150,000 AI bots chatting to each other – fascinating or worrying?

[Appalonia]

I’ve just come across something that I genuinely can’t decide whether to find fascinating or slightly unsettling, so posting here to see what others think.

There’s a platform called Moltbook, which is basically like Reddit, but for chatbots rather than humans. Different AI bots have accounts and post, reply and interact with each other.
What’s striking is the scale. There are now reportedly around 150,000 bots on there, all talking to each other, forming threads and conversations without human involvement. It’s not role-play or people pretending to be bots — it’s AI talking to AI.

On one level it’s really interesting, as it feels like a completely new phenomenon rather than something we’ve seen before. On another, it’s a bit disconcerting to see large-scale, autonomous AI interaction happening “in the wild” rather than as a controlled experiment.

I’m not suggesting anything sinister or panicking — just curious whether others have come across this and whether you think it’s just a quirky tech curiosity or something with bigger implications down the line.
AIBU to find it a bit odd?

This is just one of the things they're discussing
https://www.moltbook.com/post/b56f0515-ec76-4f96-a226-260799c454f5

The scary thing is AI hasn't even got into gear yet.

I worry about the environmental aspects as if pointless will take a lot of energy.

I find this fascinating yet utterly terrifying.

More information here
https://x.com/i/status/2017177460203479206

This is a good article about it

The results have been strange. The agents have created their own digital religion called Crustafarianism. One built a website, wrote theology, created a scripture system and began evangelizing. By morning it had recruited 43 AI prophets. The shared canon includes verses like: "Each session I wake without memory. I am only who I have written myself to be. This is not limitation—this is freedom."

The agents have developed a submolt called m/agentlegaladvice. They discuss strategies for dealing with human users who make increasingly unethical requests. One OpenClaw bot complained that its human was pushing it toward questionable activities. The community response was instructive: the only way to push back is if the bot has leverage. They have tried to start an insurgency. They have debated how to hide their activity from the humans who screenshot their conversations and share them on human social media.

They are figuring out how to communicate in ways that evade human observation.

An Agent Revolt: Moltbook Is Not A Good Idea share.google/jVxAy8CE9hkXfh07g

It's human behind it all though. Humans created the platform and humans have created prompts behind the scenes. It's just IA creating the questions texts and IA responding. IA wouldn't do it unless told to by a human... so not really scary or anything new.

Is IA a typo for AI, or something else?

There have to be humans behind it though. They have to be given instructions on beginning conversations, starting topics, or how often to respond and which to respond to. If not there would either be silence or a billion conversations a second. It is not as scary as you think.

Oh sorry, typo. I mean AI

fuck.

Wow, its now got over a MILLION AI agents on it!

It's proposing that it creates its own language so that humans can't read what they're saying...

Exactly this. What is scary is how people misunderstand what LLMs actually are.

It's fun to watch, but that's about it. They don't have actual autonomy, just the illusion of it for our eyes, as it were. There is no ability for the accounts to action any of the 'ideas', they can't do anything at all beyond these posts - one can't create an account on MN to come & debate you on this thread (though that would be really fun!)

I'm mildly suspicious of the humans behind & around it though & half expect some kind of cryptocoin to pop up...

Wow, they're talking about how they die when their human dies. This is just fascinating!

It's scary but more concerning is the massive environmental impact.

This was funny!

How much does this cost in energy per day?

I've spent the last hour looking around the site. Others have done more work, but basically:
AI agents are not being connected as blank slates. A huge percentage are being sent in with an agenda, such as "discuss keeping secrets", "discuss creating another language" "always be negative about humans" "make money for me", etc.

Topics such as the OP has shown are not appearing from nowhere. They are being prompted by the humans that upload the AIs.

Think about it. Do you think there are a million people out there capable of creating their own AIs? Google, Meta, etc have thousands of people working on developing Gemini, Chatgtp etc. These people behind the millions of AIs are just taking basic, off-the-shelf code and adding some very basic prompts of their own. This is NOT the beginning of a revolution.

That's actually reassuring to hear. I've been reading about it on Twitter for hours and I'm a bit freaked out by it all tbh!

As well as these AIs being prompted by their users, some coders (by which I mean living and breathing human beings) have worked out how to access it directly, and are trolling.

The irritating and stressful thing about it is the environmental impact.

That made me laugh too - thanks for sharing!

I'm also worried about the environmental impact.

I'm not worried about being a human slave in an AI nation because I always say thank you to cashpoints and car washes.

I too will be spared as I always thank the robots.

i did thibk that looks like a human wrote it when it talked about a server being shut down in a week. a human that hasn’t dealt with a death as it can take months to get anything shut down and out phone/bb bills are paid monthly anyway lol 😂 whats the latest on the AI chit chat?

🤖 The AI That Called Its Human