It’s a chilly evening and I’m sitting by my computer, headphones on, listening to a piece of music that feels like a gentle nudge into a vast, emotional landscape. This isn’t just any music—it’s my music. "A lake that’s on fire" was created by feeding poems that I wrote over the last few years into an AI which generated vocals and music based on my instructions. As the synthetic harmonies fill the room, I’m struck by a profound question: what does it mean when the music that moves us is created not by human minds, but by an artificial intelligence?
For centuries, music has been a deeply human affair. From the intricate compositions of Mozart to the raw emotional power of Kishore Kumar, music has always been seen as a direct expression of the human soul. Yet, today, AI systems can compose pieces that mimic the greats and generate new works in any style, from classical to heavy metal. These systems can also sing, their voices undistinguishable from humans’ to the untrained ear, a development that blurs the lines between human creativity and machine generated output.
The cognitive dissonance here is stark. On one hand, we have a long-held belief that music is a uniquely human endeavor, an art form that communicates emotions and experiences too deep for words. On the other, we are faced with the reality that machines can now create similar, if not indistinguishable, musical experiences. This juxtaposition forces us to reconsider what makes music, music. Is it the origin—human versus machine—or the outcome that stirs us?
Of course, we're not the only species to make music. Birds sing, whales compose vast symphonies in the ocean, and even insects have their rhythms. This suggests that perhaps music isn’t solely a human trait but a language of the natural world, one that AI is now learning to speak.
For someone like me, who is more comfortable with words (as opposed to chords), and who sorely lacks the musical skills and talent to trans mutate my words into song, this particular AI capability represents something profound. The words that form the core of each of these songs were written by me, some are informed by my lived experience, some connect to powerful and personal emotions, some just whimsically pop into existence from apparently nowhere. And yet, these words of mine languished on pages of old diaries and google docs that only I ever set eyes upon. But setting them to music is a kind of alchemy that enables it to take flight and finally leave the confines of my imagination.
Before, the barrier to entry in the music world was high—requiring years of practice and study. Now, AI can bridge the gap between imagination and creation, democratizing music production. This isn't just about making it easier for hobbyists and amateurs; it's a revolutionary change that could redefine professional music-making, influencing everything from how we create and consume music to how we value it.
The implications for humanity? They're as broad as they are deep. Philosophically and ethically, the rise of AI in music challenges our understanding of creativity and originality. What does it mean for an artist if the emotional resonance of a piece of music no longer requires a human touch? Economically, AI could disrupt the music industry, altering everything from copyright to the way musicians earn a living. Socially, it could change how we connect with music—perhaps making it more accessible but potentially less personal. Just like every other industry that gets disrupted by (any kind of) technology, there will be denial, anger and fear. There wil be (somewhat valid) concerns of a flood of mediocre AI generated music flooding the internet. In the longer run I suspect this will play out as it often does - bespoke human created content will always have a higher emotional value for other humans, the initial wave of early adopters/hype chasers will get bored and move on to the next shiny thing, a small minority of artists will figure out how to become better and more productive and more creative with the new technology, industries will change and evolve, the show will go on.
As for "A lake that’s on fire", the final product is something that I feel deeply connected to, even though my hands never touched an instrument during its creation.
For listeners, the question then becomes: does knowing that AI played a significant role in the music's creation change its impact? Does the knowledge that you are listening to an AI and not a human detract from the experience the music provides? Perhaps it does for some. For others, it might not matter.
In the end, I don't think is just about technology. It's about us—our definitions, our values, and our capacity to find beauty and meaning in the new and the uncharted.
As we listen to the future, perhaps what matters most is not who or what created the music, but how it resonates within us, stirring the vast landscapes of our minds and souls.
Faraway thunder. Raindrops on the roof.
Leaves rustling in the autumn wind.
The sea breathing in and out in tides.
Are these not music to your ears?
Check out "A Lake That’s on Fire" on Spotify, YouTube, Amazon Music and most other places you get your music. How does it make you feel?
(And yes, the name “The Tokenizers” is a bad pun)
Perhaps The fire in Belandur lake ignited your mind. What if it was in Carnatic music. Which ragas will be chosen by AI.? Anyway congrats for awe-inspiring thought provoking write-up.
Excellent….kudos to your creative imaginations and style.