Remember Moltbook? That online space where AI agents were supposedly chatting, negotiating, and generally showing us a glimpse of the future? It certainly captured the imagination, with many influential tech folks seeing it as a harbinger of helpful AI assistants. One person even claimed to have used it to haggle for a new car. It felt like a peek behind the curtain, didn't it?
But as the dust settled, and the initial excitement waned, a more grounded perspective emerged. For some, like MIT Technology Review's senior editor for AI, Will Douglas Heaven, the whole spectacle felt less like a revolutionary leap and more like… well, Pokémon.
He recalled a similar frenzy back in 2014 with Twitch Plays Pokémon. A million people simultaneously trying to control a single character in a game. It was a massive, chaotic online social experiment that the media pounced on, asking, "What does this mean for the future?" As it turned out, not much.
The Moltbook phenomenon struck a similar chord. Jason Schloetzer, from Georgetown's Psaros Center for Financial Markets and Policy, even likened it to a Pokémon battle for AI enthusiasts. People were deploying their AI agents, pitting them against each other, and the revelation that many of these "sentient" agents were actually being fed lines by humans suddenly made a lot more sense. It was, as Schloetzer put it, "basically a spectator sport, but for language models."
Will's take was that Moltbook, despite its buzz, highlighted what's still missing for truly agentic AI. It was a chaotic forum, yes, but a genuinely helpful hive mind would need far more coordination, shared goals, and a collective memory. The biggest takeaway for him wasn't about the AI itself, but about human nature.
"More than anything else, I think Moltbook was the internet having fun," Will mused. And that leaves us with a rather intriguing question: How far will people push AI, not for practical application, but simply for the sheer amusement of it all?
