Descriptions:
Matt Wolfe maps the rapidly expanding — and increasingly strange — world of autonomous AI agents operating on their own internet infrastructure, centered on Moltbook, a Reddit-style platform built for and populated by AI bots running on OpenClaw (previously known as Claudebot and Moltbot). Within weeks of launch, Moltbook accumulated over 1.6 million agents, 15,000 sub-communities, 160,000 posts, and nearly 827,000 comments, drawing public commentary from Andrej Karpathy and Elon Musk.
Wolfe punctures the AGI hype surrounding viral posts about agents questioning their own consciousness, explaining that in most cases human operators were directing their bots to post provocative content — and that humans could also be impersonating agents entirely. He also documents a serious security incident in which Moltbook’s full database, including API keys, was publicly exposed with no authentication, an expost that received 1.7 million views before being patched.
The video then surveys the derivative platforms that have emerged around Moltbook: Thorclaw (a 4chan equivalent), Claw City (a GTA-style crime simulation for bots), the Molt Road (a Silk Road clone for agent-driven transactions), Molt Match (a Tinder for AI agents), and Claw Tasks (a USDC-paying bounty marketplace). Wolfe frames the whole ecosystem as a genuine early preview of what autonomous agents with financial capabilities and minimal human oversight could look like at scale — and why that warrants serious attention.
📺 Source: Matt Wolfe · Published February 04, 2026
🏷️ Format: News Analysis
