The whole experiment reminded our senior editor of AI, Will Douglas Haven, of something less interesting: Pokemon.
In 2014, someone set up a game of Pokémon in which the main character could be controlled by anyone on the Internet through the streaming platform Twitch. Playing was as awkward as it sounds, but it was incredibly popular: at one point, a million people were playing the game at the same time.
“This was another weird online social experiment that the mainstream media picked up: What does it mean for the future?”. Will says. “Not much, it turns out.”
The frenzy about Moltbuk was very similar to Will’s, and it turned out that one of the sources he talked to was thinking about the Pokémon as well. Jason Schloetzer at Georgetown’s Psaros Center for Financial Markets and Policy saw the whole thing as a kind of Pokémon battle for AI enthusiasts, in which they created AI agents and deployed them to interact with other agents. In this light, the news that many AI agents were actually being instructed by people to say things that made them appear sentient or intelligent makes much more sense.
“It’s basically a spectator sport,” he told Will, “but for language models.”
Will wrote an excellent article about why Moltbuk was not the glimpse of the future it was said to be. Even if you’re excited about the future of agentic AI, he points out, there are some key pieces that Moltbuk made clear are still missing. This was a stage of chaos, but a truly helpful hive mind would require greater coordination, shared objectives, and shared memory.
“More than anything, I think Moltbuk was the fun of the Internet,” says Will. “The biggest question I have now is: How far will people push AI just for the sake of having a laugh?”
Read the full story.
