Reactions to AI-generated art have ranged from polite interest to a disavowing howl. Many artists feel as though their livelihood is being threatened by software that, in a few moments, can generate content that takes them many hours to complete. What’s more, the technology is good. Very good.
While there is much media attention on the visual arts, AI-generated music has also made its mark. Drawing from a database of existing compositions and directed by a user prompt, an AI can generate a completely convincing piece of music in a few moments. This technology has been used in experimental applications to complete some of the unfinished works of the great masters; notably, Beethoven’s 10th Symphony which only ever existed as a few sketches in the composer’s notebook.
Currently, the consensus seems to be that musical works produced by technology tend to have minor imperfections. But there’s no denying the fact that the results are very impressive. Proponents of the technology argue that AI will be a boon to artists by reducing the amount of time they spend on the mechanics of the creative process. This may be true, but the artistic process isn’t just the mindless execution of a task — it’s highly interactive and essential to the evolution of the artist’s craft.
Iconic bandleader John Philip Sousa once pitched a famous hissy-fit in his 1906 essay, “The Menace Of Mechanical Music”, to which my own article denies any ironic connection. Sousa’s essay predicted an end to amateur musicianship, at the hands of the newly invented gramophone and the player piano. Hans Zimmer, one of the world’s most influential film composers employs a staff of composers who work out the finer details of his film scores. Zimmer prompts them with a structure, melodic material, harmonic progressions, and exercises veto power on the finished product.
If AI helps an artist’s work along, by generating a “mostly completed” work for an artist to refine, then the artist isn’t exactly engaged with the process from womb to tomb and the human executant is less executor and more editor.
Many will say that, in this context, my resistance is just a masquerade for inflexibility. The argument is that I need to adapt to technological advancement. By not doing so, I’m only hobbling my own creative potential.
The prospect of AI sharing my artistic practice is troublesome because I worry it may reduce opportunity for me. But the motivation to refine AI to be a creator is mostly financially motivated and, I’m sorry to say, there isn’t really any financial opportunity for AI in the field of contemporary classical music.
Finally, I really enjoy being immersed in my own artist practice and I don’t think I particularly care that AI is out there disgorging musical excerpts in the style of famous composers, writing background music for whatever, or coming up with a new “boing” sound to advertise a product. Part of me has a hard time understanding why we would even want to create a shortcut through the creative process when it’s such an enjoyable part of being a musician.
- Will AI-Generated Art Kill Human Creativity? - April 10, 2023