“Many people in AI believe that we’re close to [a computer passing the Turing Test] within the next five years.” So said Eric Schmidt, Google’s executive chairman, speaking at The Aspen Institute last month.
The Turing Test, of course, was proposed in 1950 by brilliant computer pioneer Alan Turing as a way to decide whether a machine could be said to think.
The Turing Test has many critics, but it seems to me that if a computer convinces a panel of humans that it has human-level consciousness and intelligence then we will have to accept that it is correct. After all, that is what we do with each other.
(As an aside, isn’t it remarkable and rather wonderful that within a single human lifetime our society has changed one from which criminalised people like Turing for the supposed “crime” of homosexuality to one where most of us regard it as entirely normal.)
I’m loath to argue with someone as well-connected and successful as Schmidt, and I can’t prove he is wrong. The whole point of this blog is to get people to take seriously the idea that human-level AI is coming sooner than most people think. But within five years? If it was anyone but the boss of Google I’d say he had been smoking something. If he’s right we’d better get our skates on: we are simply not prepared for the arrival of human-level AI.