It detects text and images on the screen, then suggests possible actions, such as copying, saving, or searching. All initial image and text analysis takes place locally on the device, which means ...