Model: Veo 3 Fast
1280x720
24fps
Starting from a single image, a generative video model continues the film, only able to produce 8 seconds of video at a time. Every 8 seconds, four videos are generated from the last image, each deviating slightly, presenting different paths, different speeds. Each of these four videos then splits off into four more, and so on, growing exponentially. All of the paths produced in this branching structured are stacked on top of one another with ever diminishing opacity.