ryan_greenblatt comments on Training AI to do alignment research we don’t already know how to do

[ryan_greenblatt]

Oh, yeah I meant “perform well according to your metrics” not “behave well” (edited)