• Cethin@lemmy.zip
          link
          fedilink
          English
          arrow-up
          3
          ·
          7 hours ago

          The answer “there is no correct answer” is correct, but the answer isn’t 0%. It’s a paradox. There isn’t a solution, which isn’t the same thing as saying there’s a 0% chance of getting the right solution (when that is one of the choices).

          • Natanael@infosec.pub
            link
            fedilink
            English
            arrow-up
            3
            ·
            5 hours ago

            Essentially there’s a difference between “0% of answers will be correct (because none can be)” and “the listed 0%-option self-invalidates if selected”. The problem is the interaction between answers that shouldn’t be there in normal tests, which changes the values after selection.

            Therefore, the only winning move is not to play.

              • Natanael@infosec.pub
                link
                fedilink
                English
                arrow-up
                1
                ·
                edit-2
                52 minutes ago

                If c is correct for you then it’s also correct for those who answered at random. And if it is, then you’re all wrong because you’re answering about those who answered at random.

                The only viable approach to actually answer is to argue that the question is ambiguous and that while the hypothetical test takers face infinite recursion you the real test taker don’t have to resolve the infinite recursion and to then choose a place to halt evaluation. And if they disagree then your solution to the Halting problem invoke the all powerful oracle function “I’ll talk to the principal”.