A new AI model that renders generative visuals of what you see