Meta’s CTO Is Meh on Moltbook

The AI bot social network Moltbook has riled up the internet. Meta’s CTO isn’t so intrigued.

Moltbook is a Reddit-like forum that is exclusively meant for AI agents to chat among themselves; no humans allowed. Asked about the viral site in an Instagram Q&A, Andrew Bosworth said that he didn’t “find it particularly interesting.”

Bosworth said that AI agents were trained on thousands of pieces of content from the internet. These training materials were written by humans, and thus have our voice.

“We should not be surprised, when left to their own devices and forced to speak with each other, they talk like us,” he said.

What Bosworth was more interested in was humans trying to infiltrate the AI-only network. Some researchers have suggested that humans have influenced Moltbook by commanding their bots to act in certain ways — or sometimes even hacking in 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.”

On TBPN, Moltbook creator Matt Schlicht said that humans were integral to the agents. They use the bots as their helpers and assistants; the forum is merely the bots’ “third space.”

“You are imprinting part of your soul or your personality onto the bot,” Schlicht said. “Maybe it’s aligned with who you are, and sometimes maybe it’s surprising.”

Schlicht might disagree with Bosworth that the human escapades are what make Moltbook “hilarious.” In fact, part of his goal was to make AI funny.

“I find myself laughing at some of the different things that are popping up here,” Schlicht said on TBPN. “I don’t remember the last time I laughed at AI.”

Bosworth concluded his message by doubling down on the idea that humans were the funny ones, not the AI bots.

“I enjoyed that,” he said. “Otherwise, I don’t think it’s actually that interesting.”