x
AI alignment open problem - History — LessWrong