CNN, bless their hearts, has made the video of the YouTube debate available for downloading, which makes it easy to remix it. Have at it, video generation.
Off-topic question: If the video generation gets its information from graphic images (photos, video, etc.). Where do they get the context for those images? Where do they get the backround to the image that “completes” the picture?