x
Empathic Inference — LessWrong