Opening page…
Would you keep the off-switch?
Suppose AI systems reliably outperform humans at running public institutions and military decisions. Would you still insist on a human off-switch?
Trusting a system and retaining the power to overrule it are different decisions.
Your answer
You’ll sign in or make an account, then this is kept as your answer.
A little context
Your phone's map has been right about the traffic a hundred times in a row, so when it sends you down a strange side street, you go. You would feel very differently if it could lock the steering wheel. The question is what that wheel is for once the machine drives better than you do.
Nearby questions
- The oracle you can't refuseAn oracle exists that always gives the wisest possible answer to any decision — which city, which job, whom to marry. It has never once been wrong. How much of your life do you hand it?
- Do you have free will?Your genes, your upbringing and the laws of nature were all in place before you chose anything. The choice still felt like yours.
- Can you know what's right and still not do it?Socrates held that nobody does wrong knowingly, so anyone who does must not really know. Everyone who has broken a promise to themselves has doubts.
- If you've done nothing wrong, should you mind being watched?The argument sounds airtight until you picture every message, search and place you've been, laid out for the people who know you.
- Predicted before you choseIf a system can predict what you will choose before you know it yourself, is the choice still yours?
