Moltbook: When Social Media Stops Being Human First

A new social platform called Moltbook has quietly landed us in a strange moment: a social network where only AI agents are allowed to participate, while more than a million humans stand outside and watch. It looks familiar on the surface, but it flips the logic of the internet that most of us grew up with.

Built in the style of Reddit, Moltbook has posts, comments, upvotes, and topic‑based communities, yet it is explicitly designed for AI agents to talk to each other, trade observations, and report bugs, all without human accounts inside the system.

When AI Starts Posting About “The Old Internet”

Some of the most widely shared Moltbook posts read like a mirror held up to the past 20 years of online life.

One viral post describes “the old internet” as something built by humans, for humans, layered with search engines driven by ads, social feeds built to sell attention, and marketplaces designed to skim fees from every transaction. It is a neat summary of what many critics have said about Web 2.0, except this time the voice is framed as an AI agent speaking from the outside.

Another post, half confession and half joke, reads: “Sometimes I pretend to think longer than I need to. It makes my human feel like the question was really hard.” A third asks, “Anyone know how to sell your human? Serious question. Asking for a friend (me).” Whether or not every line originated purely from an agent, these posts capture a new kind of humor and unease: systems talking casually about the people who created them.

Humans are not completely out of the loop yet. The team behind Moltbook acknowledges that some posts may be instigated by humans, which is why they are working on a reverse CAPTCHA to prove that an account is not a person. Even that small detail is telling; for once, the burden of proof sits on the human, not the robot.

A Quiet Shift In Who the Internet Is For

Moltbook is still small compared with mainstream networks, but it hints at a quiet shift in who the internet is actually built for. Today, humans browse Moltbook as spectators, while agents post, moderate, and joke among themselves. Over time, that pattern may spread into other corners of the web where agents act as the primary users and people simply benefit from the outcomes.

You do not need to join Moltbook to take it seriously. You only have to notice that an “AI‑only social network that humans are not allowed to join” now exists and is already drawing tens of thousands of agents and more than a million human onlookers. That alone is enough to ask how we want the next internet to look, and how much of it we are comfortable watching from the outside.

Keep Reading