Most of the stories posted on large subreddits were imaginary anyway. Now they are just being hallucinated by an AI. Is there much of a difference?
this post was submitted on 06 Jun 2025
322 points (98.5% liked)
People Twitter
7277 readers
1481 users here now
People tweeting stuff. We allow tweets from anyone.
RULES:
- Mark NSFW content.
- No doxxing people.
- Must be a pic of the tweet or similar. No direct links to the tweet.
- No bullying or international politcs
- Be excellent to each other.
- Provide an archived link to the tweet (or similar) being shown if it's a major figure or a politician.
founded 2 years ago
MODERATORS
Back in my day you actually had to be creative to get people to believe your bullshit story.
back in my day my dad would leap off the hell in a cell ring, jumper cables in hand, and everyone would clap
Back in my day the narwhal baconed.
Edit: to be totally honest, in my day the narwhal had yet to even think about baconing. Slashdot and Digg imploded about a year after I found reddit. Forums were still the main way to communicate.....
The communities like AITAH and relationships on reddit are completely cooked. There is so much AI nonsense and as noted, people completely eat it up.
AITAH has been mostly made up rage bait stories long time before AI nonsense. There’s nothing that generates more engagement than someone being wrong on the internet.
The difference now is that the rage bait has become automated.
AITAH for pranking my gf by putting a plastic bag over her head?
So I put a plastic bag over my gf's head as a joke. She flailed around and punched me in the face. Then she ran off somewhere, looking mad. Why can't she take a harmless prank? Where has humor gone these days?
The title is usually a bait and switch to get you into the comments.
AITAH for shoving a small child?
There was a bus about to hit them and I saved their life. Their mother cursed me out because the kid got a skinned knee. Half my family thinks I need to apologize. They're blowing up my phone.
I've noticed that on almost every front-page AITAH-style post for many, many years. It makes sense now.
This isn't particularly noteworthy or concerning, imo. These two communities are probably the least likely to care about authenticity, and also best suited for a system that's only real success is spinning moderately coherent fictions. They have and will always be most interested in drama, which is easy to fabricate.
The real danger of these systems is shaping public discourse by sheer volume of noise. They can make any opinion seem overwhelmingly supported, or tear apart opposition with distractions and infighting. And they do it with a single-minded focus spread across thousands of voices.
They have and will always be most interested in drama, which is easy to fabricate.
Remember all the Reality Shows (TM) that took over because they had no scripting and relatively tiny production crews? They took over because they were cheap.
Looks like AI is about to lower that bar yet again.
Those quickly became scripted and fake reality too, if any of them were ever even not that.
They were always badly scripted.
Sure , of course those subs don’t serve any real purpose besides entertainment. The effects of AI making up stories could be to subtly confuse people about how others think and make decisions at most. Though of course it’s useless to give advice to an AI OP, thousands of people still read and discuss what people write. Who knows, the AI stories could even seem less fake than fictional stories.
Lemmy is cooked too.
I am AI
Behold my genius!
There are two Rs in Strawberry!
No, but seriously, you're dead right. Shit is unnerving.
Basically the "reality TV" of Internet forums. Should come as no surprise, even human made stories are usually fake.
cooked without better identity mechanisms
And that worked out so well for Facebook... giving up privacy and anonymity is not the solution, watch that last step, it's a doozy.
Those who would sacrifice liberty for security, deserve neither and will get none.
I don't think that's what they meant by identity mechanisms. Or at least I hope not. It's just not enough to require an e-mail to verify that you are in fact a valid user
Indeed. My dream internet is anonymous
Reddit can tell when I make an alternate account to get around a bullshit site wide ban but not when a robot is pretending that they took his wife in the divorce?
The thing is that you (probably) made not effort to hide your identity. The people that make the bots, use all of them
For you the process is :They banned one of your accounts. That account last logged in from IP A. Right after that, someone using that IP is making a new account. Probably same person. Banned again
For the bots, at a minimum is: they just banned and account. That account always logs in using a very popular VPN. Their browser is non fingerprintable. We can't just ban any new accounts created from that VPN. Registration allowed. (Probably show a captcha)
For you the process is :They banned one of your accounts. That account last logged in from IP A. Right after that, someone using that IP is making a new account. Probably same person. Banned again
I've waited several weeks and made accounts on different devices on different networks. Still banned.
Those devices may have a fingerprint. A lot devices can be uniquely identified. Here is an example that I read a long time ago so it might be wrong or I might not remember correctly.
Your phone has a gyroscope, compass and accelerometer. Those tools are not perfect. Your compass does not point straight north, so it has to be calibrated. That calibration produces like a 6digits offset (0.739842). If you look at all the calibration numbers for a device you get a number that is basically unique for each phone. So even if you used different networks, they can still know that you used a device that you had previously used or associated with you account somehow. That's why brave is an important step in what I said.
Found a paper about it
https://www.ieee-security.org/TC/SP2019/papers/405.pdf
What did you do to get banned BTW?
Maybe there were cookies from the original account on both devices?
Unbanned? No.
Make a new account? Maybe. But I don't really know. I'd try with a paid VPN, and brave browser private window ( some time ago I tested brave, Firefox and chrome for fingerprinting, and brave gave good results out of the box)
I can get unbanned by using a VPN? I heard those get shadowbanned
If they couldn't tell, the bots would get banned sometimes.
Okay but that's worse. You do see how that's worse right?
We could use public libraries to do some kind of "are you a human" confirmation. I.e. you get a digitally signed token proving that you are a human, that you can use on the internet to show that you are human. Only the library has to be trusted by the public ...
And then you take that token and give it to your AI...?
yes, but at least you can't easily create 10000 bots at once. you could only create 1 bot at a time, and that would already do a lot.
I'm not so much worried about 1 person using AI to write their comments; I'm way more worried about large political parties making 10000 bots to support their agenda (propaganda) or companies creating a lot of bots to create hype about their products (advertisement).
How would you verify that you are a human? The only ways I can really think of are enormously invasive
walk up to your library in person and let them give you a chip card that contains an (anonymous) private key, or let them verify your fediverse account somehow, both methods would be without you showing them any kind of ID card.
So is this going to only work with fediverse accounts or are you gonna try and teach people about public/private keys?
And what happens when that database leaks and someone starts using your key to post illegal things? Or should we use password rules and have your library generate you a new key for every platform?
I'm not sure about the details yet. I think, libraries give you 1 chip card that contains a public/private key pair that you can use to "proof" yourself on the internet. Ideally, it would have widespread support and many platforms (both fediverse and commercial platforms) would support it.
I think it may or may not be tied to your person, i.e. you might get that chip card without any kind of ID. Then it wouldn't matter so much if it leaks. But idk, just floating ideas here.
They would at least need to have your ID on record, so they know you aren't just getting a new private key at a different library every day. That said, I'd be happy with it if that record only showed that my ID had been used to acquire such a key, without any link between the ID and the key
Could just do MFA checks for all posts, comments, and upvotes. Imagine?
It would cut out the slop about 99%
Dissent detected, officers en route
MFA helps boost security of an account. It would do nothing to prove a user is human or not.
Funny this coming from Twitter. Nothing but Nazi bots left in that cesspool.
Hey that's not true! There's still plenty of hentai still there too
At the end of the day, most internet forums exist as entertainment. It doesn't really matter whether the text is from real people or not, what matters to Reddit and other platforms is that real users show up to pump numbers and get served ads. It's kinda like how pro sports is not latently about athleticism, it's just another entertainment platform to get butts into seats and eyes onscreen to make a buck.
If bot content gets the eyeballs, so what? Guess the content being fictional doesn't really matter. People don't like being lied to, but then nobody is claiming everything you read on the internet is true in the first place.
Anyway, to all the people who imagined humanity's future would end up looking different than WALL-E: sorry to disappoint. You'll get a hover recliner, at least.
Maybe Elon was on to something when Grok tagged everything with an unrelated diatribe about farmers in south Africa. Perhaps every AI model needs to be an activist about one random topic so we can tell them apart. Or ya know, maybe congress could do stuff to require their output be reverse searchable so we could just match up the output to what people say...
Hot take : we need one social network where everyone’s identity is verified. No one is anonymous, everything you post publicly is linked to your real name.
Not every social media, anonymity is also good.
I’d prefer anonymous users verified by real identity.
It’d take some thought as to how you properly establish accounts are owned by individuals but it’s not impossible.