AI Via Super Mario Evolution

Can Super Mario teach you to think? That’s the idea behind using a simple version of the game to teach artificial intelligence. [Oddball] calls this The Mario Genome and wrote at program that can take on the level with just two controls, right and jump. He gave the script 1000 Marios to run through the level. It then eliminates the 500 least successful and procreates back to 1000 using the 500 most successful. In this way the program completed the level in 1935 generations and completed it in the quickest possible time in 7705 generations. He’s posted the script for download so that you can try it yourself. It’s an interesting exercise we’d love to see applied to more random games, like Ms. Pac-Man.

[via Reddit]

Swarm Robotics


Uber-geek [James McLurkin] was in Austin recently demoing his robot swarm. He’s on tour with EDA Tech Forum. [McLurkin] has multiple degrees from the MIT AI lab and worked at iRobot for a couple of years. Lately, he has been working on distributed robot computing: robot swarms.

[McLurkin] was an entertaining speaker and had an interesting view of robotics. He is optimistic that robot parts will become more modular, so it will be easier to build them, and more importantly, faster to design them.

Some quotes:

  • “There’s more sensors in a cockroach’s butt than any robot”
  • “12 engineer years to design, 45 minutes to build”
  • “If it can break your ankle, it’s a real [rc] car.”

Continue reading “Swarm Robotics”