Computers are learning tools. Mine has taught me humility.
My last computer fried when a glass of water spilled over it. I’m tempted to do the same to my new one — and this time it will be on purpose.
The field of artificial intelligence has made great advances in the last few years. Researchers at MIT have constructed a humanoid robot named COG, to which they progressively add better parts and computing power. Their hope is that this will simulate childhood development and will eventually result in something approaching an intelligent machine.
Other researchers at universities around the world have developed software for computers that uses natural selection to do calculations. Rather than relying on lengthy equations with huge numbers of variables to solve problems, the machines make up random solutions and test them against programmed constraints. The results say the answer was good or bad, and they keep changing their answers until one becomes more efficient than the others.
Basically, these computers have been given imaginations. They have the ability to test out ideas in their “heads” before trying them in real life.
The results so far have been rather impressive. COG can play the drums by ear. Computers using natural selection software are currently running warehouses, painting pictures and playing in jazz ensembles.
Obviously a lot of work has gone into making these intelligent computers. Lots of technicians dream of a day when computers will be as smart as humans. Perhaps they’re missing the point.
An age-old rule in warfare is that you don’t have to outgun your enemy if you can bring him down to your level. This, my friends, is what my computer has done to me.
I boot up the machine and usually type for around 15 or 20 minutes before the monitor suddenly turns blue and flashes obscenities at me.
Ugh. I type three pages of a paper.
I start chatting with a friend.
You get the picture. I finally decided to run a system diagnostic with McAfee Online, which is a sort of computer first aid site that I subscribe to. It began checking my computer for errors. I waited patiently. Finally the screen changed.
“McAfee has found the following problems with your comp–”
Guess I’ll never know. Tech support doesn’t know either. Nor does my roommate who was, until recently, a computer science major.
Even working computers are perfectly capable of screwing you. Witness the “auto-save” function. This is the gadget on your word processor that saves your essay every 10 minutes so you won’t lose it in a power outage or a freak accident with a nail gun.
When I sit and stare at my computer screen for 20 minutes trying to break free of writers’ block, I can be perfectly sure that the two crappy sentences I’ve written will be safely stored away on my hard drive.
On the other hand, let’s say inspiration hits and I spend 10 minutes writing a glorious and impassioned manifesto. Suddenly, wham. The lights go out.
I’d have done better to have stared at my two lousy sentences or gone to the bathroom for that matter. It would have produced the same result.
There won’t be a great 21st century novel. Not as long as they’re producing iMacs , anyway.
So you can stop wondering when the artificial intelligence revolution will come along. It’s already happened. Computers still can’t touch our intellect.
But they can sure drag us down to theirs.
There’s just one more thing I wanted to sa–
Josh Braun is the Daily Nexus science editor. He actually likes computers, just not his own, and invites you to share in his frustration.