My favorite interview question is "how do elevators work"?
0. If they immediately blurt out something like "I know how elevators work", probably don't hire them.
1. You can find out what areas they're most interested in. Do they jump straight to the physical mechanics? The programming? The UI? The abstraction that they're solving a problem?
2. Eventually, they reach a point where they just have to say "I don't know". The rabbit hole really just keeps going with elevators. You could know this question is coming and we will still easily reach the knowledge boundary.
3. If they "don't know" you can ask them to guess how a system might work. Or give them time to research it and follow up with you later.
I did a design interview which involved drawing up an elevator control panel for an apartment building with 1000 floors. Me, trying to leverage my sense of humour I guess (oops) designed a system in which the presumed wealthiest users at the top were instantly recognized with a face scan and never had to see a panel of buttons. The rest were forced to use a basic number pad. Needless to say I didn’t get the job. Doesn’t help that when you google my name my father’s extensive work as a labour organizer tends to spook my potential boss…
Your idea reminds me of a recent novella by Cory Doctorow, Unauthorized Bread. In there, there was an apartment tower with an elevator that had separate doors for poor and rich people (with the two groups living on different floors), and the poor people couldn't call the elevator when it was in use by the rich.
0. If they immediately blurt out something like "I know how elevators work", probably don't hire them.
1. You can find out what areas they're most interested in. Do they jump straight to the physical mechanics? The programming? The UI? The abstraction that they're solving a problem?
2. Eventually, they reach a point where they just have to say "I don't know". The rabbit hole really just keeps going with elevators. You could know this question is coming and we will still easily reach the knowledge boundary.
3. If they "don't know" you can ask them to guess how a system might work. Or give them time to research it and follow up with you later.