Tag Archive | Jaron Lanier

Listening to Robots Sing: GarageBand on the iPad

I recently had the opportunity to fool around with the iPad2’s new GarageBand suite. Enticed by the intuitive touch interface I soon found myself lost within the device’s labyrinthine architecture. Every poke, prod and press brought me to a new screen with a bevy of exciting options. A touch to create a drum loop, a tickle to evoke some reverb, and a brush to strum a guitar. I was one with the machine; it was a truly cybernetic, kinesthetic moment. This may sound naïve, but I had never realized how many tools were available to electronic musicians, or how intuitive using these tools could be. As digital tools to create music become more accessible and more intuitive, what is the role of the human in understanding their use? Further, what strategies can we adopt when listening to these creations?

This question may seem a bit outdated to those who have been researching post-humanist phenomena since the digital boom in the mid-nineties. Often conflicting perspectives regarding the negotiation of the human and the digital have been considered in the last decade or so. Some like Donna Haraway, Pierre Lévy, and even Ray Kurzweil offer particularly optimistic readings of the post-human (although for radically different reasons). While scholars like Nancy Baym and Jaron Lanier have offered decisively more sober readings of the problematic. They argue that splits between the human and post-human, or analog and digital are false dichotomies. Truth be told, none of the theorists above adequately address my feelings on this topic. Producing music with a digital audio suite makes me defensive of my humanism and it is by its very nature a project of preservation.

The algorithmic tools packaged within digital audio suites encourage a sense of aesthetic preservation. Tools like GarageBand’s Smart Guitar, Smart Drums, Smart Bass, various arpeggiators and Appleloops encourage the user to program music on a high level where the nuance of serendipity and improvisation play second fiddle to the overall sonic contours of a piece. Although the user is provided the tools to intervene and program music in a very specific way, it is by default a distinctly different experience than that of playing a guitar or piano. The ghost of the algorithm haunts such performances; reminding the user that these acts of spontaneous creation are no longer the default but deliberate…. This sense of deliberate improvisation forces me into a reflexive space where I am acutely aware of the mediations occurring within my performance. Succinctly, I must defend a sense of self within my creation. If I yield to the algorithms that seek to help me compose, I destroy all sense of the human within my work. Simply turning on robots and watching them sing.

For this reason, I propose an aesthetic of preservation as a way to understand the ways in which we listen to works created by digital audio suites. As algorithmic aids become more advanced and commonplace in music, the human becomes a less essential aspect of the form. Understanding what has been deliberately included in spite the seductive algorithmic environment is ultimately a project that seeks to recover the human in the machine; perhaps even, a project doomed from the start, as we grow ever closer to the means of our artistic production.

AT

Magnasanti – Check out the results of my collaboration with Colin Germain on GarageBand!

Add to FacebookAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to Yahoo BuzzAdd to Newsvine

Like This!

%d bloggers like this: