[imranq]

I'd imagine someone could use text-to-image models like Dall-E to visualize a scenario with a few images