Because the adversarial case is different. Will there be a sort of Content ID?
j previous speech k next speech
