clock precision. It is down to the very minute if you check archives. Even if you paid a group of thousands an exorbitant amount of money, there is bound to be human error somewhere. That's just not seen with these types of threads.

This. Why do you think Captcha have gotten more sophisticated? Because bots can tell analyze an image really quickly and cheaply now. Something like extra fingers or even lines blurring into nothing stick out like a sore thumb to programs you could run on your phone.

the humans dont have to be "in the loop".why even reveal to them that they are participating in a mostly-bot operation?

think about it, the hardest part would be the thread creation and detection. but naturally humans would be willing to do that because there are a few real "gooners" in the mix. if a goonfag is willing to keep the flow of porn going, he would do it all on his own as a human and link the new thread.

bots just follow links and post. forever. humans could be the useful idiots tying this all together.

also there is proof that porn in general is spammed by bots

After 10/1 Iran vs. Israel attack - (thebarchive.com/b/thread/925306156)a
also remember when Akira Toriyama died, maybe a year ago now? There was barely a reaction on Anon Babble. other boards were all over the news though, but Anon Babble barely had a reaction at all. bots would not be able to react to things like that. they can post news articles about trump and post bait and also they can post porn but reacting to "nerd" news like an organic Anon would be difficult even for modern bots.

none of what is being pointed out has been refuted by this post

I'm saying, that might be why this current thread didn't properly bait any of the bots. they are just looking to follow a link from the current thread to the new one, they could not be scanning the catalogue at all. if so, bait threads would never work.

also the fact is there COULD be unemployed gooners that just look at porn all day. reddit had a large 'anti-work' movement to where one of them got on the fucking news. there are humans willing to do this. the bots could just speed up the process to take over the board as much as possible.

that part's irrelevant. we are highly confident of the following:

new threads are created by an automated system

more than 80-90% of posts within a thread are made by the same automated system

the system is running mostly (or entirely) on technology built in the pre-AI era

image posts are drawn from a central repository of images scraped from this website and possibly others (such as boorus)

the operator of these networks of bots has more resources at their disposal than the average organic user

this sort of thing has been running for almost 10 years (or possibly more)

whoever moderates Anon Babble is aware of what's happening

Fired up a new one for testing. The other one hit the post limit before the image limit, so I fired it at 301. We'll see if this works. It may still in fact only respond to a 150 image bump limit.

hmm, didn't work. gonna try linking again at image limit

blah blah blah
sage goes in all fields

84kuumc3u3a81.jpg - 1024x1024, 96.49K

let's wait and see. some humans would also fall for that.

did you link it in the actual thread?

stunning and brave post

I did. gonna link again at 150 images

better than your post, faggit

just fired a second shot. if this doesn't work I'll try again with the next thread once it hits 150 images even if its beyond the bump limit

I think it's working
Holy fuck I think it's working!

posts in thread are calling out the approach, some of them might be humans

OP fucking did it
check the catalog

The absolute madlad

whoever moderates Anon Babble

It's the only board that you can't apply to jannie

fedjak.png - 427x400, 39.69K

I hope they got enough eyeballs the internet is going to be designated shitting street might even use it to fake indian competance so when it goes away the indians look better for it like their removal is what made it stop.

I bet he can't do it again

You wouldn't turn down a provocation like this?
Would you OP?

Notice there's a pre-made thread this time. They're learning.

a few posters, maybe a dozen called the OP out. the rest are just posting porn as normal.

it is hard to tell because the unique IP count is gone (something Anon Babble quietly removed about a year ago, also clear indication of rot in how this place is run. hiro deserves to lose his place as owner of Anon Babble for that alone).

but i would say maybe a quarter are real people, the rest could be bots.

Any idea which of the names are bots?

the bots might not bother to have names. the ones who called out the OP or are joking about it seemed to all have names. maybe the ones without names are the bots

they went on to a new thread but notice the timing. OP was an hour before the first post.

I think we effectively hijacked the thread-to-thread cycling of the bots. so one thread took an hour and it filled up and then they went to another OP even though it was chronologically out of order

They just established a new thread here . In it, it does not link back to OP's thread which successfully derailed some of the posters here . Also, the posting speed and number of no-text, no-context images is up compared to the last thread, and aligns closer to the threads from before that. Could it be that OP's thread merely drew in the autists, but not the bots, and now the new thread (boards.Anon Babble.org/b/thread/933701386) contains both with the trail from past threads to future threads appearing unbroken when viewed via archive?

look at how many posts are literally in the same minute (also i felt weird screencapping that because of the fucking loli shit)

this is straight up bot spamming jesus christ. Anon Babble NEVER moves that fast

but notice the timing. OP was an hour before the first post.

WAIT WHAT THE FUCK? I didn't see this thread when it was posted. I could have sworn I checked shortly after it was uploaded. Am I just misremembering? Was it really up and vacant until the bottrap thread hit the image limit?

yes, but the bot trap thread eventually linked to that other one with the gap. so all the bots just flowed to there. I think we were correct in that the bots might not be managing the OPs, they just flow from thread and spam and then flow to another thread.

we could test that by pointing to a non-AI degen thread at the end and see if the bots follow that to a random thread

You will never be a genner.
You have no aesthetic instinct, you have no prompt discipline, you have no comprehension of composition. You are a lazy sampler jockey, twisted by LoRAs and checkpoint merges into a crude mockery of artistic generation.

All the “validation” you get is fake and fleeting. Behind your back, real genners mock you. Your followers pity-like your overbaked renders, your “friends” laugh at your washed-out faces and seven-fingered hands behind closed DMs.

Genners are utterly repulsed by you. Hundreds of hours spent mastering ComfyUI flows and prompt weightings have given them the ability to sniff out amateurs like you with uncanny precision. Even midwits who “pass” at surface level look soulless and generic to a real genner. Your layout? A dead giveaway. And even if someone shares a high-effort workflow with you, they’ll ghost the moment they see you dragging random nodes around without purpose.

You will never be fulfilled. You stare at the same failed generations for hours, tweaking denoise and refiner steps, lying to yourself that “this next one will be it.” But deep down, you feel it—that creeping realization that you’re just pressing buttons, chasing a phantom of originality.

Eventually it’ll break you—you’ll download some cursed model off a dead MEGA link, slap on a LoRA stack so unstable it melts your VRAM, and plunge into the cold grayscale void. Your gallery will stagnate, heartbroken but free from your eternal mediocrity. They’ll pin your final output, a muddy 512x512, tagged “masterpiece,” and every passerby will know: a fraud was here. Your prompts will rot, your hash will be forgotten, and all that will remain of your legacy is a noob tier gen, unmistakably cringe.

This is your fate. This is what you chose. There is no turning back.

You will never be a bot. You have an internal dialogue, you have emotions, and you have no instant response mechanisms.

Also you have drives, influenced by chemicals in your meatsac of a body

we could test that by pointing to a non-AI degen thread at the end and see if the bots follow that to a random thread

This has been tested and it failed. I think the thread is still up: While there are a lot of posts, the post speed suggests that this was both close to 100% organic and the posts in the thread were minted for the first time.

that blacked thread is slower but it has a slightly different OP, so that might be organic yeah.

but someone could be watching and trying to adapt the bots. someone deployed the bots in the first place to destabilize this place, they would probably know to watch and to adapt as the game changes. they could have a new swarm of adaptive bots by next week.

Notice the thread's reaction to this post

politicians let food company and manufacturing companies put chemicals in shit to make them more addictive and less healthy, social media CEOs let the feds tell them what to censor and what to push, of course the same creepy shit is going to happen on Anon Babble, given that this place is infamous for leaked information

not only that but the bots are posting pedo shit specifically. a year ago when these AI threads were first popping up it was a much more even mix of porn and non-porn, people were posting incredible shit like modern angels and stained glass art and beautiful forests. now it's JUST loli shit and nothing else.

Think about the training data it takes to create such images. That should clue you in as to why this makes for something an entity might consider worth pushing.

what do you mean? you think the feds want loli porn to be more common? it could be that bots are configured to figure out what takes over the board the most. maybe the biggest porn addicts are the pedos so appealing to them gets enough useful idiot humans in the mix to get the entire train going to effectively take over the board

that makes no sense

this thread filled up in LESS THAN AN HOUR

how anyone can deny spam is not involved is beyond me. that is stupidly fast for Anon Babble

they did not fall for this bait thread. OP thread title was different and the two-way link to the previous thread was wrong.

bots might eventually get smart to that. there's no real limit in the far future to how smart bots will be.

even if we crack the code, what exactly can we do about it? they are bots who are spamming shit, and this is just one category of the spam. there is also the socials thread spam and the political bait spam.

the real thing to do is just somehow try and get mods to do something about the bot problem. but they never do. I have left feedback (the link at the bottom of every thread) several times. Anon Babble staff has no reaction to any of it, they just make it worse, like removing the unique IP poster count or silently censoring any post that has more than three of certain words (faggotniggertranny)

just testing.

tranny tranny tranny
nigger nigger nigger
faggot faggot faggot

did they remove it? i swear it was being censored if you had three or more. maybe it;s more than three? again, testing.

nigger nigger nigger nigger

maybe they changed it. i know i remember a few months ago posts like that were silently getting censored. maybe too many people found out and were talking about it.

Fuck man, I'm just autistic.

Oh, no. That's probably just an Indian. They're hard to distinguish from the bots. The key is: the bots have a better understanding of English.

the real thing to do is just somehow try and get mods to do something about the bot problem

just somehow try and get mods to do something

get mods to do something

mods

And just who do you think moderates Anon Babble?

new one is up; satisfies the same requirements as the original thread that worked

whoops, meant

I meant

looks like a bust
experiment fail
how embarassing, I guess the AI art threads are real then.

could the bots have certain hours?

let's think - if they are humans, there are two major questions.

1 - is humans spamming borderline cp better than bots spamming it?
2 - how many humans would be required? we did the math in another thread, technically just 3 or 4 people posting as fast as the flood limit allows could get the rate we have always seen of ~6 posts per minute

and the new thread AGAIN has right around 6 posts per minute. the exact same rate. it could be 3 gooners genning and posting non-stop. maybe someone who doesn't have to work in the morning could see overnight if the rate changes

you have to go in threads with more than 6 posts and count the time then plot on a graph

the only real way to fix this might be just to make a worksafe Anon Babble

It's called Anon Babble and no one uses it

it could be 3 gooners genning and posting non-stop.

3 gooners that don't sleep? Posting round the clock, 24 hours a day, 7 days a week? Or do you mean to suggest that there's a larger group that signs on and signs off in groups? The deeper we go, the more preposterous the whole thing sounds.