I really enjoyed reading Peter Norvig’s short piece in the New York Post (the Post? yes, the Post!) about the age of machines.  Artificial intelligence (more than games) was what first got me excited about computers and I still have a bunch of the early AI books (Patrick Winston anyone?) on my shelf.  We are now actually getting AI although people seem loath to call it that because of the history of over-promise.  Yet, computers are clearly doing exactly the kind of things that we had hoped decades ago they would be able to: translate texts, understand voice commands, recognize faces, answer questions, etc. 

This progress is arriving in relatively mundane packages (eg voice commands on Android) but that should not obscure the immense magnitude of the progress that’s been made.  Nor should the fact that computers don’t “understand” what they are doing distract from this actually being a feat of intelligence.  After all, we (as conscious individuals) don’t really “understand” how we recognize faces.  Yet we do it all the time and that is good enough.

I am looking forward to this week’s man vs machine Jeopardy challenge.  Whether or not the machine wins tonight, I am convinced that our days as the best Jeopardy players are numbered.  And I consider that a good thing.  While we are at it, let’s please dispense with spelling bees also.  We should focus our attention on efforts in which our trillion connection wetware has a meaningful competitive advantage and can find meaning.

I really enjoyed reading Peter Norvig’s short piece in the New York Post (the Post? yes, the Post!) about the age of machines.  Artificial intelligence (more than games) was what first got me excited about computers and I still have a bunch of the early AI books (Patrick Winston anyone?) on my shelf.  We are now actually getting AI although people seem loath to call it that because of the history of over-promise.  Yet, computers are clearly doing exactly the kind of things that we had hoped decades ago they would be able to: translate texts, understand voice commands, recognize faces, answer questions, etc. 

This progress is arriving in relatively mundane packages (eg voice commands on Android) but that should not obscure the immense magnitude of the progress that’s been made.  Nor should the fact that computers don’t “understand” what they are doing distract from this actually being a feat of intelligence.  After all, we (as conscious individuals) don’t really “understand” how we recognize faces.  Yet we do it all the time and that is good enough.

I am looking forward to this week’s man vs machine Jeopardy challenge.  Whether or not the machine wins tonight, I am convinced that our days as the best Jeopardy players are numbered.  And I consider that a good thing.  While we are at it, let’s please dispense with spelling bees also.  We should focus our attention on efforts in which our trillion connection wetware has a meaningful competitive advantage and can find meaning.