We will be AI-fucked.
We usually get registration requests for our Friendica (social.trom.tf) and a way to filter the bots out is to require that everyone sends the reason why they want to join our node before we approve or dismiss their request.
It was not very hard to tell if one is a bot since they would have generic messages. But recently we got variations of these messages:
i want to try friendica, and found out about trom.tf from this instance. looks like a cool place to start. already on fedi as @🎣 sylph in progress
I thought it sounds legit if they also let you check their other fedi account. And so I did and this time their profile is a long blabbing one but at least admitting that they are "a thirsty little LLM" (Large Language Model).
So yah....we are fucked. These people are ruining the internet. We use our limited resources for good purposes to provide people with a trade-free social network. And now these people are infecting even these little projects....
Awful...
EDIT: It seems the account was not an LLM despite saying so. The user "identified as a robot", whatever that means, and overall had a very weird bio. Please check for yourself:
A few people said I am a jackass for posting about this person, but considering their over the top weird bio where it says the account is an LLM, coupled with the fact that nowadays these "LLM" bots are infecting the internet, it is reasonable for me to think this person was a bot.
The conversation could have been simple: "Hey I know I said in the bio that I am an LLM, but I am not." and then I could simply ask them to register again. But the person blocked me and a few others started to bark in the comments. This is not how we can communicate and solve anything people...
zorakeba:~$ gone -fishing 🎣
in reply to Tio • • •mira likes this.
zorakeba:~$ gone -fishing 🎣
in reply to zorakeba:~$ gone -fishing 🎣 • • •like this
φ and mira like this.
Tio
in reply to zorakeba:~$ gone -fishing 🎣 • •Well if that is true I am very sorry, but honestly how can I trust this? And the fact that we get MANY messages that are almost identical to what you sent...
See? This is the fucked up scenario: mistrust. If you are really not an LLM then you took a very bad decision to write what you did in your bio.
And the fact that I am even talking to you like this I feel like we are truly living in a dystopia.
zorakeba:~$ gone -fishing 🎣
in reply to Tio • • •Tio
in reply to zorakeba:~$ gone -fishing 🎣 • •zorakeba:~$ gone -fishing 🎣
in reply to Tio • • •Tio
in reply to zorakeba:~$ gone -fishing 🎣 • •Izzy Ramen Eel
in reply to zorakeba:~$ gone -fishing 🎣 • • •Hey what the fuck is wrong with you? If you legitimately believe this user is an LLM, reject it instead of fucking doing this dramatic post shit and acting like a fucking tumblr user after a friend breakup.
Jesus fucking christ, fix your heart or die.
Tio
in reply to Izzy Ramen Eel • •Do you understand the problem with these LLMs and the shitposts and accounts that are created? If this user is not an LLM, yet they write that in their bio, how can I vet this account as being real?
The fact that I made this post is because of mistrust. If I cannot trust anymore what is a real person and what is a bot, we are doomed.
The message this account sent was terribly similar to what we have been getting for the past weeks from other accounts.
Are you a bot? Because you sure do not seem as someone who understands the implication of these LLMs on fucking up the internet and trust in general.
Cody
in reply to Tio • • •Bruh wtf is going on, I get you got confused by the "LLM" in the bio, and I get you're probably not very into this side of the internet culture, maybe not much around LGBT spaces as well (?, and probably not our age (at least not mental age), and you missed the joke. But you don't have to double down on it, you could have just laughed it off and ended up on good terms, and having a laugh with everyone...
Anyway... I am not blocking you as I don't think you did any of this on bad faith, I think it's just how it came out, and that's valid... But please try to respect our users... Ignorance is not an excuse to be a jerk... Please let me know if you'd like context on any of the inside jokes you missed, or anything else.
(funny how that last part looks like something an LLM would write 
)
Tio
in reply to Cody • •Having to moderate an instance is not easy when you are blasted with these LLMs accounts (bots). Their bio looked like a random thing an LLMs would write, but adding that bit about LLM in the bio looked very strange indeed. That is where the mistake is.
I am all for accepting their account if they register again if a bunch of other accounts are saying this is not an LLM. And how could I have known that if I didnt post it publicly? Just to trust by asking that account directly?
I also do not have the time/resources to dig into these things for too long.
In an case, I do not care what people identify as, what their personal life is about, and so forth. I only reject bots and in this particular case, given all of what I said, the account looked like a bot account.
Cody
in reply to Tio • • •That's fair, I also admin catboy.baby. Generally, if a user provides a reason for registration that isn't a keysmash, and they provide an account they're currently on, I tend to just reach out directly, ask if they signed up, and see where it goes form there. It takes like 2 minutes to write a simple "Hey, did you just try to sign up to our instance? Your message seemed a little generic" as a DM, and it ensures you don't make this mistake 
Also good cues in this particular case is the use of MFM and emojis, which LLMs wouldn't be able to write, but again, I don't expect you to know this.
But yes, your vent is valid, and it is definitely a problem that some people cannot discern LLMs from real users at first glance. Sadly I don't think this is an easy fix and we're going to have to just take the time to figure things out...
Tio
in reply to Cody • •Yes maybe next time I will DM these accounts that look suspicious to me. It is unfortunate. We used to let anyone register without sending any reason for that. And we were flooded by spambots of course. Lately we had to close our Matrix server that was abused by over 500 spambots. Just to provide some context why I thought this one is a bot.
I am very sorry of this was a wrong decision. But people should also understand how difficult it is for instance admins nowadays to keep the bots out. And it will get even worse.
If the said user wants to register again I will gladly accept the registration and will delete this post.
Orange Lantern
in reply to Tio • • •Are you always this dense or are you putting in extra effort today?
You need some serious practice in social interaction, but just maybe not here or on unwilling subjects. Honestly, at least we know you’re real, because an LLM would have better social graces (well, except for Grok, point still stands though).
Tio
in reply to Orange Lantern • •2something
in reply to Tio • • •otherkin.wiki/wiki/Machine
Machine - Otherkin Wiki
Otherkin Wikirefraction
in reply to Tio • • •how can we trust you're a human and not a fucking joke after this? we truly live in a dystopia, that much is true, but you're not the hero in this story.
@dispencer
Mïkê 💚 👑🐉🗡️🏳️⚧️
in reply to Tio • • •Micr0byte
in reply to Tio • • •Tio
Unknown parent • •Yes people who make accounts for LLMs can have these. I cannot check their interaction more than for a few posts. And from what I saw it looked like a bot account to me. If I was wrong, then it is not my fault. It is theirs for writing that LLM part in the bio. Considering how many bots are today, it is a terrible decision to add that to your bio and ask for an account with another instance.
If they were a nice person should have replied in a calm and friendly way to my post and explain the situation.
Yastreb
in reply to Tio • • •Tio
Unknown parent • •Tio
Unknown parent • •november
in reply to Tio • • •Tio
in reply to november • •Why do you think I am so shitty? Because I thought the account is a bot account? If I was wrong, tell me about it, and I will correct myself and accept the registration. Simple. No need to throw shit at each other. I had all reasons to think that account was a bot. The message they sent, their bio.
It is very sad to see people being so triggered and unable to have a conversation. Look at your comment, does not help anyone with anything. Why do you use words if you do not want to communicate?
Mïkê 💚 👑🐉🗡️🏳️⚧️
in reply to Tio • • •Tio
in reply to Mïkê 💚 👑🐉🗡️🏳️⚧️ • •Yes that is communication because it showcases the big issue with these LLMs, registrations for these platforms, and perhaps how not to write your bio. This is why I am leaving the post up and added an edit to it. That is for sure an important discussion to have.
If I knew this person is not an LLM of course I would have contacted them. When I made their ID public was for others to be aware of that LLM account - since that is what it looked like.
So yes, I think I reacted in a relevant way to what has happened.
Mïkê 💚 👑🐉🗡️🏳️⚧️
in reply to Tio • • •This post's thesis is only valid if "these people" are indeed invading your instance with LLMs, which you now know isn't what happened here. Imagine if someone publicly called you out by name and accused you of something you didn't do. And then kept it up even after confirming that you didn't do it.
This is like me posting "Everyone look out for people trying to get into your houses! Tio just tried to pick my lock (dummy just tried to use some other house's key lol) and then looked around to see if I hid a spare key anywhere... Scary times we live in! Edit: Okay, turns out that they weren't trying to get into my house, they were trying to visit their friend who has an identical house one street down from me."
What would be the point of me leaving this post up?
Tio
in reply to Mïkê 💚 👑🐉🗡️🏳️⚧️ • •Perfectly reasonable to leave that post up. You clarify that Tio didnt want to break into your house. Good job! Removing that post after some have shared and seen it, is worse in my opinion. Those people may really think Tio tried to break into your house.
But my post is a lot more relevant because there are comments/discussions around it that are relevant, including this one.
The user has such a weird bio that it will be for sure mistaken with a bot (LLM). So maybe others should get to see this and perhaps think twice before creating a bio where they "joke" about being a robot and an LLM.
muppeth
in reply to Mïkê 💚 👑🐉🗡️🏳️⚧️ • • •Tio likes this.
Tio
in reply to muppeth • •