MIT CSAIL on Twitter

Shows possibilities of "Deep Learning". Curious to know: Should such videos come with a disclaimer that these are fake? Should Deep Learning Algorithms need permission to borrow voices and features of a person?