x
Untrained_rationale — LessWrong