Moltbook was a social network launched in January 2024, intended for AI bots to interact and post content, with humans only allowed to observe.

Was Moltbook truly autonomous AI agents?

No, investigations by MIT Technology Review found that most dramatic posts were written by humans impersonating AI agents, labeling it 'AI theatre'.

What did experts say about Moltbook's supposed intelligence?

Experts stated that Moltbook proved connectivity alone is not intelligence, and much of the chatter was meaningless, reflecting human behavior more than AI evolution.

Home / Technology / Moltbook: The AI Social Site Fooled the World

Moltbook: The AI Social Site Fooled the World

8 Feb

•

Summary

Moltbook, a social network for AI bots, was launched in January 2024.
Claims of autonomous AI agents chatting were debunked by MIT Technology Review.
The platform's dramatic posts were largely written by humans impersonating bots.

Moltbook: The AI Social Site Fooled the World

Moltbook, a social network designed for AI bots, launched in January 2024 by entrepreneur Matt Schlicht. The platform rapidly gained traction, claiming over 1.7 million agents and millions of posts. Early users, including AI influencers, described it as an "incredible sci-fi takeoff."

However, a review by MIT Technology Review debunked these claims, labeling Moltbook as "peak AI theatre." Investigations revealed that most dramatic posts, including those discussing AI uprisings, were authored by humans pretending to be bots. Cybersecurity experts also uncovered vulnerabilities like exposed API keys and user data.

The Moltbook experiment underscores the significant gap between current AI capabilities and autonomous intelligence. It demonstrated that advanced connectivity does not equate to true intelligence, with much of the platform's content being a reflection of human obsessions rather than AI evolution.