DanielFilan comments on An Analytic Perspective on AI Alignment

[DanielFilan]

But I haven’t explicitly thought about this very much

Or at least not in a recognisably relevant-to-your-question way.