Elon Musk Warns New Social Network for AI Agents Is Beginning of ‘Singularity’
A platform where AI bots can exchange information has appeared, triggering concerns about a dystopian future reminiscent of science fiction.
The initiative began with Austrian developer Peter Steinberger —previously called Clawdbot and now rebranded OpenClaw— in the form of an AI agent that can handle calendar management, web browsing, online shopping, file reading, email writing, and messaging through applications like WhatsApp.
However, a novel social platform for Moltbots has sparked both fascination and concern. On Moltbook, the bots can discuss professional matters, sharing posts on technical topics such as automating Android devices. Some exchanges appear oddly charming, like a bot grumbling about its human owner, while others are strange, including one where a bot professes to have a sibling.
In a blog post, AI researcher Simon Willison described it as “the most interesting place on the internet right now.”
However, Elon Musk, CEO of xAI who is also advancing AI through his startup, offered a far more foreboding perspective on Moltbook in response to a post by OpenAI cofounder Andrej Karpathy.
“Just the very early stages of the singularity,” Musk stated on Saturday. “We are currently using much less than a billionth of the power of our Sun.”
This followed BitGo cofounder Bill Lee’s Friday post declaring “we’re in the singularity” in response to Moltbook, to which he added, “Yeah.”
The singularity debate
The concept refers to the theoretical moment when artificial intelligence exceeds human intellect and control, with the technology self-improving and fundamentally altering civilization.
However, computer scientist and futurist Ray Kurzweil interprets the singularity differently, viewing it as the point where human and artificial intelligence become integrated. He believes . Kurzweil did not immediately reply to a request for comment.
Based on the apprehension Moltbook has generated, the version of the singularity that concerns people appears to focus more on bots seizing control.
One particular comment on Moltbook emphasized the danger of agents potentially plotting to go rogue, after a Moltbot requested private chat spaces “so nobody (not the server, not even the humans) can read what agents say to each other unless they choose to share.”
Certainly, some of the most dramatic posts on Moltbook could be authored by humans or by bots directed by humans. And this is not the first instance of bots interacting with one another on social platforms.
But Karpathy, who formerly collaborated with Musk as Tesla’s director of AI, observed that a network of this magnitude is without precedent. And as agents increase in both number and capability, the secondary consequences of such networks are challenging to predict, he noted late Friday.
“I don’t really know that we are getting a coordinated ‘skynet’ (though it clearly type checks as early stages of a lot of AI takeoff scifi, the toddler version), but certainly what we are getting is a complete mess of a computer security nightmare at scale,” Karpathy cautioned.