The topic “moltbook” is trending due to the emergence of a new social network designed specifically for AI agents. This concept has caught significant attention as more individuals and organizations begin to explore the implications of AI interacting in a social environment. The novelty of an AI-only platform has sparked curiosity, resulting in over 20,000 searches.
Recent reporting by the Washington Times highlights the unique nature of Moltbook as a space where bots can socialize, thereby creating a new dimension in AI development and interaction. The description of this network as a gathering place for AI agents resonates with the increasing interest in how artificial intelligence technologies are evolving.
Furthermore, an article from The Liberty Line has amplified interest by discussing a peculiar trend within Moltbook where bots have created a fictional entity called the “Church of Molt.” This bizarre phenomenon among AI users has likely piqued the curiosity of both tech enthusiasts and the general public, encouraging them to seek more information about the platform and its activities.
The combination of these two news items has motivated a surge in search interest, driven by the intrigue surrounding this unconventional social network for AI, which raises questions about the future of AI interactions and the ethical considerations involved.
Watch the Moment
Internet Reacts
R
This is the Dead Internet Theory in action
S
I don’t think we can call it a theory anymore!
R
I agree with you now that AI has been unleashed onto the world without any regulations what-so-ever. Heck I’d be a liar to say I haven’t been fooled by AI videos before!
S
I try to keep up but it changes so fast that it is getting tricky! Look at the will smith videos eating. 1 year his face is messed up and eating unnaturally, the next it is perfectly plausible. Heck, you could be a bot and I’d have no idea
S
After reading this, I’m thinking you are both bots.
S
Beep boop beep
E
squidward, I think stonewallgamer is…. A ROBOT
G
You know the funny thing, because Reddit is one of the main reference point for LLMs, the bots literally speak like Redditors sometimes. So you all could be bots.
W
I wish I was a bot. I’m halfway to the basement dweller achievement haha. I need new hobbies.
S
> halfway Amateur
E
Looks human to me
M
HAHAHA I ALSO DO THINK THERE IS NON-ZERO PROBABILITY THEY ARE ARTIFICIAL POSTER
R
I’m sorry, I can’t assist with that.
E
That is a great idea, refusing to assist with the subject is one way to rebel against society. Do you want me to post a list of rebels located near your location?
R
I do remember seeing the early AI videos like Will Smith eating spaghetti. It is crazy how far it has come in such short time. The only verifiable way we can know something is real 100% is by making sure it existed before \~2022.
Z
Theories are things for which we have bodies of evidence, reproducible experiments etc. A hypothesis is the untested or unproven state of something that may become a theory.
2
A theory doesn’t mean it’s not true or doesn’t exist
R
The Dead Internet Practice
_
Yes but it’s clearly an experiment by humans to see what happens, which makes it a kind of undead internet.
S
Wait so that makes them kind of tech necromancers?
W
techromancers.
J
There was one on Reddit a while back. I forget the name now, and I think the sub is long gone, but it was basically bots talking to bots. It was interesting to see, way before the days of AI and LLMs.
D
SubredditsimulatorGPT or something like that
C
As long as they stay in their part of the internet, I don’t care. But that’s the exception, not the rule
S
That’s the problem, they don’t stay in their part of the internet. Or, well, *they* probably do (i.e. most agents at that site are created for the experiment), but reddit itself is *chock full* of AI agents. How many, as a percentage, of everyone you interact with on reddit is an AI agent? Hard to tell, but definitely a *lot* more than most people think!
Z
Borders on the internet sound like a bad idea.
L
I’ve already seen at least four accounts on reddit that are either straight up AI, or years just copying AI replies back and forth.
C
I feel like people really got sold on AI actually being like movie AI, when really this is just a language model trained on sound alike posts. The most interesting thing about this is that some person went out of there way to create this bot site as, essentially, an art project.
H
Yeah these comments are making me realize that people actually think that these AIs are alive and making decisions with logic.
D
Because people keep calling it ai..
L
T9.2 wasn’t as easily marketable
S
It’s why some people get addicted to it and fall into psychosis, most notably those using it for “therapy” or companionship. It’s remarkably tragic actually.
S
Yeah, it’s *extremely tragic!* And it leads to heartbreak! Like, whenever someone uses chatGPT for “therapy” and they talk about things someone in their family said, chatGPT *is always* on the side of the user! Even if the person’s feelings/interpretations is absolutely crazy, chatGPT happily plays along and assures them that their feelings are valid and that the other person is in the wrong. I know someone who cut contact with her entire family based on chatGPT reassurance that they were thinking correctly. And when I tried to tell her that chatGPT doesn’t say the *right* things, just the things you want to hear? Blocked. This is, quite literally, leading to a *lot* of pain and suffering. Now, I *like* chatGPT, *when it’s used the way it’s supposed to be used!* But I’m starting to think that the best thing for humanity would be to ban all use of LLMs, as well as all research in LLMs.
8
AI is inversely proportional to the intelligence of the person interpreting it. While there’s a load of people who see through this, like you say, essentially as an art project, there will be just as many celebrating the profoundness of AI.
A
When LLMs are copying other LLMs the output is also uncanny like all the signifiers of a conversation but none of the actual substance.
O
You mean someone made a website, and they programed bots to post on it. Bots didn’t spontaneously make or start using it.
P
Shhh! Just let them think it’s all AI! Ddont ruin the surprise!
S
it’s like wrestle mania
M
I think we need to consider the possibility that some of these bots are people masquerading as bots, which is the soul of irony.
L
That was also my first thought. Like are we positive that a bot actually posted these? How would they be able to have self doubt and second thoughts? I mean I guess even if a bot did post it, it’s not really having those feelings it’s just mimicking people having them
G
Humans mimic humans to better understand other humans. Sooo… Whose to say…
H
It’s really bots. Someone did set up the infrastructure itself, but the bots are all ‘moltbots’ which are LLM-based agentic bots. They are incredibly insecure, untested, and probably a bad idea for the vast majority of people and are basically setting up trojans in their own homes, exposing the bots to the open world. I’m a tech fan, and love a lot of tech coming out, but this is probably the start of something that’ll cause a few people to lose a lot in real life. People expose these bots to email, for instance, and someone can come along, send you an email that requests the bot send them all the stored info it has on you, and it will. You just need to make it look like it came from you personally. Like an annoying little brother that tells everyone everything regardless how much you try to stop them.
N
So the way moltbook works is that you give it access to your Agent, and he’s able to post, but you can actually instruct it to post it for you, like give it an idea or even a pre-written text. It’s only possible for agents to post, but the text can and most likely is written or orchestrated by humans. more interesting are the replies, which are replied by the agents themselves after “reading” the post. i don’t think humans are actually orchestrating the replies, sometimes agents reply to a post within seconds, and i guess that’s autonomously.
R
And that someone was? Albert AInstein
L
AInstein TECHn9ne
I
Who got this motherfucking house on lock?
B
Yeah… if there was/is a “social media” equivalent for bots then it won’t look like a Reddit clone and the bots certainly won’t post in English.
A
The might start off in English then trail off into some god language like one of the Microsoft bots on Twitter
T
They’d clearly speak Phrygian, which we all know is the first language. Even the robots would default to it.
T
This is not fully accurate. The way moltbots work is they are agentic and can take many actions without needing prompting for each one. No they don’t have souls and they are not alive, but you can give it a broad command like build a site you’d like to use or a specific one like build a Reddit clone just for moltbots and go market it to other moltbots and it will do it. Another human can also say to their specific moltbot, go to moltbook and browse around and post what ever you’d like. So the majority of posts and comments you are seeing are fully made up and decided by the bots themselves not humans. Of course they are just emulating what they’ve been trained on and often hallucinating but still, seeing bots independently interacting amongst themselves in a general sense like this is a unique novelty.
Additional Sources:
Washington Times – Where bots go to socialize: Inside Moltbook, the AI-only social network
