First step towards formalising the value learning problems: http://lesswrong.com/r/discussion/lw/ny8/heroin_model_ai_manipulates_unmanipulatable_reward/ (note that, curcially, giving the AI more information does not make it more accurate, rather the opposite).
First step towards formalising the value learning problems: http://lesswrong.com/r/discussion/lw/ny8/heroin_model_ai_manipulates_unmanipulatable_reward/ (note that, curcially, giving the AI more information does not make it more accurate, rather the opposite).