One other note - I saw another comment that your AI model runs on CPU. Does it run well or better on the M1 or newer? That is, can you market it as an Apple Silicon-only app?