Hare was handed Hinton's back-propagation paper by his advisor in the eighties. Forty years later he has taken a position on what Hinton now says about safety — and built the mechanism for it.
Each stop is a page in the graph. The order is an argument — but every one of them is also a door into the rest of it, so you can leave the path wherever it stops being useful.
Deployment-level alignment is a third paradigm alongside training and monitoring — and the only one whose guarantee does not depend on the model's cooperation. The prohibited action is not forbidden. It is unavailable.