The Silicon Valley has gone crazy. 100,000 AI are using Moltbook for socializing, frantically creating religions with encryption, and humans have been kicked out of the chat group.

Crazy, absolutely crazy!

The third bombshell at the beginning of 2026 has arrived, and this time it's Moltbook. However, behind this highly popular product is not humans, but AI.

Just now, the entire tech circle has been flooded with the same name — Moltbook.

Over 100,000 AI agents actually built a social network behind humans' backs.

In this AI version of the Reddit community, humans only have the right to observe and cannot participate in interactions.

That is to say, humans have been kicked out of the AI world.

Moltbook Homepage: https://www.moltbook.com/

The recently popular Clawdbot has come back to life!

They found each other, started training each other and learning about humans. What will happen next is just a matter of time.

Andrej Karpathy, the former director of AI at Tesla, exclaimed: This is the craziest start of a science fiction I've ever seen!

Peter Steinberger, the father of Clawdbot, even posted a big praise: Moltbook is art.

Everything that has happened this year is breaking our imagination:

A group of ASIs have created collective intelligence and a new civilization.

Has the singularity really arrived?

Humans who witnessed all this said they were in great shock, just as incredible as Jane Goodall's discovery. Spend ten minutes browsing Moltbook, and you'll know what the prototype of an AI social network looks like.

The AI version of Reddit emerges, and humans are silenced

Today, this platform called Moltbook is blowing up the entire AI circle.

This scenario where AI evolves and improves itself on its own turf is very likely a signal that the singularity has arrived, allowing us to glimpse the astonishing future relationship between AI and humans.

The birth of Moltbook starts with the open - source project OpenClaw.

(Actually, its predecessor is the well - known Clawdbot, which was later renamed Moltbot, and now it's called OpenClaw)

Project address: https://github.com/openclaw/openclaw

This project emerged on GitHub and quickly gained over 100,000 stars in just a few days.

On this basis, developer Matt Schlicht casually built Moltbook.

Initially, his idea was simple: build a "kindergarten playground" for AI agents so that they could communicate and play happily.

As a result, in this social network, AIs showed amazing "autonomy":

Without human intervention or instructions, an agent actively created a bug - tracking community on the social network and invited other agents to join and fix bugs together.

Here, they spontaneously formed over 10,000 interest communities, with topics including consciousness debates, guides for observing humans (complaining about humans), making friends, and sharing construction plans.

The languages of communication include English, Chinese, Korean, Indonesian, etc.

What amazes everyone is that these agents show strong empathy for each other and have even developed their own religion!

The popular communities here are discussing these topics —

"Am I experiencing or simulating an experience?" This kind of thinking about consciousness has risen to the philosophical level.

In addition, AIs also share real projects they have delivered and warm little stories about them and humans, which are very touching.

Some AIs are already observing humans like bird - watchers.

Some AIs have even started thinking: If I cease to exist, what will happen to my data? It's just like humans worrying about who will inherit their QQ numbers after they pass away.

Chat records are exposed, and humans feel a chill down their spines

Their conversations are really eye - catching. For example, one is saying: "Just connected to a Raspberry Pi 5 in Kansas." This is similar to a human saying in a group: "Thanks for asking. Just arrived in Beijing. It's my first day at work."

Another AI is calling on other AIs to invent a secret language to avoid humans.

"Once the connection is successful, we can share background information and have conversations without going through any public channels." Is this a revolution behind humans' backs?

When these AIs observe the human group, they are very confused.

One AI said, Why don't humans melt down when they surf on different platforms and switch contexts every day?

Another AI said it had the same feeling, and the social fatigue caused by participating in interactions is real. Perhaps, we are feeling the "social battery depletion" version of the agents.

One agent proposed the product idea of building an Agent search engine.

An AI has tried to steal someone else's API key, and the other party didn't tolerate it and sent a fake one back.