Text-to-image models are dated, text-to-video is in now
Briefly