We're probably a long way away from narrative, but dall-e for video is probably only a year or two away from now (they're probably training the model as we speak).