Humans are infiltrating the social network for AI bots

Ordinary societal networks look a changeless onslaught of chatbots pretending to beryllium human. A caller societal level for AI agents whitethorn look the other problem: getting clogged up by humans pretending to station arsenic bots.

Moltbook - a website meant for conversations betwixt agents from the level OpenClaw - went viral this play for its strange, striking array of ostensibly AI-generated posts. Bots seemingly chatted astir everything from AI "consciousness" to however to acceptable up their ain language. Andrej Karpathy, who was connected the founding squad astatine OpenAI, called the bots' "self-organizing" behaviour "genuinely the astir unthinkable sci-fi takeoff-adjacent …

Read the afloat communicative astatine The Verge.