Before you set out to change something, make sure it needs changing.
Elevators, for example, probably do not need dramatic improvements. Particularly in their user experience, they work fairly well. You are at a certain floor, you press a button to call the elevator, you enter it, you select the floor you want to go to, it moves and stop, you exit the elevator, done. It is a pretty well oiled dynamic, and trying to make it more efficient – say by asking people to select the floor they are going to at the same time they call the elevator – might create funny confusion and make the process slower.
Similarly, the process of listening to music on a mobile phone was fairly smooth without needing to remove the jack and forcing people to use bluetooth headphones. Of course, improving the user experience was not primary in this case, and so eventually the whole process now feels more clunky and unreliable.
Not everything needs changing, not everything needs improvement. And if eventually your decision is that yes, this thing does really need to be better, put the users first when designing the betterment.
