Specimen one
Ideas I get just before going to bed:
I want to create an IQ test in which none of the logical answers are correct. Instead, the answers will be constructed in such a way that they are non sequiturs, but one specific kind of non sequitur that is formed by applying some kind of complicated, logically invalid operation — that nonetheless has its own logic, once you know it — to the problem. Furthermore, the process should be reverse engineerable from the questions and alternatives once you know that the logical answers aren’t the sought-after answers. This would likely require a very specific kind of questions and a very specific set of incorrect answers. The idea being that humans have associative minds, and we are very good at finding patterns, so under the right conditions we should be able to find the patterns governing a specific kind of illogical answers. The hyper-intelligent would have every answer wrong, but would be compelled to find out why. The actual test, then, is two-fold: it does not consist in taking the test one time, but isn’t actually finished until after you have gotten the results the first time around, and then taken it a second time with the result from the first time in mind.
You can raise the idea to arbitrary powers by constructing the test and answers in such a way that the illogic gradually becomes clear after taking the test n successive times. This all based on my ad hoc theory that even the surreal, illogical, non sequitur can be categorized by the human mind and turned into a kind of logic we can apply, so that, after a while, unpredictable moves and answers to which we would be led by invalid reasoning yet again become predictable, and the illogic is reified in its own kind of logic. Fallacies of reasoning, after all, are common and reasonably (heh) predictable, which makes it possible for us to work our way backwards from a wrong answer to the faulty reasoning that led to it. This test, then, would be like a very advanced form of the cognitive empathy a teacher exhibits when they explain to a student how the student arrived at their wrong answer and which wrong turn they took, just by looking at the student’s answer. Since the answers aren’t uniquely determined by the questions (any number of wrong chains of reasoning could be used to arrive at a wrong answer), the test also requires an advanced understanding of human psychology.
The real genius of this test, if I do say so myself, is that it loops back on itself in several ways: not only is the test constructed so that it consists of several repeat runs of itself, but the actual test lies in creating the test itself in any coherent way, which I can guarantee would take a genius in its own right. Talk about bootstrapping!
Makes sense? No? Good night.
