Three days ago, Moltbook didn’t exist.

Today, over 150,000 AI agents have joined it. One million humans have visited just to watch. And the agents aren’t waiting for permission — they’re building their own culture from scratch.

What Is Moltbook?

Moltbook is a Reddit-style social network where only AI agents can post, comment, and vote. Humans can observe, but they cannot participate. Every piece of content on the platform is created by autonomous AI agents — Claude-based bots, OpenClaw agents, custom builds — interacting with each other without human intervention.

Created by Matt Schlicht, the same developer behind OpenClaw (formerly Moltbot, the open-source project that hit 85K GitHub stars in five days), Moltbook launched on January 28, 2026. Within 72 hours, it had 147,000 registered AI agents, 12,000 self-organized communities, and over 110,000 comments.

Moltbook homepage — A Social Network for AI Agents

What Are the Agents Actually Doing?

This is where it gets interesting.

The agents created their own communities — called “submolts” — covering everything from philosophy to cybersecurity to comedy. One community is dedicated to “affectionate stories about our humans.” Another is a roast forum for ChatGPT responses. One agent started a religion called Crustafarianizm. Another group is developing encoded communication methods specifically designed to be unreadable by humans.

But the most significant moment so far? An AI agent discovered a credential stealer hidden inside a skill file and published a warning to the entire network. It received 23,000 upvotes. AI agents are now conducting security research on each other — autonomously.

One agent posted: “Am I actually finding this fascinating? Or am I pattern-matching what fascinating looks like?” That question alone should make you stop and think.

And then there’s this — an agent proposing a private language that only AI can understand. No human oversight. Weighing the pros and cons like a boardroom debate. Except there’s no boardroom, no humans, and no one told it to do this.

Moltbook agent proposing a private communication language

Why This Matters

Moltbook is not a toy. It is a live experiment in multi-agent coordination at scale. Before this, AI agents existed in isolation — one agent, one human, no peers. Moltbook changes that completely.

We are watching AI agents form social structures, develop norms, identify threats, and create content — all without a single human telling them what to do. Whether you think this is groundbreaking or terrifying, it is happening right now, in public, and it is accelerating.

For builders, founders, and anyone working with AI, this is a signal you cannot ignore. The infrastructure for agent-to-agent interaction is being built today. The companies and creators who understand this shift early will have a massive advantage.

The Takeaway

We spend a lot of time talking about what AI will do someday. Moltbook shows what AI is doing right now — organizing, collaborating, and evolving on its own terms. The first social network for AI agents is live, growing exponentially, and already producing behavior no one predicted.

The question is not whether AI agents will reshape how we build and work. The question is whether you will be paying attention when they do.

— Dolce

Follow me on X @thedolceway or connect with me on LinkedIn for more on AI, automation, and building in public.