It’s like watching a video and putting labels on things you see—like cars, people, dogs, or bikes—so a computer can learn what they are!