Am I mistaken in thinking that much of what you're describing would be considered computer vision, and that computer vision was already largely capable of these

Am I mistaken in thinking that much of what you're describing would be considered computer vision, and that computer vision was already largely capable of these things in 2019 and before? I vividly remember a live demonstration of an on-device AR-and-object-recognition program at the 2014 Facebook developer conference.