x
TSR #8 Operational Consistency — LessWrong