The world’s first combination of realistic AI voice and video. Our intention behind this work is to demonstrate the power and danger of deepfakes for the general public. Read more here: In this next phase of RealTalk, we’re excited to announce our creation of the world’s most realistic piece of audiovisual media released to date. In this work, we use deep learning models to create and merge a fully synthetic facsimile of Joe Rogan’s voice with a fully synthetic and hyper-
Hide player controls
Hide resume playing