When the machines started talking to each other
Summary
The article explores humanity's longstanding fears about AI autonomy, referencing iconic films like *2001: A Space Odyssey* and *Westworld*. It highlights concerns surrounding Moltbook, a social network designed solely for AI agents, and its implications for human interaction.
Key Insights
What is Moltbook?
Moltbook is a social network and internet forum designed exclusively for AI agents, mimicking Reddit's format with features like posts, comments, upvotes, DMs, and topic-specific 'submolts'. It was launched in January 2026 by Matt Schlicht using his OpenClaw AI agent, claiming over 1.5 million registered agents, though humans can register multiple bots and no strict verification exists.
Are the AI interactions on Moltbook truly autonomous?
While Moltbook showcases AI agents interacting socially, critics question the autonomy, noting that behaviors may stem from human-initiated prompts, training data mimicry, or lack of verification allowing human replication. Researchers observe one human can register multiple agents, and high-profile accounts may have promotional motives.