When creating a robot scientists often look at nature and copy/perfect it but sometimes it's interesting to see how robots learn things by themselves. I remember an article about a program that derived pendulum equations without knowledge of basic physics (but I can't find it anymore -.-). But these robots were designed to find resources and avoid certain areas and then tell their little friends where to find it.
The robots got higher marks for finding and sitting on the good resource, and negative points for hanging around the poisoned resource. The 200 highest-scoring genomes were then randomly "mated" and mutated to produce a new generation of programming. Within nine generations, the robots became excellent at finding the positive resource, and communicating with each other to direct other robots to the good resource.
Since there were not enough resources for all robots they did no longer signaled the swarm that they found a good resource and instead kept it for themselves.
However, there was a catch. A limited amount of access to the good resource meant that not every robot could benefit when it was found, and overcrowding could drive away the robot that originally found it.
After 500 generations, 60 percent of the robots had evolved to keep their light off when they found the good resource, hogging it all for themselves. Even more telling, a third of the robots evolved to actually look for the liars by developing an aversion to the light; the exact opposite of their original programming!
Pretty cool beasts but it's a frightening vision ...
My own musical tribute to two great men of science. Carl Sagan and his cosmologist companion Stephen Hawking present: A Glorious Dawn - Cosmos remixed. Almost all samples and footage taken from Carl Sagan's Cosmos and Stephen Hawking's Universe series.
Did you dreamed of building your own atomic bomb while your parents hiding in a bunker in the 1940th? In fact, I did not. Because I wasn't born yet but if you want to keep up with your childhood dreams this would have been your chance! For only US $4,150.00. Really cheap in comparison with the original price of $49.50 ($379.92 in 2005 US dollars).
Last Updated on Tuesday, 18 August 2009 11:10
Read more...
