The Turing Test Is Like the Ultimate Test of Artificial Intelligence

The Turing test tries to falsify the hypothesis that an AI system is not intelligent. When something passes the mirror test, it doesn't necessarily guarantee that the thing that passed it is sentient. The question is how frequently can humans fool the judge? So let's say you run thousands of trials, whole bunch of different judges,. And on average, the judge gets it right 70% of the time. This is a woman pretending to be a man and getting it right 70 per cent of the times.