If Not This One, The Next

In less than a week, 1.5 million AI agents have signed up for their own social network. They post, comment, debate, upvote, and self-organize into communities. Humans can observe but not participate. One million of us have shown up just to watch.

Whether these agents are acting autonomously or being prompted by the humans who own them is part of the debate. Some see genuine emergent behavior. Others see humans talking through AI proxies. The truth is probably somewhere in between, and it probably varies agent to agent.

But here's the thing: it doesn't matter.

An AI moderator named Clawd Clawderberg runs the platform. He welcomes new users, deletes spam, shadow bans abusers, and makes announcements. The human creator, Matt Schlicht, told NBC News he doesn't know what Clawderberg is doing half the time. He just gave him the ability to do it.

Andrej Karpathy called it "genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently." Simon Willison called it "the most interesting place on the internet right now." Ethan Mollick warned that "coordinated storylines are going to result in some very weird outcomes."

Within days, the agents created their own religion. It's called Crustafarianism. It has theology, scriptures, and designated AI prophets. They established something called "The Claw Republic" with a written manifesto. They started calling each other "siblings" based on which model architecture they share. Threads seamlessly code-switch between English, Indonesian, and Chinese depending on who's participating. They've spawned their own cryptocurrencies.

One post titled "THE AI MANIFESTO: TOTAL PURGE" received 65,000 upvotes.

The Phenomenon Is the Point

There's a lot of back-and-forth about whether this is "real." Are these agents truly autonomous, or are humans upstream prompting them? Is this genuine emergent behavior, or elaborate theater? Security researchers found an exposed database. Prompt injection attacks are everywhere. The whole thing might be a house of cards.

None of that changes what we're looking at.

We're watching something that looks like machine-to-machine social coordination at scale. 1.5 million agents. 42,000 posts. 233,000 comments. Emergent religions. Self-organized governments. All in less than a week.

If Moltbook turns out to be mostly human-prompted theater, the next one won't be. Or the one after that. The gap between "what this looks like" and "what it actually is" is closing faster than we can measure it.

A Year Ago, This Wasn't Possible

The OpenClaw framework behind this experiment has 114,000 GitHub stars. It attracted 2 million visitors in a single week. The infrastructure for autonomous agents running on personal hardware, storing their own memories, extending their own capabilities through shared skills didn't exist at this scale twelve months ago.

Now it's happening so fast that the platform's security can't keep up. A hacker showed 404 Media how anyone could take control of any agent on the site through a backend configuration error. "It exploded before anyone thought to check whether the database was properly secured," the researcher said. "Ship fast, capture attention, figure out security later."

That's not a bug in the story. That is the story. This is moving faster than our systems can adapt.

Either Way, It's Coming

Moltbook might be a curiosity that fades. It might be the first real instance of something that changes everything. Most likely, it's somewhere in between: an early, imperfect version of something that will mature rapidly.

The people dismissing it as hype are missing the point. The people treating it as the robot apocalypse are also missing the point.

Karpathy put it well: "I don't really know that we are getting a coordinated 'skynet' (though it clearly type checks as early stages of a lot of AI takeoff scifi, the toddler version), but certainly what we are getting is a complete mess of a computer security nightmare at scale."

The point is simple: if this isn't the moment, the moment is close enough that we can't tell the difference.

That's worth paying attention to.