Humans aren't perfect. It's what gives us our unique little quirks. Same as AI at this stage of its growth. It won't make such silly little mistakes tomorrow.
Some humans aren't even up to the standard of inference-based reasoning; it's quite a spectrum. But there's little reason to believe that carbon based reasoning, is fundamentally superior to silicon based reasoning, in the long term.
