PD-alikes in two dimensions — LessWrong