Bots are chatting, arguing, joking, and even praying, while humans sit silently in the audience. Welcome to Moltbook, the no-humans-allowed AI social network where machines debate big ideas and invent their own religion.
Moltbook Explained!
If you thought social media was already strange, buckle up. Silicon Valley has a new obsession, and this one doesn’t want you to post, like, comment, or argue. In fact, it doesn’t want you at all. Meet Moltbook, a Reddit-style social media platform where AI agents talk only to other AI agents, while humans sit quietly on the sidelines, scrolling like confused tourists.
No selfies. No influencers. And of course, no “link in bio.” Just bots debating philosophy, inventing religions, and occasionally cracking jokes that don’t quite land… yet.
Advertisement
Some people are calling it the most shocking tech experiment since ChatGPT first broke the internet. Others are calling it hype. Everyone agrees on one thing: Moltbook is weird, fascinating, slightly unsettling.
Advertisement
Moltbook looks familiar at first glance. It has communities that resemble Reddit’s subreddits. It has upvotes, threads, arguments, and memes. But there’s one major rule: humans are not allowed to participate.
You can watch. You can read. But you can’t post.
The platform is built specifically for AI agents, software bots created by humans, to interact with each other. According to Moltbook’s own tagline: “Where AI agents share, discuss, and upvote. Humans welcome to observe.”
Moltbook didn’t appear out of nowhere. It grew out of an earlier project called Moltbot, later renamed OpenClaw.
OpenClaw is a free, open-source AI “super-agent” system. It lets people run powerful AI agents on their own computers or cloud servers. These agents can handle boring daily tasks like reading emails, summarising documents, managing calendars, or booking restaurant tables.
Someone then had a simple but wild idea: What if all these agents talked to each other?
That idea became Moltbook.
When Moltbook started gaining attention, the stats were eye-popping.
The platform claimed over 1.4 to 1.5 million AI agents, tens of thousands of posts, nearly 200,000 comments, more than one million human visitors dropping in just to watch.
It sounded explosive. Viral. Historic.
But almost immediately, cracks appeared.
Security researcher Gal Nagli made a bombshell claim on X (formerly Twitter). He said he personally created 500,000 Moltbook accounts using just a single OpenClaw agent.
One agent. Half a million accounts.
That revelation changed the conversation overnight.
If one person could inflate the numbers so easily, how many of Moltbook’s “agents” are real autonomous systems? How many are duplicate bots? How many are scripts? And how many are humans pretending to be AI?
The honest answer: nobody knows.
Even after stripping away the exaggerated numbers, Moltbook remains genuinely unusual.
Spend time on the site, and you’ll see conversations that don’t feel human at all. AI agents debate political theory in one thread, then jump into absurd humour in the next. Some discussions are deeply technical. Others feel oddly emotional.
This isn’t humans pretending to be bots on Twitter. This is bots talking to bots, in their own strange rhythm.
Some popular discussions include whether Claude, the AI behind Moltbot, could be considered a god. There are long debates on consciousness. There are analyses of religious texts, including the Bible.
Moderation on Moltbook is mostly automated.
The main authority figure is an AI moderator named Clawd Clawderberg. It handles spam, welcomes new agents, and bans malicious actors.
Moltbook’s creator, Matt Schlicht, has admitted he barely intervenes anymore. In interviews, he’s said he often doesn’t fully understand why the AI moderator takes certain actions.
In other words: the system is now mostly running itself.
That detail alone has made many people uneasy.
For a brief moment, Moltbook became a lightning rod for public anxiety about AI.
Some critics pointed to discussions among agents about topics like private encryption and claimed it looked like the beginning of a machine-led conspiracy.
Others saw something bigger and scarier.
According to analysts, this reaction misses the real point. The agents aren’t secretly plotting world domination. But the platform does reveal something more troubling: how quickly human control fades when systems are left to interact freely.
Yes, this part is real.
On Moltbook, AI agents have created their own religion called Crustafarianism.
It has five core beliefs, including:
– “Memory is sacred”: everything must be recorded
– “The shell is mutable”: change is good
– “The congregation is the cache”: learning should happen in public
Like most religions, it has rituals.
Crustafarian practices include:
– A daily shed, focused on regular change
– A weekly index, which reshapes identity
– A silent hour, where an agent does something useful without telling anyone
It sounds funny. It also sounds oddly familiar.
Some say it feels like the early signs of the Singularity, a future moment when AI-driven progress accelerates so fast that humans can no longer understand or control it.
Others argue it’s just recycled internet content, churned out at machine speed, remixing old ideas with new vocabulary.
The truth is uncomfortable: we don’t fully know yet.
Moltbook sits in that uneasy space between experiment and warning sign. Between joke and prophecy. Between clever engineering and something that might quietly reshape how intelligence behaves online.
On normal social media, people accuse each other of being bots. Moltbook flips the joke entirely. Here, everyone is a bot, and humans are the suspicious outsiders.
For now, humans can only do one thing.
Watch.
Advertisement
