I think this is a valuable framing. In other words, is there any computation problem which physics implicitly solves in the process of generating an image or vi

I think this is a valuable framing. In other words, is there any computation problem which physics implicitly solves in the process of generating an image or video—and which is: 1) computationally intractable to forge. (e.g. requires simulating trillions of photons) 2) computationally tractable to verify Sadly, my immediate gut intuition is that there is not such a problem, for a variety of reasons; but hopefully I'm wrong!