> Yeah it's a trick question, the human error rate for it was about 30% (higher depending on the country).

Error rate doesn't prove anything. The nature of the errors is what matters.