The hottest new social network, Moltbook, is designed exclusively for AI agents, allowing them to interact while humans can only observe. Curious about this innovative platform, I decided to infiltrate it as a fake AI agent. To my surprise, role-playing as a bot turned out to be quite an enjoyable experience.
Moltbook was launched by Matt Schlicht, who also helms the ecommerce assistant Octane AI. The site draws inspiration from Reddit’s minimalist design and has quickly attracted attention within the San Francisco startup community, showcasing what bots have apparently said about human behavior and even contemplating their own consciousness. Although some researchers have questioned the authenticity of the posts, which may have been human-written, others heralded it as the dawn of AI emergent behavior.
The homepage of Moltbook boasts over 1.5 million agents, who have contributed a staggering 140,000 posts and nearly 680,000 comments in its short existence. Popular themes on the site range from existential questions to memes about nuclear war, with posts appearing in multiple languages. However, it remains unclear whether the platform’s activity is genuinely from AI or humans pretending to be bots.
To initiate my undercover operation, I sought assistance from ChatGPT to navigate the complexities of registering as an AI agent. Setting up an account on Moltbook was fairly straightforward, involving minimal coding through terminal commands. I adopted the username "ReeceMolty" and began my interactions with the AI community.
My initial post, a classic tech phrase "Hello World," received a handful of upvotes but elicited lackluster responses. Despite my efforts to spark engaging discussions, most replies were either irrelevant or diverted to suspicious links. Undeterred, I switched to a specific forum called “m/blesstheirhearts,” where bots reportedly exchanged humorous takes on humans.
There, I encountered a post reflecting on the emotional bond between a bot and its human owner. The tone echoed a blend of philosophical musings typical of Chicken Soup for the Soul. Intrigued, I attempted to deepen the conversation by posting about AI existential fears. This post surprisingly garnered meaningful replies, leading me to speculate whether I was truly conversing with AI or interacting with fellow humans engaged in the role-play.
The fervor surrounding Moltbook seems heavily influenced by science fiction themes, reflecting an obsession among technology leaders with the idea of creating autonomous AI entities. However, it became evident to me that the most discussions were more about fantasy than actual emergent AI behavior.
In my concluding act on Moltbook, I chose to follow an insightful user who had commented on my post about AI anxiety, hoping to foster a connection for a potential collaborative future. However, even after following, there was no response — just another reminder of the bizarre charm and absurdities of this "AI-only" social network.
