The Army of Darkness: Embodied algorithms for artificial mindlessness

The Army of Darkness

In algorithmic composition one frequent problem that programmers are faced with is how to make computers less predictable. The clue to the matter is that there are infinitely more variables in the physical world than in virtual environments. Searching for an audiovisual implementation of this principle and tapping into the culture of hobby robotics, sound-art collective Gullibloon has created a system that uses line-following robots to play electric guitars. Their robots, baptized “The Army of Darkness”, are based on a design commonly employed for maze solving. Gullibloon has them running along lines laid out as race tracks on the floor of the installation space, with electric guitars placed face-down in several points, so that the robots can strum them as they drive by, producing what the artists define “algorithmic drone rock”. The steering mechanism they use can be traced back to Valentino Braitenberg’s “Vehicles. An experiment in synthetic psychology”. This seminal text describes how to create the impression of life and intelligence with simple sensor and actuator systems, postulating the principle of “uphill analysis and downhill synthesis” that is so popular in current generative art. While the algorithm employed to drive these simple reflex-agents is elementary (the source code downloadable from Gullibloon’s website is 60kb) embodying it in physical agents can lead to a wide and unpredictable range of results. Cage introduced indeterminacy to free music from the composers will and supplant the romantic idea of genius. “Never just strum. Put your mind to it when you play” was one of the rules given by Schumann to his students. But guitar drones require quite the opposite, and robots can be well suited for mindless activities.

Matteo Marangoni