There is an amazing article on Popsci.com about what is being called an invention machine. John Koza has essentially mated the programming concepts of artificial inteligence and genetic programming, creating something that sounds like its straight out of The Matrix. You feed the program some basic properties of a device, give it what your desired results are to be, and it utilizes a process of natural selection to create a new and improved device. Read this quote for an example.

Lohn got his hands on the antenna specs for the Space Technology 5 mission. He plugged in an antenna’s basic requirements and let the software run. What he got, several hundred generations later, appeared to be a mistake. “It looked like a bent paper clip,� he remembers. Lohn had no background in antenna design, but that was beside the point. “There’s no chapter in the textbooks on crooked wire antennas,� he says. He had his bent paper clip prototyped and put it in a test chamber. Sure enough, it provided the tricky combination of wide bandwidth and wide beam that NASA required. Like the duck-billed platypus, it looked preposterous but proved perfectly suited to its niche.

Is this good? Bad? I’m not really certain yet. Towards the end, the article talks about how the machine, at least thus far, can’t create its own problems to solve. We, as humans, still need to tell it what we’re looking for. The invention machine merely crunches through the numbers in a new revolutionary way to get somewhere we would probably have gotten in time. It just gets there much faster. Being able to run through 1000’s of generations of code using a Darwin-esque process of natural selection. Getting rid of bits of code that don’t hold up to the ends and mating together the ones that do.

This entry was posted on Friday, April 21st, 2006 at 6:08 am and is filed under Technology. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.