Super Artificial Intelligence

Last week, I read a two part piece, by Tim Urban, describing super artificial intelligence, it’s origins and inevitable ramifications on human life.

The AI Revolution: The Road to Super Intelligence

The AI Revolution: Our Immortality or Extinction

Tim does a very good job of putting everything in context, whereas, we might not be able to envision a future incredibly enhanced by super AI (most experts agree we’ll see super AI in the lifetime of the average millennial), but that’s perfectly human.  I certainly had a hard time wrapping my head around some of these notions.  He describes artificial intelligence that will appear to have been a sudden discovery, but we’ve been chipping away at it slowly and methodically (but faster and faster as technology progresses).


Kind of on a related note, I look at where technology was when I used a computer (that IBM PS/2 Model 30) for the first time, then fast forward to today, an incredible amount of progress that is very easy to be taken for granted.  Technology has progressed rapidly over the past 20-30 years, and it’s advancements, when summed up, are pretty incredible.  I carry a device in my pocket (and on my wrist) that has more computing power than a lot of the computers I’ve owned, combined.  

The advancement of artificial intelligence will be like this.  Seemingly tiny iterations that we all just get used to like Gmail spam filtering, Google Now, Google Photos face detection, Siri, credit card fraud detection systems, Google Maps navigation, music suggestion / playlist creation, etc.  One day, we’ll have super artificial intelligence.  Now I don’t know what it’ll look like.  Will my Google Maps navigation all of a sudden just become awesome?  Or will Google market their branded super artificial intelligence solution.  Who knows?

Consequently, today Google’s DeepMind division made a particularly big advancement in artificial intelligence with AlphaGo, a computer that can beat the world’s best Go players.