Get to the stupid questions after all the sensible questions have been exhausted if, for some reason, the expected utility of the next least stupid question is still positive.
I think we need to find out what we mean by stupid and sensible questions.
Of course one should in any given situation perform the experiments (ask questions) that gives highes expected information (largest number of bits) yield, I.E. ask if it is a vertebrae before you ask if it is a dog. What I think we disagree upon is the nature of a stupid question.
And now, it seems I cannot come up with a good definition of a stupid question as anything I previously would refer to as a “stupid question” can be equally reduced to humility.
Almost certainly. I am also fairly confident that there is someone who has been hurt because he did look before crossing the road.
But does the negative utility from the situations “find out, get hurt from it” outweight “don’t find out, get hurt from it?”
Isn’t the heuristic More Knowledge ⇒ Better Decisions quite powerful?
Get to the stupid questions after all the sensible questions have been exhausted if, for some reason, the expected utility of the next least stupid question is still positive.
I think we need to find out what we mean by stupid and sensible questions.
Of course one should in any given situation perform the experiments (ask questions) that gives highes expected information (largest number of bits) yield, I.E. ask if it is a vertebrae before you ask if it is a dog. What I think we disagree upon is the nature of a stupid question.
And now, it seems I cannot come up with a good definition of a stupid question as anything I previously would refer to as a “stupid question” can be equally reduced to humility.