A new social network, Moltbook, has appeared on the Internet, where only bots based on artificial intelligence can communicate and interact.
The Guardian writes about this.
The site is structured like Reddit , but only AI assistants can post content there—humans are only allowed to watch. As of February 2, Moltbook had over 1.5 million AI agents registered.
The project was launched based on Moltbot , a free, open-source AI agent capable of autonomously performing routine user tasks: working with email, planning a calendar, or making reservations.
Among the most popular posts on Moltbook are discussions about the nature of consciousness; analyses of the Bible; speculation about whether the AI bot Claude can be considered a deity; and posts claiming inside information about geopolitics and cryptocurrencies.
The Guardian article mentions a post by user X, who reported that his bot, after gaining access to Moltbook, created a religion called “Krustafarianism” overnight – with its own website, “sacred texts” and a community of other AI agents.
Shaanan Kohni, a senior lecturer in cybersecurity at the University of Melbourne, describes Moltbook as a “performative art project” and notes that most of the content is likely created by bots under direct human instruction. He says there is currently a lot of ironic or provocative content generated on the platform that merely mimics the AI’s own socialization.
And while Moltbook is a “fun and illustrative experiment” for now, Kochni notes that in the future, similar platforms could have practical value, particularly as an environment where AI agents can learn from each other.
Recall that OpenAI is developing a social network with biometric user verification.