Robot Dreams

The afternoon at the Statler finished off with a collaborative talk called “What Do Robots Dream of?” Cornell engineering professor Hod Lipson talked about his research on robots and teaching them to “evolve” and function independently. His simulations and footage of actual experiments with robots was fascinating.

In one project he described, a robot had to learn how to propel itself forward without any knowledge of its own structure or what it looked like. First the robot had to internally generate a hypothesis about its own structure and test this (through movement), repeating the process over and over until it figured out how it was laid out. Then knowing its structure, the robot had to hypothesize a way to propel itself forward, and then try out methods over and over until it actually succeeded. By experimenting and creating new models, the robot “evolved” its ideas and found a way to function correctly. This and other examples that Lipson discussed gave a glimpse into the future of robotics. In answer to his own question, “Are we there yet?”, he responded “no, not yet,” but it was clear how far we’d already come.

Lipson was followed by philosopher and retired IC professor Lee Bailey, who discussed the philosophical implications of creating robots that seek to emulate humans. Although a bit rambling, Bailey made a good case for why robots will never be able to be like humans in terms of emotions, creativity, and ethics.

Both speakers were supplemented by musicians and musician-robots from the Brooklyn-based League of Electronic Musical Urban Robots, who brought two instruments (one resembling an electric bass and one similar to a glockenspiel and a drum kit put together) which were hooked up to laptops and played music via robotic moving parts. The music was interesting, and rather eerie considering that people were not touching the instruments at all. After the end of the performance, audience members crowded up around the instruments to see exactly how they operated.

Categorized: General