What if you could set up a 3D scene, product, environment, character, pose and let AI generate the photorealistic lifestyle photo from that composition? The 3D scene acts as the reference layer. The AI handles the photorealism on top. The 3D model is the input, the product shape is always exact. The AI only needs to make it look real. That's the moat.