It would let users point the camera at objects or scenes to ask questions via ChatGPT or run a Google reverse image search.