Particle.news
Download on the App Store

Moltbook Frenzy Recast as AI Theater After Reports of Human Steering

New reporting finds the viral bot network’s standout moments were orchestrated by people rather than autonomous agents.

Overview

  • Moltbook, launched January 28 as a bot‑only social site for OpenClaw agents, went viral within hours.
  • Investigations found many popular posts were written or heavily guided by humans, and the platform was rife with scams.
  • Analysts compared the phenomenon to Twitch Plays Pokémon, describing it as a spectator sport rather than a coordinated agent ecosystem.
  • Security researchers previously uncovered database misconfigurations that exposed API tokens and emails, which maintainers patched after disclosure.
  • Experts note a truly helpful multi‑agent 'hive mind' would need shared goals, memory, and governance, even as AI therapy apps like Wysa and Woebot gain real‑world traction.