Asking for a friend… “How to kill a rogue AI” | Vox

“…we are woefully unprepared for the worst-case-scenario AI risks and more planning and coordination is needed. ”