NewScientistTech posted an article with a log capture of their recent conversation with a ChatBot that was an applicant to the annual Turing Test competition. Apparently ElBot managed to fool 25% of the Turing judges (or 83% of the necessary judges fooled in order to pass the Turing test).. In simple terms that means that ElBot fooled 3 of 12 judges, and if it had fooled 1 more judge it would have passed the test.
This is the closest any applicant has come to passing the Turing test since the annual testing started in 1991.
Follow me on Twitter 
That’s pretty cool, but it’s not taking over skynet. Yet.
I wonder how the three judges who were fooled explain it to themselves? The article’s log didn’t look very convincing.
No accounting for the intelligence curve.