Could video transformation become more important than video generation?
I tried Google's latest video model, Gemini Omni, using my own footage as input and asked it to transform the scene from warm lighting to a cool blue mood, and then to bright studio lighting.