Part of the problem is that the confidence for “cardboard box” was probably quite high. It’s hard to properly calibrate confidence (speaking from experience, speech recognition is often confidently wrong).
I think there are strong indicators in much neuroscience research that indicate our brains are quantum entangled in a way that a cpu currently is not. I project the quantum entanglement part from the experiments such as the precognitive ones, which usually don't make that claim.