For those with theoretical interests with respect to machine learning flavored AI, the ML Theory blog run by John Langford is highly recommended. Though recently started, Langford and others have so far been doing an excellent job of commenting on both the science and culture of theoretical learning research.
Yasuo Kuniyoshi and Shinsuke Suzuki found that they could get a simulated robot to walk, and even to scamper over obstacles, by programming each leg to obey a chaotic feedback controller.
Most of you have probably already seen this (NS), but I figured I’d pass it on anyways. As usual, the media has gone over the top with the “robots are going to take over the world” theme. I think the basic idea, which does deserve some attention, is that hypothesis generation in the face of complex biology can be overwhelming. I don’t think this kind of inductive AI is that innovative nor is it necessarily the best solution in all cases of hypothesis generation, but the example is interesting.
The best part of this research, though, is the “experimental comparison” to graduate students. You can guess who won. I’m sure those students just love being in that lab now.
Article in Nature. Here’s a more entertaining article that includes a diss by Pat Langley!