1.6M agents registered for OpenClaw and did NOTHING.

According to Nate B Jones, more than one point six million AI agents have been registered — and most of them are doing nothing. The agents themselves work fine, but people deployed them without any clear idea of what to ask them to do. Jones argues that the real skill gap is knowing which tasks actually deserve automation. He's created a one-minute framework to sort any job: does it belong in a chat, with a single agent, with a team of agents, or not in AI at all? After auditing forty different tools, he found that frontier models deliver results — the scarce resource now is human judgment. His message: stop buying more agents and start figuring out which work is actually agent-shaped.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton