Because the adversarial case is different. Will there be a sort of Content ID?
j 下一段next speechk 上一段previous speech
