30
submitted 1 week ago* (last edited 1 week ago) by dead@hexbear.net to c/feedback@hexbear.net

I am going to preface this by saying that this is not a dispute with a moderator. I am not asking for my post to be undeleted. I will write the post from the perspective that Hexbear moderators are correct.

I am writing this post for clarification on rules of future posts.

I like to post news stories which are published by DPRK public news agencies ie DPRK state-owned media, usually KCNA. I started doing this because I saw many misinformation posts about DPRK, whether good or bad. Many of the misinformation posts about DPRK cited DPRK public news and misinterpreted the article. So I figure, I should just start posting DPRK public news directly to demystify people's ideas of DPRK.

Examples of misinformation about DPRK:

  • Anti-DPRK sites which may be secretly funded by US government, such as DailyNK and NKNEWS. These sites make up bullshit about DPRK and are considered legimate by western media because their sole purpose is making bullshit.
  • Some western media articles, such as published by Fox News or BBC. These articles will typically cite the Anti-DPRK sites or misinterpret information published by real DPRK sites.
  • Screenshots of twitter posts with uncited sources: "Kim Jong Un said DPRK executed 2 israelis" (fake), "Kim Jong said Netanyahu is bad" (kinda real, DPRK has hated Israel since at least the 1980s), "Kim Jong Un said US is a cancer-like entity" (real quote from a much longer speech)
  • Collections of images of DPRK posted without context. There are many twitter posters who seem to get images from KCNA articles and then remove the watermark and post the images without the article, so that they can take credit for the images which were published by DPRK media.

I think that News Agencies inside of DPRK are a much more reliable source of information about DPRK than outside sources. Yet, I never see other users posting links to the News Agencies that are actually inside DPRK. Other users tend post things from the other 4 categories which I mentioned.

Do hexbear users believe that DPRK is a reliable source of information about DPRK or not? Are they afraid to visit DPRK-hosted website? Do they not know that DPRK has internet and websites? Are the DPRK-hosted websites blocked in your country?


Again, I'm not disputing any decision of Hexbear mods. I'm not asking for the post to be undeleted. I want clarification for future posts, which I'm sure would also be helpful to other users.

Yesterday I posted an article from DPRK-owned media with 2 photos by DPRK government and 1 photo by the Russian government. My post was removed because Hexbear believes that the photo which was provided by DPRK government is AI-generated.

The photo which is suspected of being AI-generated was broadcast on DPRK National Television (KCTV), published in WPK's official newspaper (digital & physical) (Rodong Sinmun), and published on DPRK's publical owned news agency (KCNA).

The reason given for the photo being suspected of being AI is that it looks weird. I can't disagree with that. Many photos that come out of DPRK look weird, even before generative-AI was a norm. DPRK just looks weird in general which I ascribe to culture shock.

Hexbear removed the post, saying that the DPRK government used AI to generate the image. This means that DPRK government published an AI-generated image in a physical newspaper and broadcast it on National Television.

exhibit 0) My post which was removed for containing a photo produced by DPRK
https://hexbear.net/post/8321242

exhibit 1) Photo published on KCNA, DPRK's nationally-owned news agency.
https://archive.is/xepRu

exhibit 2) Photo published on Rodong Sinmun, DPRK WPK's official newspaper.
https://archive.is/uvSei

exhibit 3) Photo broadcast on KCTV, DPRK's national television channel. Screenshot 1 shows the image plainly. Screenshot 2 & 3 shows the image as part of a newspaper, which I would guess to be Rodong Sinmun.
video (timestamp 13:15 and 25:30) https://kcnawatch.org/kctv-archive/69ea2614e62a8/
screenshot 1) https://hexbear.net/pictrs/image/553ce8ba-c7fb-4529-8687-129dd84cc347.png
screenshot 2) https://hexbear.net/pictrs/image/bf839602-ef40-466d-ab59-13943b6ea932.png
screenshot 3) https://hexbear.net/pictrs/image/c2becb72-b6b0-4085-b4ab-0a512296911b.png

https://en.wikipedia.org/wiki/Korean_Central_News_Agency
https://en.wikipedia.org/wiki/Rodong_Sinmun
https://en.wikipedia.org/wiki/Korean_Central_Television


Here's my concern, AI-generated does not mean "looks weird". AI-generated means that the image was not produced with a physical camera. AI-generated images could look totally normally, just as photos made with a camera could look weird.

My post was removed. I am being told that I broke a rule by posting an image that was created by the DPRK government and seemingly reported by DPRK as being a legitimate image.

Now I have to be concerned about whether photos posted by DPRK government are considered to be legimate by Hexbear, which seems to be the greater authority on what actually happens in DPRK.

For clarification on my future posts, I need to have some questions answered.

  • Does Hexbear consider DPRK to be a reliable source of information?
  • Will my posts in the future be removed for containing images published by DPRK government?
  • If Hexbear considers images which look weird to be AI-generated. How do we know that images that don't look weird are not AI-generated?
  • Will Hexbear users face consequences for posting information provided by DPRK government because it looks weird?

These questions could also apply to other governments.

top 50 comments
sorted by: hot top new old
[-] jack@hexbear.net 32 points 1 week ago* (last edited 1 week ago)

Generally yes, but that first image is definitely AI. It's meant to be a vibes-inspiring photo.

Don't think the post should've been removed though because there are plenty of real photos and the news is obviously real.

[-] StillNoLeftLeft@hexbear.net 8 points 1 week ago

Yeah it's clearly there to highlight the co-operation.

[-] Demifriend@hexbear.net 28 points 1 week ago

Why are you treating it as if it is all or nothing? The DPRK government and DPRK state-owned media are obviously capable of misinformation, same as any other authority. I would consider them to be reliable and trustworthy, but that does not mean they are infallible or that we should be uncritical of what they say. You also didn't engage with the multiple people saying to you that the image in question being AI generated does not necessarily mean that the DPRK government is trying to deceive people, though the post may have been removed before you got the chance. I disagree with the removal of the post, and also dislike that misinformation that is unsourced or from untrustworthy sources is posted here on Hexbear just because it sometimes says things people want to hear, so I hope you will keep posting news from official DPRK sources, but refusing to engage with news media in a critical way because its source is generally trustworthy and from a state we support is no way for a communist to act.

[-] Abracadaniel@hexbear.net 25 points 1 week ago

Generally I'd say we do yeah.

That photo is obviously AI generated, you're the one insisting that fact has broad implications and crashing out over it.

Your post and your comments didn't deserve to be removed. The mods here will often come in after an argument and remove the comments of the person that was "wrong" even if they weren't breaking any rules or being reactionary. The mods just decide you're not allowed to post because you had strong feelings and were in the minority of an argument. It's fucked. It's like a a parody of communists, literally 1984

[-] spectre@hexbear.net 11 points 1 week ago

IMO even "being reactionary" needs to be allowed to stand for longer. Let people dunk on the chuds, it's fine! Ban the account after a bit, but even then I don't think its necessary to remove the posts for having dorkass opinions.

None of what I said above relates to this situation, which is even worse that a known user of this site was over-moderated for being wrong about something. I don't see how that is rule breaking?

Yeah i agree like theres a difference between being cringe and wrong and being, like, harmful?

[-] spectre@hexbear.net 10 points 1 week ago* (last edited 1 week ago)

I know some may disagree, but even if the content is harmful (but maybe not egregiously so, that's all subjective of course), it feels like an insult to my intelligence when its removed. Like Abra said, it comes off as a parody of "communism big brother 1984".

Again, some may disagree with my personal threshold and maybe we should hash that out elsewhere, because the incident that initiated this discussion isn't even close!

[-] Kuori@hexbear.net 9 points 1 week ago

why should we host content that's harmful to our userbase on our own instance? it's not like you can't check the modlog if you need to indulge in bile fascination (i do this)

[-] Acute_Engles@hexbear.net 23 points 1 week ago* (last edited 1 week ago)

I don't think the whole post should have been removed but come on

[-] starkillerfish@hexbear.net 19 points 1 week ago

it is a generated image, but i disagree with the removal of the post. it was a real story with other real images included

[-] TheoryofChange@hexbear.net 17 points 1 week ago

The image is patently obviously AI, that is proven by the impossible angle of the bridge segments among MANY other things. It was used as an illustration by the North Korean journalist, and shouldn't be interpreted as a real image, which is a trend amongst reporters of all nations. Most folks at hexbear don't support this practice but it doesn't make the article inherently unreliable in this context

load more comments (4 replies)
[-] NinaPasadena@hexbear.net 16 points 1 week ago

I think weather the image is AI doesn't matter to me. The post should not have been removed because it was reporting on real and positive (which is nice since so much negative shit happens constantly). The news was reported from a reliable source... The official dprk channels.

If the dprk wants to illustrate their article with AI, weird photos, or something in between such a highly edited photo I don't mind at all. I'd simply scroll to the photos I found more interesting. Which I did on that post and really enjoyed.

It was a good post

[-] Lemmygradwontallowme@hexbear.net 16 points 1 week ago* (last edited 1 week ago)

This seems like a motte and bailey argument. You obfuscate the fact of the media containing AI with the media's writer and publisher itself, the DPRK's people

Just acknowledge, 'my bad, there may have been AI on the article, though not all, I'll put a CW for anyone who doesn't like it', that's the least you can do.

The article itself is fine, despite the mild AI usage.

[-] dead@hexbear.net 2 points 1 week ago* (last edited 1 week ago)

I explained pretty clearly what my concern is. My concern is that future posts of KCNA will be removed on a whim.

I'm not trying to argue with people about what the contents of the photo are. I did not try to argue with people yesterday about the contents of the image. I said that I have not known the DPRK to post ai-generated images and that the image was published by DPRK. Then I stopped posting in the thread because I didn't want to argue with people.

I'm choosing to take an agnostic position. This also my position yesterday. I said that sometimes normal photos just look weird. Could it be AI-generated? Maybe. Was the image created by the DPRK government? Yes.

Let's say Trump makes an announcement and his announcement has an AI image attached to it. Does the Trump announcement get removed for having an AI image? Of course the post doesn't get removed.

I want to continue posting articles from KCNA and I don't want my future posts to be removed. I want to know what I have to do to keep posting KCNA articles without having my posts removed. The only reason that I made this thread is because i want to know what I will prevent my posts from being removed in the future.

[-] Drewfro66@lemmygrad.ml 13 points 1 week ago

The DPRK's official media is a reliable source. You're not asking the question in good faith - you're salty a mod deleted your post and reading way too much into it.

An article from the DPRK broke site rules by having AI-generated content. A mod decided to remove it. If you believe this was a bad move, the correct method of resolution is to do the exact thing you're constantly proclaiming you're not doing, which is to ask for an appeal on the decision.

Moderators are, generally, just people doing their job. And just like sometimes at work I'll trim a steak and it looks like shit and I just toss it in a tray and say "Fuck it, ship it, someone will buy it on markdown in four days", sometimes moderators will remove a post they shouldn't or not remove a post they should and it's just not that big of a deal.

[-] JoeByeThen@hexbear.net 13 points 1 week ago

Hexbear seems to think AI bad, but makes exceptions for Iran because Iran is in the middle of war with US. We're not allowed to use AI to do our own Class War agitprop because the average Hexbears don't take Class War seriously let alone see themselves as actually resisting. Or at least that's my current running theory.

fanon

Personally, I give far more faith to the DPRK and other AES media than I would to western media. A single AI image is no reason to delete the post.

[-] plinky@hexbear.net 16 points 1 week ago

It’s more meant to safeguard slopification of the website, i can generate dozens of images i find funny/educational but their worth, especially if they are posted constantly is just nil.

[-] JoeByeThen@hexbear.net 6 points 1 week ago

Yeah that argument doesn't work so great when we've got bad posting and they once gave Melania an anti-union comm.

[-] plinky@hexbear.net 8 points 1 week ago

which aren't infested with ai, but more idiosyncratic posters and bits.

[-] JoeByeThen@hexbear.net 3 points 1 week ago

Well as long as it's human made trash I guess that's still quality. soviet-playful

[-] plinky@hexbear.net 5 points 1 week ago* (last edited 1 week ago)

it's at least intentional and interesting, quality is in the eyes of beholder innit.

returning to point of topic starter, ai generated image of workers meeting at a bridge would be much more interesting as either stylized socialist poster (how do they marry symbols of juche and russian liberal dipshit flag, what poses are workers in etc) or just as a real foto. ai doesn't pose people with any intention, it just does slop. (and don't tell me they used deep pose there, they obviously didn't)

use depth estimator to make 100 designs of your room or whisper to automate home assistant, fine whatever, or make a bot triggering chuds on twitter, it's either self-serving or hostile action to engage other brain in time wasting.

load more comments (3 replies)
[-] jack@hexbear.net 15 points 1 week ago

're not allowed to use AI to do our own Class War agitprop because the average Hexbears don't take Class War seriously let alone see themselves as actually resisting. Or at least that's my current running theory.

It's not useful with an American audience who largely fucking hate it

[-] TraschcanOfIdeology@hexbear.net 8 points 1 week ago* (last edited 1 week ago)

It's not useful with an American audience who largely fucking hate it

Seems to be working for nazis and chuds. Agitating on social media already seems futile to me, but let's not act like a lot of people have lost their mind because AI slop tells them to hate immigrants and trans people or else shrimp jesus will not get 1000 likes.

We can argue all we want about how AI slop just affirms their prior beliefs and that's why it works, and why AI slop isn't good for deprogramming people, but it does get a rise out of people, enough to make them engage with the ragebait.

[-] jack@hexbear.net 12 points 1 week ago

We can argue all we want about how AI slop just affirms their prior beliefs and that's why it works, and why AI slop isn't good for deprogramming people

Don't even see why we'd need to argue about that since it's obviously true. We can't use all the same tools our enemies get to because we have different objectives.

[-] JoeByeThen@hexbear.net 4 points 1 week ago

Lmao. And yet...

duck-dance 🎶Your government is run by pedophiles.🎶 duck-dance

Might wanna start squaring with those contradictions running around out there.

[-] jack@hexbear.net 7 points 1 week ago

What is your point? If you want to reach people potentially sympathetic to socialism in the US, those are exactly the people who are most opposed to AI. It is counterproductive. Obviously Iran has a very different relationship the technology and a very different purpose with it.

[-] JoeByeThen@hexbear.net 3 points 1 week ago

Lol the same people "hating ai" are the same people calling these bangers. And then your average normie doesn't give a fuck about ai one way or the other. They just don't want it shoved into everything. They're still using chatgpt for free and forwarding images of animals doing impossible things to each other. You go on insta to some vid of a person being hugged by a panda bear and the handful of people pointing out it's AI are getting dog piled.

Obviously Iran has a very different relationship the technology and a very different purpose with it.

A very different purpose? Do they though? Because then you and I have very different goals. And I point you back to my original comment and theory.

Frankly, the general consensus of how hexbear talks about not using AI for agitprop is like libs talking about why the West can't have high speed rail. There are hurdles and then are examples out there defying them, and y'all keep making excuses rather than adjusting to reality.

[-] tocopherol@hexbear.net 9 points 1 week ago* (last edited 1 week ago)

Some anti-imperialists enjoy Iran's use of AI because they are using the empires tools against them, it's like a taste of their own medicine after all the generated garbage posted by the US gov, AI generated images posted by western leftists wouldn't have the same effect.

But one big reason we should discourage AI use right now is just because of the massive environmental impact, to be consistent with our ideals of sustainability it should be discouraged, and most AI generated stuff looks like shit and irritates the people who would listen to a communist message. Instagram is not a good metric for what people actually care about or what effective agitprop would be anyway, it's almost completely inorganic.

[-] JoeByeThen@hexbear.net 2 points 1 week ago

Some anti-imperialists enjoy Iran's use of AI because they are using the empires tools against them, it's like a taste of their own medicine after all the generated garbage posted by the US gov,

Yeah, that was what I was implying, which continues into the rest of my comment.

AI generated images posted by western leftists wouldn't have the same effect.

Again, high speed rails. It wouldn't hit the same is not a legitimate reason, that's a skill issue.

But one big reason we should discourage AI use right now is just because of the massive environmental impact, to be consistent with our ideals of sustainability it should be discouraged

Absolute can-excuse-1 shit. You can excuse all the genocide minerals and existing data centers that have been making people's lives hell for decades to partake in this digital hellscape we call the internet but you're gonna draw the line at a massive force multiplier? And I mean ABSOLUTELY MASSIVE, they're pumping out a Lego video almost every 24 hours now, I think? Deeply unserious. can-excuse-2

most AI generated stuff looks like shit and irritates the people who would listen to a communist message.

Again, skill issue. Which is not gonna be overcome by sticking our heads in the sand.

Instagram is not a good metric for what people actually care about or what effective agitprop would be anyway, it's almost completely inorganic.

Blah, neither is hexbear a good take on leftists. I live in a red state, I'm surrounded by libs and Nazis same-picture , they don't give a shit about AI.

[-] thefunkycomitatus@hexbear.net 10 points 1 week ago

You have completely ceded to the techbro view of the world, that social media is the defactor mediator for social interaction, even in meatspace, and is where social/political influence is. But that then makes you wonder how people were able to socialize and carry out politics before social media, even in my lifetime.

Right wing AI slop gets engagement because the platforms are designed to do that. Engagement sounds like a really cool thing. We want people to be engaged in the socialist project. But it's a misnomer because it's not engagement in practical activity, in the Marxist sense. It's a simulation. It's voting, subscribing, views, clicks, replies. To a techbro those are the metrics of social interaction. They have that view because it makes them money. The platforms they created to privatize social interaction rely on those metrics. That doesn't mean that you, as a Marxist or normal human being have to accept those terms. In order for them to maintain control they have to convince people that their version of engagement is real and important and at the center of human nature itself. Hence why they push it as social media, as some cultural aspect of human nature where we come together and communicate. It's so ingrained at this point that it actually now mediates meatspace interaction as well. The current regime now carries out political action in response to online grudges. They do this because they have financial and social ties to the owners of public discourse and the owners need to make it all the more real by creating legal consequences that they control.

The answer here isn't to beat them at their own game. It's not to make our own social media where we get to be mods and admin and updoot our own AI memes and do raids like we're 14 year olds from 2005. The answer is to get people to stop using these platforms. Get people to log off except for vital communication, because the internet as a steadfast and long-distance means of communication is genuinely good. The more people log off and re-establish meatspace social ties, the more social media will be seen as weird and off putting. These sites will implode and it won't matter who owns the AI slop machines because it would be like owning a machine that produces noxious farts while everyone enjoys fresh air. The trick they're trying to pull right now is to get everyone to accept AI as the defacto means of producing content, as if nobody could ever create anything without it. Part of that is also creating fomo (which you are here demonstrating) and treating it like some kind of political necessity. If you put your political ambitions towards getting people to log off, you don't have to worry about the AI meme arms race.

"But what if people will never log off" Then we all die. It's the same answer as if people can never be organized. Do you really think it's just some "capitalist realism" thing where logging off is impossible? It's kind of self-fulfilling right? If you just assume you can't pursuede people to log off, and people don't, you lose social interactions in meatspace, which forces you to rely on social media more, which makes alienation worse. It's got to break somewhere. Might as well start with getting people offline. If capital is completely invested in turning me into a pod person, I should probably try to resist that somewhat rather than accept it and consider doing everything from inside a pod.

[-] JoeByeThen@hexbear.net 2 points 1 week ago

Lead the way, comrade who just made a shitload of assumptions. stalin-approval

[-] tocopherol@hexbear.net 2 points 1 week ago* (last edited 1 week ago)

How is AI a force multplier? Because they can pump out more memes per day using it? Relying on bullshit AI slop for memes is a skill issue. Just using the internet is not the same impact as generating things with AI. You admit you live in a red state where people don't care, where I'm at plenty of people do hate AI. What is the connection to the high speed rail argument?

load more comments (1 replies)
load more comments (2 replies)
[-] Thordros@hexbear.net 13 points 1 week ago

My "I'm not crashing out over my post getting moderated on chapo dot chat" t-shirt etc. etc.

You aren't offering feedback in good faith. You're being mad that Korea used an AI image in a newspaper article and don't want to admit it. And you're framing it as Hexbears just don't trust Korea I guess! That's disingenuous and you know it.

Move on, brother. This is not a hill worth dying on.

load more comments (3 replies)
[-] razgriz@hexbear.net 13 points 1 week ago

That image is definitely AI because I don't think the DPRK wouldn't allow workers to stand on those dangerous narrow beams without scaffolding or guardrails for a photo

And also the other images don't look like a mountainous area and it seems like a single span bridge not two bridges with one offset and lower than the other

[-] tocopherol@hexbear.net 6 points 1 week ago

They each have different hi-vis vests and hardhats too, it seems unlikely they would outfit their workers like that, and each standing in the same generic pose. It seems obviously AI but I don't know the context, did DPRK claim it was a real photo or just a promotional sort of thing for the project? They have murals with idealized images of workers and socialist progress, they might use AI in the same way.

[-] Acute_Engles@hexbear.net 7 points 1 week ago

Now I have to be concerned about whether photos posted by DPRK government are considered to be legimate by Hexbear, which seems to be the greater authority on what actually happens in DPRK.

I missed the snark the first time around.

lol

[-] AssortedBiscuits@hexbear.net 7 points 1 week ago
  1. It was obviously AI.

  2. It shouldn't have been deleted. AES media doesn't have a hater boner towards AI like Hexbear dot net. I've seen other AES media like Xinhua and Vietnam Plus post articles with AI images. Nobody fucking cares. Those Iranian Lego agiprop should've been the clue that AI is a tool like any other.

This whole saga is just Western leftists being useless moralists per usual while the rest of the world moves on. And unlike Western leftists crying about AI while doing fuck all to actually stop AI, Iran actually does anti-AI praxis like blowing up Amazon data centers in the UAE. And one of the reasons why Iran was able to light the Zionist's asses on fire is because of the DPRK. Iranian engineers had help from Korean engineers when all those Iranian underground bunkers and fortress cities that shelter their missiles and drones Iran used to bomb data centers were designed and build. DPRK official state media posted one (1) AI image and suddenly that all gets flushed down the drain.

[-] dead@hexbear.net 2 points 1 week ago* (last edited 1 week ago)

Many of the images which are frequently published by DPRK feel unrealistic. There is often photos of massive gatherings with coordinated dancing, firework shows, etc. There are often photos of the Pyongyang circus, which features acrobats and trained bears. DPRK often posts photos of military training exercises, often with martial arts performances and the solders breaking bricks with their bare hands. DPRK often posts photos of military equipment which seems unfathomably large, such as a 5000-ton warship and a 8700-ton submarine. These types of photos were posted also before AI was a norm.

What am expected to do when posting images from KCNA? Am I supposed to inspect every pixel for logical inconsistency?

April 24, 2026 - DPRK people compete at tug of war in a large stadium

April 15, 2026 - DPRK people enjoy firework show to celebrate birthday of 114th Birthday of Kim Il Sung

April 8, 2026 - DPRK exhibition for Industrial Arts.

April 7, 2026 - Korean Electronic music concert. Appears to be drone show in the back.

March 26, 2026 - Two Tourists explore Ryongmun Cavern

[-] Acute_Engles@hexbear.net 19 points 1 week ago

I don't understand the point you're trying to make.

If one LLM generated image makes you treat the entire DPRK as untrustworthy that's on you

[-] mathemachristian@hexbear.net 15 points 1 week ago* (last edited 1 week ago)

Please, look at the right crane, it's behind the bridge, it's facing away from the "camera" but the ropes still are holding up the bridge? Like the load of the crane is behind it?? The left crane doesn't even have a trolley???? It is AI generated, that was no reason to remove the post nor does it mean that the DPRK news agency is untrustworthy, but I am incredibly certain the image is AI generated.

[-] Biddles@hexbear.net 8 points 1 week ago

Your post wasn't deleted because it had an AI photo in it. It was deleted because you crashed out when people mentioned it was an AI photo.

[-] tocopherol@hexbear.net 7 points 1 week ago* (last edited 1 week ago)

Maybe I have seen too many AI gen images, but none of these look fake to me at all, abnormal and impressive but real. There are several clear signs that show the image discussed in this thread is not a real photo. I don't think the DPRK gov were being misleading with it necessarily, they have often used idealized fictional imagery in art and I don't know if they are claiming this image to even be real.

None of us are expected to be perfect in determining what is true these days. In my view you should post what you believe is reliable, we can comment and discuss as a group if it is or not and we can develop our skills for determining the truth together fidel-salute

load more comments
view more: next ›
this post was submitted on 24 Apr 2026
30 points (100.0% liked)

feedback

7906 readers
1 users here now

Tell us how we're doing and report bugs. If you're technically inclined, please submit bugs to the chapo gitea project.

If you want to request a new community the place for that is over in !commrequest

All post must follow the Hexbear Code of Conduct.

Hexbear Warrant Canary

founded 5 years ago
MODERATORS