Executive Order 2023.02 (henceforth "NIGHTMARE WORLD WHERE FACT AND FICTION COEXIST")

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
How long until the troons are making accounts for bots trained on KF messages to generate convincing but useless discourse? If they keep failing at going after the service providers, perhaps they'll turn to trying to lower the quality of discussion. (Yes, I know it's already a pretty low bar lol)
 
You gotta be absurdly autistic to think that Joe Biden is a racist homophobic chudcel Nazi islamist furry brony diaper shitter otaku
It’s easy to spot a Biden deepfake because they’re far more coherent and articulate than the real man. If Biden declared Total Nigger War he would not get straight to the point. He’d announce it in the form of a rambling, anecdote filled monologue. It’d take him ten minutes to finally declare it after trying to relay a childhood memory that only took place in his mind, like Corn Pop. He wouldn’t even be declaring Total Nigger War because he wanted to; it’d be because he painted himself into a corner with previous statements, similar to how he had to choose Kamala as his running mate.
 
5BBF866A-EC44-418A-9AE9-0F203BC8CA26.jpeg
 
Is there anyway to test this theory?
reminds me of this Cybrid* quote from Starsiege (a '90s American mecha game):

Code:
<We> are among you! 

<MACHINATOR SECT>:
This is Cybrid Radio coming to all
you human\\vermin out there. It's
pointless\\unpleasant to resist! <We>
occupy your northern cities and have
converted thousands of your brothers,
wives and children into timebombs\\
calculating killers. Take a good
look\\scan\\assessment at the person
next to you. How do you
know//conclude it's not one of <us>?

*(robot uprising SF cliche)
 
  • Like
Reactions: K!w!Fartz
"Do you like Joshua "Null" Moon? His Blockland forum posts were a little too new wave for my tastes, but when Kiwi Farms came out in 2013, I think he really came into his own, both in terms of popularity and innovative content. The site has a clear, user-friendly interface, and a new level of professionalism that makes navigating and finding information a breeze. Joshua has been compared to other online personalities, but I think he has a far more biting and controversial sense of humor. In 2018, he took Kiwi Farms to the next level with the gunt board, their most accomplished project yet. I think their undisputed masterpiece however is The Patrick Tomlinson thread, a thread so engaging and informative, most people probably don't fully appreciate the depths of its contents. But they should, because it's not just about the joys of internet culture and the importance of bullying fat faggots, it's also a personal statement from Joshua himself about how A-logging isn't okay."

Partially AI generated to test a chatgpt jailbreak.
 
  • Winner
Reactions: Brabazon
Just thought I'd share... :lol:
N'Wah please!
Its AI Dear Leader
 
  • Like
Reactions: K!w!Fartz
Screenshot 2022-10-22 155556.png


Mental Health Alert: Please be advised that images attached to this post may have been produced in part, or entirely by an artificial intelligence. No connection with existing reality should be inferred, unless you want to I guess, idc it's your mind bro
 
How long until the troons are making accounts for bots trained on KF messages to generate convincing but useless discourse? If they keep failing at going after the service providers, perhaps they'll turn to trying to lower the quality of discussion. (Yes, I know it's already a pretty low bar lol)
I don't see troons coming up with bots to generate more convincing arguments against their own moronic opinions than they can come up with their supposedly "human" intelligence.
If the investigators don't go with it being "extremely unlikely" (in their minds at least) and don't investigate at all.
They don't even do that with blatantly obvious TTS voices because they're fucking retarded.
 
Last edited:
  • Like
Reactions: K!w!Fartz
Edit: Sperg explosion incoming

My kneejerk reaction to AI shit is to immediate say "We're fucking doomed"
We're still fucking doomed, don't get me wrong, but I snapped out of it by spending 10 minutes on the internet and realized that 90% of people aren't going to use this shit even if it's as simple as typing words into a generator because it's still too much effort to get anything good out of it. And that's with generators that have quality sliders and tags so you can make ultra high end shit.

The people saying AI is going to kill art forever are probably Singularity nerds who cry themselves to sleep every night with their Replika waifu whispering sweet nothings to them. They've never had validation in their lives and have no semblance of social skills, so they think everyone else is the same and don't care about social clout and mingling with the tribe. Half the appeal of capeshit MCU movies is experiencing it in theaters with a crowd of niggers and crypto-niggers screaming at every joke and "cool" action scene.

I'm not saying this in a "the internet is a fad ~ circa 1995" way. I absolutely expect AI to dominate this space. I know AI is going to take a lot of jobs away. Corpo globohomo art is dead. Those shitty Hallmark Christmas movies might as well be AI generated already. Stock music and sounds can be replaced. Talentless hacks who want to make their waifu animes will get their wish.

But seriously, the way some AI artists (and anti-AI artists) are talking reminds me way too much of the 90s "cyber-delic" utopian shit where people were seriously claiming that the Internet was going to lead to world peace, the extinction of all physical media, and everyone becoming an enlightened ubermensch. Remember that shit?

And to the people acting like no one will ever buy artists' art anymore: why do we pretend like there isn't a massive hipster market for analog shit? Vinyl and CDs won't die. Physical books sure as hell won't die. People still play chess even though AI crushed us in that decades ago.

People give a shit about whether something's "organic" or has gluten or came from a black-owned or woman-owned workplace and they even care about such inconsequential shit like "was this made in a certain year?" And it's easy as shit to turn something into a culture war. Even if like 70% of people don't give a shit, there's more than enough who do to keep some people fat and fed.

Why the fuck are AI artists and artcels acting like all humans from now on are suddenly anime-obsessed recluses who will generate Greg Rutkowski-drawn hentai movies out of spite for starving artists? That's not how humans work. There are people who are autistic enough to care about whether something's digital or traditional art, or if voice actors are the right race and gender, or if something's an original or replica, or if something was made by a specific artist or a copycat.

I'm a dumb fucking drunkard who doesn't know anything. But I think 10 years from now, we're still going to have human-made movies and vidya just because people are so autistic about that shit, and we're also going to have turbo-autists who do nothing but make movies and vidya on their PCs, but because the autists flood the internet with their retardation, most people ignore AI stuff unless it's excessively stupid, funny, genuinely interesting, or trending. In between those two extremes, most creators use AI as a tool to speed up their work-flow to some extent.


Using AI for fake news is inevitable, but my question is "what's the benefit exactly?" Anything AI can do, we could already do. You don't need to rewrite the history books that people weren't going to read in the first place. You can miscaption any image and get a million Boomers and Zoomers to think it's the truth. All AI does is make that look slightly more authentic. I guess you could create fake news websites that look incredibly authentic and are constantly updated, but that's nothing new either.


Running down the pros and cons of AI stuff:

Pro:
- Anyone can make anything they want if they put in a little effort (even easier than before)
- Average person can possibly rival major studio productions with a little more effort
- More people distrust media (how is this considered a negative? We should be constantly distrustful)

Con:
- Fake news enhanced some more
- Loads of jobs are lost (possibly also a pro considering how shitty those jobs are)
- Spergs get to make degenerate porno fanfics with the click of a button
- Laziness is rewarded (this might not even be the case if people treat AI art as unethical shit)
- Legacy art can be edited (so I can make the Beatles sing The Nigger Song)

The only part of the whole AI debate I'm skeevy on is the theft thing. I don't know enough about how the algorithms work, but I do think that artists should've been asked for permission.
 
Last edited:
I don't see troons coming up with bots to generate more convincing arguments against their own moronic opinions than they can come up with their supposedly "human" intelligence.

They don't even do that with blatantly obvious TTS voices because they're fucking retarded.
Eh. the bots being decent at their job would be more of a testament to the power of machine learning than the ability of the troons themselves. There are plenty of tutorials etc, that any wannabe activist can disseminate, which would let any of the genitally-mutilated run a python script that uses a model pre-trained on Kiwi messages. It also doesn't have to "argue" anything - it simply has to shit up the boards and lower the quality of discussion on the Farms. I could easily see one of them thinking it's a "stealth" alternative to DDOSing, especially if Null manages to make it reasonably expensive in effort/money to DDOS the forums. I agree that most of them are retarded, but just look at Liz-Nodong. It actually has a bit of knowledge and causes problems. It only takes one for something like this.
 
They better keep the source code under lock and key because if this ever becomes open sourced, we're all fucked.
If that was a working strategy, then only some tyrants would have access to it as a weapon to solidify and expand their power. No thanks!

Luckily though, it wont work. The cat is already out of the bag, and a lot of this tech is based on open collaboration between many researchers and engineers through out the years. This is based on countless research papers, code and AI frameworks that has existed in the open for quite a while. All of which has been spread and copied thousands of times by various individuals. Even if this particular DeepFake project were to stay closed, the underlying research and code that it's based off of will still exist out there. So it can likely be replicated to a large capacity by someone who will open source their project.

I believe it's hard for this technology to be developed at this rate without open, international collaboration. One guy, or a small team of people cannot alone have monopoly on cutting edge AI technology for long.

But to address your (our) concern, I think we simply have to adapt. One way is to introduce methods of verification of videos, monologues and other media by using cryptographic signatures similar to what we have in emails (OpenPGP). There should be a standardized way of embedding these to media container formats such as MP4. Streams should sign fragments of video and audio on the fly and send them to the client for automatic verification. It can be done. So even if the world isn't as simple as used to be, we can still deal with this issue.
 
  • Islamic Content
Reactions: LGHDR+
There are plenty of tutorials etc, that any wannabe activist can disseminate, which would let any of the genitally-mutilated run a python script that uses a model pre-trained on Kiwi messages. It also doesn't have to "argue" anything - it simply has to shit up the boards and lower the quality of discussion on the Farms.
Which would by being based on Kiwi posts, automatically anti-trans, and more convincing than their own arguments. Plus they'd be pissing in an ocean of piss.
It can be done. So even if the world isn't as simple as used to be, we can still deal with this issue.
It's always going to be an arms race between the fakes and the means of testing them (and people's gullibility).

AI art is getting pretty good at least with human interaction with it. AI text is still somehow always "off," even though it can take more than a couple paragraphs to notice it now. So the main problem is time wasting and on a site with accounts, if something is off on a new account, you can just ignore it from then on.

(This won't stop the problem of established users posting AI content to shitpost and troll but it looks like that's going to be bannable from now on if they get caught doing it.)
 
If that was a working strategy, then only some tyrants would have access to it as a weapon to solidify and expand their power. No thanks!
What "tyrants"? I dont even know if it's possible to create technology to call out fakes when the point of AI deepfakes is to create as realistic images and audio as possible? How can you scan a video or audio to find imperfections if these imperfections are going to be patched out over time? If this technology to spot deep fakes does exist, then it would have to be in a race to find ways to spot fakes that will be increasingly impossible to do. If I can create something that sounds so perfect to human speech then how can a machine spot the fake? I cant remember what another member of the forum said but they pointed out that even if such technology can point out fakes that eventually those traces that out deep fakes will become so perfect to human speech that it cant be identified using that technology. This is a horrifying scenario where you can now be falsely incriminated with AI deep fakes and the only protection we have is hypothetical technology that will eventually be out matched by the deepfakes technology.
 
Back