Between 1986 and 1988, the jazz musician and experimental music pioneer George Lewis created the first version of Voyager. Through encountering the music of David Behrman, Jim Horton, and others in the 1970s, Lewis became fascinated with the possibilities afforded by using computers in creating music. After spending some time making work that involved compositional programmes in Paris, Lewis returned to the US and began work on Voyager. His aspiration was not simply to use computers as a tool or raw material, but to create software that could take an equal improvisational role to the other (human) musicians in the performance.
Anyone who listens to Voyager’s improvisational jazz will not have a problem with calling its performances ‘art’. But there remains something unsettling about the ascription. Even today we tend to think of art as a distinctively human creation, arguably one of our highest and defining achievements. But works created by software such as Voyager call that assumption into question, asking us to re-evaluate our understanding of the nature of art, the role of the artist and our relationship to these new technologies.
The phrase ‘computer art’ has taken on varied meanings. Since the rapid expansion of electronic music in the 1970s, computers have taken on a central role in the production, distribution, and reception of music. However, in the case of Voyager the computer isn’t just a tool used to produce or present the work. Nor is it simply playing a composition that has been dictated to it by a composer or mp3 file. Instead, Voyager appears to be an improvising performer like every other musician involved in making the piece, creating and writing the music as the work unfolds. In short, the computer programme has the constitutive role of being a creator of the music.
The view that art is a distinct product of human activity can perhaps be traced back to the Enlightenment period in eighteenth-century Germany. As the philosopher Theodor Adorno noted in his Aesthetic Theory, during this time art began to be recognized as a distinct sphere of value and activity, independent of its relationship to religious practices and state power. Artists were no longer mere craftsmen or the vessels of divine inspiration, instead they came to be seen as ‘geniuses’ whose exceptional mental capacities and talents enabled them to create art. Immanuel Kant, one of the many proponents of this view, claimed that the artist possessed “an inborn predisposition of mind” in which they drew on their imagination, understanding, taste, and “spirit” to create original and exemplary works.
Although the idea of the artist-as-genius has now largely fallen out of favour, many still think of art as a distinctly human creation. One of the main reasons for this is the belief that creating art involves intentional actions and mental states: an artist is someone that creates art by virtue of intending to do so when composing the work. Moreover, artworks often express meanings in a way that other products of human activity don’t. For Oliver Roeder, even computer art is the product of the intention of the artist. The artist, after all, is the one who created the computer system in the first place.
If this is the case, then it would be difficult to think of Voyager as anything like an ‘artist’ or creator of the music it performs. At best, it might be said that Voyager is a tool used by George Lewis to create music. Scientists and programmers have yet to create a computer that has anything like the cognitive states or capacities that are presupposed in intending to do something and certainly the early versions of Voyager, created in the mid-1980s, lack these.
But why should we think that all art must be the product of intention? It is far from clear why we should favour rejecting Voyager as a co-creator of the music, instead of adapting our theories of art and its creation to accommodate it. The idea that art changes over history and requires this to be acknowledged in our theories is hardly anything new. Hegel saw art as something that progressed and developed through human history, just as Walter Benjamin claimed that the birth of photography fundamentally changed the nature of art in modernity. The fact that we now have art produced by computers and call it ‘art’ should equally give us reason to think that it is our theories that are now in need of revision to accommodate it, not our language to exclude it.
This question becomes more pressing with the recent resurgence of computer programmes designed to create their own music, with projects such as Google’s Magenta and Ji-Sung Kim’s Deepjazz being just two examples. Given the status we have ascribed to artistic creation since the Enlightenment, it is hardly surprising that computer scientists are pursuing this goal. It is now certainly true that programmes such as these and Voyager can now create music that is indiscernible from that created by an exceptional human artist.
But the achievement of developing a computer programme that can do this does not mean, as some have concluded, that we are witnessing an age in which there is no longer any real difference between ‘human art’ and ‘computer art’. Instead, that distinction has become essential to understanding what art is and why it matters today. Voyager raises questions about the nature of art, and its relationship to humanity, that could not have been asked before its emergence. As George Lewis once said: ‘Rather than asking if computers can be creative and intelligent… Voyager asks us where our own creativity and intelligence might lie—not “How do we create intelligence?” but “How do we find it?”
Featured image: Computer Code by Markus Spiske. CC0 via Unsplash.