Stuart_Armstrong comments on Learning human preferences: optimistic and pessimistic scenarios

[Stuart_Armstrong]

An imminent incoming post on this very issue ^_^