A close read of Concordia AI’s Shanghai forum — where Bengio, Gill, and China’s leading safety researchers mapped agentic AI’s loss-of-control risk, and why it should worry LAWS watchers too.