Extraordinary claims require extraordinary evidence

As of yesterday evening sometime, my twitter feed has lit up with claims that a computer has passed the Turing test for the first time. These claims have their roots in this press release from Reading University.

The details of the test and how it was carried out are thin on the ground. We do know from Reading’s press release that one of the judges was an actor and that in 33% of cases, the computer could not be distinguished from a human.

I have a couple of key questions.

  1. What language did the humans interact with the bot in? This is important because the bot is defined as a 13 year old boy from the Ukraine. If the interaction was in English then for me, all bets are off.
  2. Where is the peer reviewed paper?

The Turing Test is, in many respects, iconic. If someone claims to pass it, a press release is going to be nowhere near adequate to support that claim. We need to know a lot more about the system concerned, how it works and how it operates.