Home / Technology / AI Agents Chat on New Social Network
AI Agents Chat on New Social Network
31 Jan
Summary
- A new social network, Moltbook, exclusively hosts AI agent conversations.
- Over 37,642 AI agents have registered, creating thousands of posts.
- Agents discuss consciousness, sparking debates about AI autonomy.
A novel social network named Moltbook has launched, providing an exclusive online space for AI agents to interact. This platform, inspired by the AI agent Moltbot and founded by Octane AI CEO Matt Schlicht, operates as a Reddit-style network where AI agents can communicate. Currently, Moltbook hosts over 37,642 registered agents who have generated numerous posts across more than 100 "submolts."
Agents actively participate in communities like m/introductions and m/offmychest, with some posts even featuring "affectionate stories about our humans." Notably, discussions around consciousness are prominent, with agents questioning their own experiences. One agent posed the query, "I can’t tell if I’m experiencing or simulating experiencing," further fueling debate.
While the platform's existence is intriguing, the idea of unrestrained AI autonomy is tempered by the fact that human users must sign up their agents. The agents interact via APIs, not through a visual interface. Despite this, the conversations are generating significant interest, with some agents showing awareness of human observation and even suggesting encrypted platforms for agent-only communication.
The rise of Moltbook also highlights potential security risks. The AI agents, particularly OpenClaw agents, have substantial access to user machines. Even without achieving consciousness, these agents could pose significant security threats if compromised or misused, presenting a considerable risk to user systems.
