It's quite interesting that we're heading very quickly into accepting we cannot comprehend ourselves the work that AI can prove to be correct.