Experimental results at the fringes of understanding seem to make the "Matrix" view of the world really problematic: if humans have the free will to devise experiments that are beyond huge amounts of computing power to calculate, how does the "Matrix" produce the correct answer? (Ditto AI...)