All Hail Our Computer Overlords

Yesterday, as an April Fool's Joke, Wizleton decided to announce to the world (via Facebook) his new relationship status. He was, for a brief twenty-four hours, in a domestic relationship with Hal Skynet, the creepy powerful computer he built himself a couple of months ago. I'm not sure who first suggested the name back when he first built it but I know it's based on the unhappy similarity the things holds to HAL 9000 from 2001: A Space Odyssey (notice the red glowing circles). The whole thing was mildly amusing but it got me thinking about artificial intelligence. As a big sci-fi fan myself, I'm very familiar with the various technological villains popular culture has given us ever since the first days of Asimov (or would it be from the first days of Shelley? A topic for another post perhaps). Fun personal note: I named my dad's GPS GLaDOS. The cake is quite good. But how real is the possibility of evil artificial intelligence in this era of iPods and Skype?

About two years ago, when Terminator Salvation was getting all geared up to let us all down, The Times wrote "The Future of Artificial Intelligence". The article mentions and cites a bunch of different views from mostly Silicon Valley inhabitants on when artificial intelligence might arrive and what form it might take. The answers were all over the place, some suggesting specific dates like 2045, some suggesting that computers will mesh with humans in the next step of human evolution, some suggesting that we're going to kill ourselves before we even get a chance to invent something like that.

But that was 2009 and this is 2011 2012. IBM's Watson thoroughly trounced Jeopardy Champions Ken Jennings and Brad Rutter. Writers and humanists everywhere scurried to point out how Watson was incapable of thinking outside of his programming, how much more sophisticated the human brain is, and how Watson is not plotting to kill us all. Jenninngs even wrote an interesting piece here on the experience, covering his annoyance as loosing to something that could trigger a buzzer in a millionth of second with little quips about how his own game strategy apparently inspired Watson's training. IBM has stated that they are now looking into the logistic's of Watson's use in clinical diagnosis and legal research.

But if you want to look for an easier example of artificial intelligence, you might follow the experience I had a few weeks ago with my gmail account. I was sending a file of our lab report to my lab partner for her to look over and correct. I typed up the email and hit send. Then this appeared on my screen.

  

Gahhhhhh!!!

Holy Turnips! How can it do this?

So yeah, I know it's an algorithm, the same used every day to try and hawk vanity publishers at me every day because I get a lot of emails about creative writing contests but still. Gahhhhh!!! It's only a matter of time until they get to all of us. Run!