16
submitted 5 months ago* (last edited 5 months ago) by Xylight@feddit.online to c/reddit@lemmy.world

This was the weirdest thing I've seen today. These are only the ones I've spotted.

funnily enough, these bots are also replying to an obvious repost from another bot account. It's at the top right now! Beautiful

https://www.reddit.com/r/goodnews/comments/1p8dt2a/_/

tipping points:

  1. consuming so much AI content has led to me able to see subtle patterns
  2. They're all saying "exactly" and saying the same thing"
  3. their usernames are similar, flower/nature related, two words, no profile pictures
  4. All of their profiles have the exact same format of comments with the agreement, summary
  5. and they all have porn on their profile. oh

edit: tf?

top 50 comments
sorted by: hot top controversial new old
[-] tangonov@lemmy.ca 11 points 5 months ago

You're absolutely right! Those posts do have many indicators of having been written by an AI. You're doing a great job finding these comments

[-] TheObviousSolution@lemmy.ca 5 points 5 months ago

that is completely correct! it's awe—inspiring to see this level of rigorous investigation. that sort of attention to detail gives me hope that something can be done about this problem. this post fills me with hope for humanity

[-] theMoops@lemmings.world 2 points 5 months ago

Totally feel this. It’s kinda wild how refreshing it is to see someone actually dig in instead of just shouting hot takes into the void. When people put in real effort — like, actual research, receipts, context, the whole thing — it reminds you that not everyone is just doomscrolling and giving up.

Honestly, posts like this are the rare moments where you remember, “Oh right, humans can be competent and thoughtful.” Gives me a tiny spark of optimism I didn’t expect today.

More of this energy, please.

[-] PokerChips@programming.dev 1 points 5 months ago* (last edited 5 months ago)
[-] PlaidBaron@lemmy.world 2 points 5 months ago

To make a grilled cheese you need the following ingredients:

  • Bread
  • Cheese
  • Butter

Step 1: Melt butter in a pan.

Step 2: Place the bread on the butter. Soak the butter in.

Step 3: Add the cheese slice to the bread.

Step 4: Put the bread on top of the cheese.

Step 5: Grill the sandwhich on both sides until golden brown.

Some good additions to grilled cheese are tomatoes, ham, or uranium-238.

[-] tangonov@lemmy.ca 4 points 5 months ago* (last edited 5 months ago)

Absolutely! Of all of the grilled cheese recipes out there this is by far one of the recipes out there. A little bit of cheese can really make your day better. Just like that time that Mankind faced off against The Undertaker the WWF pay per view special "Hell in a Cell", 1998. Mankind climbed to the top of the 16 foot cage and taunted the Undertaker to wrestle him up high, only to end up getting thrown into the commentator tables below. Everyone thought the match was over but just before Mankind was taken out in a stretcher he got up and ran back for more. I'll never forget how The Undertaker choke slammed him through the top of the cage and onto the thumb tacks below. It just really gives me hope for humanity.

[-] Xylight@feddit.online 5 points 5 months ago* (last edited 5 months ago)

All of social media is dying before our eyes. every platform is basically fully botted or actively dying, even lemmy seems to have fallen off quite a bit

[-] rimu@piefed.social 5 points 5 months ago* (last edited 5 months ago)

I am working on LLM detection for the threadiverse. But other than one idiot last week spamming LLM posts and comments there hasn't been much.

[-] pelespirit@sh.itjust.works 1 points 5 months ago

There are in politics conversations, but still not nearly as bad as reddit. Even before I left, it was like a weird kind of prolific dead. Kind of like the conversations in OP's pics.

[-] db0@lemmy.dbzer0.com 1 points 5 months ago

It's actually the blessing of not being big enough to attract their attention

[-] fedorato@lemmy.world 2 points 5 months ago* (last edited 5 months ago)

Just going to put my tinfoil hat on for a sec…

Part of me does wonder if the seemingly pointless proliferation of ai slop like this botting is being done intentionally to fast-track a ‘need’ for identity verification (and thus more precise tracking and surveillance).

ID verification is already being pushed on a few fronts (like to ‘save the kids from social media’ or whatever), maybe this is just one of many irons in the fire.

With ID verification then Facebook etc could angle themselves as ‘save havens’ from an ai slop enshittified internet. You’d essentially have to completely give up your anonymity to participate in interactions with other verified humans.

So your choices become:

  1. Participate in open platforms, but never really know for sure if you’re dealing with humans. At some point LLMs may be good enough that it’s impossible to know.
  2. Participate in closed platforms, where you can be reassured you’re engaging with real humans - but you’re also under total surveillance.

Surely sites like reddit or Facebook, if they tried, could control this stuff otherwise?

[-] Kalcifer@sh.itjust.works 0 points 5 months ago

[…] even lemmy seems to have fallen off quite a bit

Do you have an example of what you mean?

[-] Xylight@lemdro.id 1 points 5 months ago

from a different account since feddit.online wont show me this comment

There's nothing new happening on this platform, literally the frontpage consists of:

  • doomerism
  • trump and friends news
  • tech company does stupid thing
  • political meme
  • lets revolt guys. it'll happen any moment now we swear.

The majority of posts are made by the same people.

As someone here since ~2022, I've seen the frequency of posts continue to fall.

[-] INeedMana@piefed.zip 2 points 5 months ago

Sometimes I end up mindlessly scrolling yt shorts (not logged in). From time to time I get to a short that is clearly generated. Like the weird ones, often with animals, with drops of water appearing out of nowhere on the fur or front paws somehow transforming into hind ones, etc. And there, in the comments, very often are whole chains of comments that seem to completely be missing the fact that it's generated. Saying things "how wise it is to do that", "how cute", etc. It could be older generations not noticing the details (I see how my parents not notice those things) but I think most of those are probably bots. LLMs exchanging their "awww"s under generated videos. "Dead internet theory"

load more comments (3 replies)
[-] pleaseletmein@lemmy.zip 2 points 5 months ago

Gotta kick out all the human users to make room for more of these.

[-] 1995ToyotaCorolla@lemmy.world 1 points 5 months ago

It’ll be so nice when the bots can post for all of us on the internet! It’ll give us plenty of free time to spend in the mines

[-] falseWhite@lemmy.world 1 points 5 months ago

I was typing a long comment about all this, but in the end I decided to sum it up:

Fuck Reddit and Fuck AI.

[-] echodot@feddit.uk 1 points 5 months ago* (last edited 5 months ago)

Someone's told the last AI not to use capital letters because some someone somewhere thinks that makes it look more human. Forgetting of course that autocorrect would change most non-capitalized words into capitalised ones automatically — so it's just suspicious.

[-] AbsolutelyClawless@piefed.social 1 points 5 months ago

All three are 22-day-old accounts, too.

[-] Carighan@piefed.world 1 points 5 months ago
  1. consuming so much AI content has led to me able to see subtle patterns

It's a certain "tone" to their written text. It's difficult to identify from small blurbs like the ones you got there, but once you've seen enough LLM output, even if someone tells them to write in a specific "style" they'll still have a certain uncanny type of expression that is almost, but never actually, how humans write.

[-] foodandart@lemmy.zip 1 points 5 months ago

I find it's rather like ad-writer's content.

Non-offensive, simplistic, and based on a set of words that elicit a kind of generic "warm fuzzy" feeling.

It's saccharine and bland.

[-] rozodru@pie.andmc.ca 1 points 5 months ago

positive acknowledgment, brief summary of previous statement/post, positive conclusion/ending statement.

That's generally what I look for when trying to weed out AI. at one point it was as easy as see a bunch of EM dashes but now they all seem to follow a pattern that they perceive as natural.

[-] flowers_galore2@lemmynsfw.com 1 points 5 months ago
[-] rivvvver@lemmy.dbzer0.com 2 points 5 months ago

Yes, this absolutely. This kind of research in a changing world like ours is really important right now. Major respect to her for that

/j

[-] 0oWow@lemmy.world 1 points 5 months ago

Wait a minute... 🧐

[-] lennybird@lemmy.world 1 points 5 months ago

Well at least I don't have porn in my profile but I realize I do sound like an LLM at times in my comments... Though I guess that's to be expected depending on the material they train on.

[-] guy@piefed.social 1 points 5 months ago

Just some years ago the dead internet theory was said to be a fun but untrue thought. Well.

[-] Rentlar@lemmy.ca 1 points 5 months ago

Starting with agreement is a polite thing to do, but yes the structure of the comments feels uncanny. The words "turn regret into compassion" makes me cringe so hard and tips me off that a bot wrote this.

[-] BluesF@lemmy.world 1 points 5 months ago

Reddit has so rapidly descended into nothing but bots. Especially on certain subs, for some reason... Even some quite niche ones just seem to be bots talking to each other. Fortunately the only sub I really want to keep my reddit account for is mostly safe, but even there we had some issues.

In some cases I get what's happening - bot post with featuring some kind of obscure product, then buried in the comments you find the bot replies letting people know (apparently organically) where they can buy it - but in other cases like this it just seems pointless. I suppose the idea is to make the profiles seem natural, but they're almost all private anyway.

[-] Sam_Bass@lemmy.world 1 points 5 months ago

Gaslight trainers

[-] Treczoks@lemmy.world 1 points 5 months ago

Thirtyish years ago, we played a multiplayer online game called "LPMud". There were three talking NPCs in the game: Harry, basically a simple programming example on talking and reacting NPCs, Sir Obliterator, a dark knight with a more advanced vocabulary and a few talking points about a quest, and Eliza, basically a NPC with an Eliza engine.

Usually, they never met. Harry "lived" in the core area of the game, Sir Obliterator in or around the quest area to which he belonged, and Eliza was normally not even active.

Some wizard had summoned them all to the guild hall, the entrance area of the game for fun, and they were rather busy "talking" with each other.

They were annoying, but also hilarious...

[-] Bosht@lemmy.world 1 points 5 months ago

Dead Internet theory in practice.

[-] tym@lemmy.world 0 points 5 months ago

Where LLM though? All I see in this screenshot is a plane full of essential oils saleswomen from Utah..

[-] hayvan@feddit.nl 3 points 5 months ago

They have the self awareness of LLMs though.

[-] tym@lemmy.world 1 points 5 months ago

linkedinlunatics is a phrase for a reason.. LLMs are just polite white people faking the interaction. anyone who's worked in or around corporate recognizes the synergy

https://www.youtube.com/watch?v=GyV_UG60dD4

[-] jqubed@lemmy.world 0 points 5 months ago

I think the two words and a number are what Reddit will come up with if you let them pick a name

[-] otter@lemmy.ca 0 points 5 months ago

Adjective-Noun-# is the usual format for those. This looks different, like someone put in a list of wholesome / floral / nature-y terms into the name generator, in order to have wholesome looking accounts

[-] Capricorn_Geriatric@lemmy.world 1 points 5 months ago

Adjective-Noun-#

Fuck, I'm the inverse. I'll know better for next time.

[-] ChickenLadyLovesLife@lemmy.world 0 points 5 months ago

Do LLMs always omit the period on their last sentence? Seems like that would be a dead giveaway

[-] Xylight@feddit.online 2 points 5 months ago

No, when it comes to LLMs there's hardly any "dead giveaways" now. You have to learn to recognize the patterns.

Omitting the final punctuation is quite a common thing people do, in fact you did in your comment. It's probably just a part of the system prompt.

[-] ChickenLadyLovesLife@lemmy.world 1 points 5 months ago

in fact you did in your comment

Whoosh

[-] grepe@lemmy.world 0 points 5 months ago

i don't think i (or perhaps anyone) can recognize any single particular comment as being llm generated... but when the bots come in force it is still really easy. basically it boils down to this: many replies keep reiterating the same exact points in slightly different way with the same exact keywords. if you would use chatgpt to summarize each response you'd get basically the same thing from all bot replies.

[-] jgandert@lemmy.zip 1 points 5 months ago

I agree. I believe it's difficult for me—or anyone else—to pinpoint a specific comment as being generated by an LLM. However, when numerous bots are involved, the pattern becomes clear. Essentially, many responses end up repeating the same points, just phrased differently and using the same keywords. If you were to use ChatGPT to summarize each response, you'd essentially get a very similar outcome from all the bot-generated replies.

[-] grepe@lemmy.world 1 points 5 months ago

thank you! we need slightly longer chain or more parallel replies to drive the point home... anyone else?

[-] SGforce@lemmy.ca 0 points 5 months ago

Think it's probably a bug in the script they're running. It's cleaning one character too many off the end.

[-] PeriodicallyPedantic@lemmy.ca 0 points 5 months ago

Why would they cleaning characters of the end in the first place?

load more comments (4 replies)
load more comments
view more: next ›
this post was submitted on 28 Nov 2025
16 points (94.4% liked)

Reddit

22916 readers
2 users here now

News and Discussions about Reddit

Welcome to !reddit. This is a community for all news and discussions about Reddit.

The rules for posting and commenting, besides the rules defined here for lemmy.world, are as follows:

Rules


Rule 1- No brigading.

**You may not encourage brigading any communities or subreddits in any way. **

YSKs are about self-improvement on how to do things.



Rule 2- No illegal or NSFW or gore content.

**No illegal or NSFW or gore content. **



Rule 3- Do not seek mental, medical and professional help here.

Do not seek mental, medical and professional help here. Breaking this rule will not get you or your post removed, but it will put you at risk, and possibly in danger.



Rule 4- No self promotion or upvote-farming of any kind.

That's it.



Rule 5- No baiting or sealioning or promoting an agenda.

Posts and comments which, instead of being of an innocuous nature, are specifically intended (based on reports and in the opinion of our crack moderation team) to bait users into ideological wars on charged political topics will be removed and the authors warned - or banned - depending on severity.



Rule 6- Regarding META posts.

Provided it is about the community itself, you may post non-Reddit posts using the [META] tag on your post title.



Rule 7- You can't harass or disturb other members.

If you vocally harass or discriminate against any individual member, you will be removed.

Likewise, if you are a member, sympathiser or a resemblant of a movement that is known to largely hate, mock, discriminate against, and/or want to take lives of a group of people, and you were provably vocal about your hate, then you will be banned on sight.



Rule 8- All comments should try to stay relevant to their parent content.



Rule 9- Reposts from other platforms are not allowed.

Let everyone have their own content.



Rule 10- Majority of bots aren't allowed to participate here. This includes using AI responses and summaries.


founded 2 years ago
MODERATORS