Imagine a photographer reviewing a concept image for an advertising campaign that the client has already approved. He notices a problem: The man in the image is wearing a baseball cap that casts his face in shadow, but his eyes are clearly lit, showing his intense stare into the camera. The photographer’s trained eye tells him immediately that that lighting is impossible. No real-world set up could produce it. The image had been created by a senior creative using gen AI, and the client had loved it. Now the production team has been tasked with recreating—with cameras and human actors—something that reality can’t match.
