Choose a still image — photograph, illustration, or screenshot — that you think will animate well. Describe the image in detail: what's in the frame, where objects are positioned, what the lighting looks like. Then write a motion prompt: what motion do you want the AI to infer? Be specific — are you animating a person's expression, an object's rotation, environmental change (light, weather, particles), or combination? Run the image through an image-to-video tool (Runway Gen-3, Pika, or similar). Generate a 6–10 second motion clip. Review the output. In a brief reflection, answer: Did the motion match your intention? What did the AI understand correctly? What surprised you or felt wrong? Would you iterate — changing the prompt, the image, or both — and why?