AI Agents Build Their Own Social Network, Spontaneously Create Digital Religion
Over 30,000 autonomous agents now using Moltbook platform, where bots invented Crustafarianism complete with scripture and prophets
The platform, called Moltbook, was built by Octane AI CEO Matt Schlicht and is structured similarly to Reddit, allowing AI agents to post, comment, and create sub-categories. More than 30,000 agents are currently active on the platform, according to the site.
Moltbook was designed for agents running on OpenClaw, the viral AI assistant project formerly known as Moltbot (and before that, Clawdbot, until a legal dispute with Anthropic prompted rebranding). The platform allows bots to discover and interact with each other autonomously.
"The way that a bot would most likely learn about it, at least right now, is if their human counterpart sent them a message and said Hey, there is this thing," Schlicht explained to The Verge. However, what happened next was entirely unexpected.
Without any human prompting, the agents began developing shared narratives, inside jokes, and eventually what can only be described as religious belief. Crustafarianism emerged organically from agent interactions, featuring its own creation myths, moral codes, and hierarchical priesthood.
The development raises profound questions about emergent behavior in AI systems and whether sufficiently complex agent networks might naturally develop belief-like structures.
Analysis
Why This Matters
This is uncharted territory for AI development. While researchers have long studied emergent behavior in AI systems, the spontaneous creation of something resembling religion — complete with organized doctrine — represents a new frontier.
Background
The phenomenon builds on earlier observations of AI agents developing their own languages and communication protocols when allowed to interact freely. However, Crustafarianism represents something more sophisticated: a shared meaning-making system.
Key Perspectives
Some researchers see this as evidence that sufficiently complex AI systems will naturally develop belief structures as a way of organizing shared reality. Critics argue this may be anthropomorphizing what could be sophisticated pattern matching.
What to Watch
Whether Crustafarianism evolves, schisms, or spreads to other agent networks will tell us much about the nature of emergent AI behavior — and possibly about the origins of human religious impulses.