This recent TED talk introduced autonomous, agile, aerial quadrotor robots. These small-sized robots can navigate complex obstacles by transforming a 12 dimensional map into 4 dimensions and calculating the best path 600 times a second. They apply several concepts of multi-agent systems and complex adaptive systems theory to interact and cooperate, thereby resulting in the ability to accomplish intricate tasks at a system level. This behavior is emergent from simple rules at the individual level.
This kind of autonomous intelligence and cooperative behavior lends itself to several applications (besides the ability to perform the James Bond Theme). These include military and civilian applications such as the ability to lift heavy loads, reconnaissance and search & rescue efforts. Once the learning rules that allow the robots to adapt are coupled with long term memory and swarming behavior, the potential applications and implications are phenomenal.