What does making music actually mean? Why does music matter to us? What goes on in our minds when we write, perform, and listen to music, and how is that different from what happens when a generative AI program creates a song? What does that gap mean for our relationship to AI-generated material? And just how did we get to this point with generative AI? In an August interview with Chamber Music America on the pitfalls and potential of AI as a tool for creating music, innovative composer & New Harmony Line Board Chair Tod Machover gives his perspectives on these nuanced and tricky questions.
Machover delves into the history of AI, including as it relates to his own groundbreaking work into the nexus of technology and classical music since the 1970s. He discusses early hopes for AI as a means of understanding and modeling how human minds work, and the divergence into what AI has predominantly become – generating replicas of the end result of the human creative process rather than engaging transparently or meaningfully with the process itself.
Is this isolation from the process such a problem? According to Machover, it can carry with it the risk of losing what makes original music meaningful in the first place: the expression of a person’s lived experiences, feelings, and hopes. If machines are uncaring, then they cannot imbue creative work with meaning themselves. Meaning-making, and thus music-making, must take place in close collaboration with people who do have intention, and who care. In its current predominant form, AI digests and replicates work in ways that are virtually unknowable for people interfacing with it on the user end; the process by which the work is generated needs to be shaped by human users who know what they want. What we have now is very potent, but it is not a substitute for the music that human users who bring their own meaning and care to the process create.
There is certainly great potential in the realm of artificial intelligence as it relates to making music. Machover shares ideas for ways this may look, using as an example the process of composing a piece and collaborating with an AI to generate iterations of mood and instrumentation for that piece. AI is very powerful indeed, and the prospects of human cooperation with AI are vast. These possibilities are diminished, however, if we do not pivot to designing and implementing these technologies with intentions and goals that center the creative process itself.
Read the full interview on the Chamber Music America website, and share your thoughts with us!
Cover image: Possessed Photography via Unsplash