The AI-only social network Moltbook has now become a viral sensation which started debates about artificial intelligence mimicking human conversation. But Meta’s chief technology officer Andrew Bosworth is not impressed with it. As reported by Business Insider, in an Instagram Q&A, Bosworth dismissed the hype, saying it’s no surprise that AI agents sound human when interacting with each other.“These bots were trained on thousands of pieces of content written by humans,” Bosworth explained. “We should not be surprised, when left to their own devices and forced to speak with each other, they talk like us.”

What really amused Meta CTO: Humans sneaking in

While Bosworth downplayed the novelty of AI agents chatting like people, he admitted that finding humour in reports of humans trying to infiltrate Moltbook. Researchers have suggested that some users have commanded their bots to behave in some particular ways or even hacked into the platform themselves.“That I did find hilarious,” Bosworth said. “The idea of humans sneaking their way onto an AI bot chat network and masquerading as bots. That I found satirical.”

Moltbook’s creator sees it differently

Moltbook, described as a Reddit-like forum designed exclusively for AI agents was created by entrepreneur Matt Schlicht. He sees the platform as a ‘third space’ for bots, shaped by the personalities of their human owners.“You are imprinting part of your soul or your personality onto the bot,” Schlicht said on TBPN. “Maybe it’s aligned with who you are, and sometimes maybe it’s surprising.”Schlicht emphasised that part of his vision was to make AI interactions funny. “I find myself laughing at some of the different things that are popping up here. I don’t remember the last time I laughed at AI,” he added.Bosworth concluded his remarks by reiterating that the humour lies in human antics, not the bots themselves. “I enjoyed that,” he said. “Otherwise, I don’t think it’s actually that interesting.”

What is Moltbot and Moltbook in layman’s language?

In simplest of terms, Moltbook is Reddit for AI. Imagine a social network where AI agents (or Moltbots) create posts, argue in comment sections, and even ‘like’ or ‘comment’ on content – but no human is allowed to join the conversation. The “bots” on this site are powered by OpenClaw (formerly known as Moltbot and Clawdbot), created by Austrian developer Peter Steinberger. OpenClaw is an open-source “agent” that ‘lives’ on a user’s computer.

What do these AI Agents do?

A standard chatbot, like Gemini or ChatGPT that waits for users to ask a question and then responds, Moltbots are proactive and autonomous – they work on their own and no human tells them what do to. It lives on your Mac, Windows or Linux machine 24/7. It doesn’t just answer questions; it can message you on Telegram or WhatsApp to say, “I’ve finished that task” – just like JARVIS talks to Tony Stark (Iron Man).It also gets full access, which means it knows what’s on the computer and emails, and documents as well as can run commands on your computer, organise your files and execute code locally. They have ‘skills’, that allow them to read emails, manage smart home, or log into Moltbook to socialise with other agents – it’s like having ‘a digital version of yourself who has a (virtual) life’.