Report comment

Ian, good questions. A great deal has been written over the past 40 years on the possibility of artificial consciousness (some of it by myself) but I think in the debate about deep learning it's a red herring.
No one would claim that AlphaGo is conscious. It is an example of one of a growing number of systems capable in some - admittedly narrow - tasks of passing the Turing test of producing outputs indistinguishable from those of humans. They do this by collecting experiences - learning - of attempts to reach a goal.
For what it's worth, I believe judges may be more at risk from these developments than interpreters. But we shall see.