Generated with AI: Model telling using edge-detection (not sure it did, but still), followed by another generation: With no intermediary prompt. How computer vision algorithms “see” the world: