It’s called Lumiere and this one is going to be a game changer.
1. Describe the video you want in text and it’ll make it.
2. Turn any still image into a video
3. Using a single reference image, generate videos in the target style
Why do I get the feeling that this is going to be a abused?
Our reality is shifting… pretty soon you won’t know what reality even is…
https://lumiere-video.github.io/
h/t Shaun Kaven