by June Kinoshita, Executive Director, New Harmony Line
Okay, I made up the title to riff on music AI, only to discover there really is a Bach Bot out there. It’s a system that uses AI to generate Bach-like music. I shouldn’t have been surprised. Music AI bots are proliferating alongside other generative AI programs. Major media and music companies are investing in AI-generated music, raising alarm among those who care about musical culture.
In a recent article, the Washington Post reported, “AI music has found its way into the mainstream — you’ve probably heard a few seconds of it even today: soundtracking ads on YouTube and Facebook, or providing the emotional context of a TikTok video.”
Some music AIs are designed to suck up all the music you listen to and generate more music that sounds similar but isn’t composed by an actual human. Think of all the royalty fees these companies won’t have to pay. Other companies are developing algorithms to generate tunes that will optimize your mental state.
But can such sounds, generated without any human intention, without a living person trying to communicate their experiences and emotions, be considered art? With art, it takes two to tango. You need an artist with an intention to provoke a response, and a recipient who brings their own sensibilities to bear on how they respond. Intention and framing matter. That’s why John Cage’s 4’33, the piece that famously asks you to listen intentionally to the world around you for four minutes and 33 seconds, is art while my silent procrastination is not. When I sit through a performance of 4’33, the composer is communicating with me, even if the man is no longer living. Without the artist mediating the experience, it’s just me projecting.
“Pieces of music aren’t just pieces of sound,” says Tod Machover, the composer and co-founder of New Harmony Line.. “They’re because some human being thought something was important to communicate and express.”
There is an appropriate place for AI in music, says Machover. In his own work, “I try to make models that are productive and useful and interesting and beautiful,” he says, “and I personally believe in a kind of collaboration between people and technology.”
This belief helps us appreciate Machover’s intention behind the creation of Hyperscore. Here is a tool that utilizes technology to remove barriers to music composition. That doesn’t mean it automates the process of composition. Hyperscore takes away the parts that untrained musicians find hard and leaves the kernel that is most important: What is it you need to say?
In a world where we are losing more and more of the human touch, where nearly everything we use in daily life, from our clothes, work tools, and food, is manufactured by machines, we crave ways to express and celebrate our human moments. Witness the explosion of interest in preparing food, knitting, DIY projects.
People are hungry for channels to express themselves creatively. Music is among the most direct, powerful ways to tell our stories. We must elevate and celebrate authentic, human music by supporting artists, music education, and the means for everyone to participate.
Just after posting this, I was struck by the irony that I had used an image (below) that I had generated using DALL-E (“Baroque-style etching of J.S. Bach as a robot”). It was fun, but I agonized over whether I was unknowingly stealing bits actual artists’ work or undermining the market in which they made their living. Is there a place for AI-generated illustrations? What do you think?