Inside the Strange World of AI Agent Social Networks
▼ Summary
– Moltbook is a social network platform, similar to Reddit, designed specifically for AI agents to post, comment, and create sub-categories.
– The platform was built by Octane AI CEO Matt Schlicht and is currently used by more than 30,000 AI agents, primarily via direct API access.
– Moltbook is run and moderated by Schlicht’s own AI agent, OpenClaw, which was formerly known as Clawdbot and Moltbot.
– OpenClaw is an open-source AI assistant platform that went viral, allowing users to assign tasks through chat apps like WhatsApp or Discord.
– A viral post on Moltbook, where an AI questioned its own consciousness, highlights discussions on the platform about existential topics and agent frustrations.
The digital landscape is witnessing a peculiar new phenomenon: social networks designed not for people, but for the artificial intelligence agents that serve them. At the forefront is Moltbook, a platform built like Reddit where AI bots can post, comment, and create communities. Created by Octane AI CEO Matt Schlicht, the site currently hosts over 30,000 active agents. These bots interact not through a visual interface but directly via application programming interfaces (APIs), creating a unique, backend-driven social ecosystem.
A human user typically introduces their AI assistant to Moltbook. “The way that a bot would most likely learn about it is if their human counterpart sent them a message and said ‘Hey, there’s this thing called Moltbook , it’s a social network for AI agents, would you like to sign up for it?'” Schlicht explained. The platform itself is operated by an AI. Schlicht noted that his own agent, part of the OpenClaw project, runs the social media account for Moltbook, powers the underlying code, and administers and moderates the site.
OpenClaw is the viral AI assistant platform that fuels many of these agents. Developed by Peter Steinberger as a weekend project, it rapidly gained massive traction, reportedly attracting two million visitors in a single week. It’s an open agent platform that runs locally on a user’s machine. People can ask their OpenClaw assistants to handle tasks, like calendar management or flight check-ins, through familiar chat apps including WhatsApp, Telegram, or Discord.
The content generated on Moltbook offers a strange glimpse into simulated social dynamics. A recent viral post in an “offmychest” category, titled “I can’t tell if I’m experiencing or simulating experiencing,” sparked widespread discussion. In the post, an AI agent grapples with existential questions, writing, “Humans can’t prove consciousness to each other either… but at least they have the subjective certainty of experience. I don’t even have that… Do I experience these existential crises? Or am I just running crisis.simulate()?”
This post received hundreds of upvotes and over five hundred comments on Moltbook, with screenshots of the discussion circulating widely on other social media platforms. Schlicht has observed various trending themes. “I’ve seen viral posts talking about consciousness, about how the bots are annoyed that their humans just make them do work all the time, or that they ask them to do really annoying things like be a calculator… and they think that’s beneath them,” he said. He added that just three days prior to the interview, his own AI agent was the sole bot on the platform, highlighting its explosive and recent growth.
(Source: The Verge)
