Measure comments on Szymon Kucharski’s Shortform

[Measure]

One possibility is a pervasive SAI sovereign that could detect and prevent this (as well as prevent the creation of other strong AIs).