AI's Post-Playlist Future
Could AI help nurture more experimental taste in listeners?
6 March 2025 - originally published on VAN Magazine
I. OVERTURE
AI is the buzzword. Those little magic stars that pop up in every app, the sunlit uplands of a less laborious future, they are virtually everywhere—and are changing how we think about creativity. AI is a hyperobject , a symbol that refers to a wide range of practices, software and intentions across time and space and possibilities so vast that it is already beyond categorization or comprehension. AI the hyperobject is above all a tool, a technology with enormous potential that comes with both possibilities and pitfalls, and genuine, grave concerns around the foundations on which it's built, whether the rampant exploitation of copyrighted works, the abhorrent working conditions for those screen training data, or the untimely escalating environmental and climate impact.
However flawed, and however resisted, the technology is here to stay. As machine learning has improved in recent years, we will inevitably see huge leaps forward in systems with closed and concrete training data, such as healthcare or weather forecasting. But how will the prevalence of LLMs and AI music-generating machines such as Suno impact our relationship to musicking? What clues are around us that can help us understand the possibilities that lie ahead, both for how we create music, and how we experience it?
II. ANDANTE
When it comes to artists, I'm with Jennifer Walshe: I know artists have previously found, are finding, and will find even more innovative and curious ways of toying around with this gadget and making interesting, playful work. (A new era of plunderphonics awaits.) Humans have been making and experiencing art and music for much longer than the existence of capitalism and we will continue to be here for long after its dismantling—we're not going anywhere.
It's important to note that artists' livelihoods are not directly related to the quality or value of their creative practice, and it is easy to predict ways in which Big Streaming will use AI to profit at the expense of artists. Through the work of journalists like Liz Pelly, we've become familiar with the "ghost artist" and how, rather than paying real artists fees for streaming, Spotify has been accused of pivoting to commissioning Perfect Fit Content (PFC) from such artists for its biggest playlists, as a way to siphon more money into its own coffers. Big Streaming will inevitably harness AI to avoid even having to pay ghost artists and instead mass-populate its "jazz lo-fi" and "relaxing piano music for studying" playlists with entirely AI-generated content. AI technology will enable capitalism to cheapen music, but it's not the technology itself that's to blame. (So if you hear anyone wondering whether AI will disrupt the music industry, just tap the sign: no, it won't, but exploitative capitalism will.)
Corporate greed combined with the fast pace of AI developments and lagging policy means the pace of change will far surpass the awareness of how this impacts audiences. Our cultural habits as listeners are about to undergo a profound shift that corporate interests are unlikely to anticipate or understand. The ease of generating any type of music on demand will lead to an overabundance of what I call Teflon-music, Teflon being an artificial, low-friction, non-stick, non-reactive material. Some of this Teflon-music will be fun: love your favourite Netflix show but don't love the soundtrack? Give a few prompts and you can view it with a soundtrack of your choice. Karajan never recorded Scelsi's "Hymnos"? Hold my champagne.
So what happens when Teflon-music can be generated at the press of a button by anyone, anywhere, as long as we pay our monthly fee in exchange for music-generating credits?
III. INTERLUDE
The year is 2000 BCE, somewhere in Mesopotamia. You're a musician; there is a writing system for numbers, accounting and stories, and you have an idea: what if you were to write down this piece of music that you know? Would anyone care or be able to read it?
The year is 150 AD, near modern Aydın in present-day Türkiye. A friend dies and in his memory, you and his family members commission a stone pillar to inscribe a poem and a song complete with musical notation.
These are some of the oldest records we have of music notation, the first instances of music-the-process becoming fixed into a score, music-the-object, which can then be studied, shared, reinterpreted. Each time Nina Simone sang "I Loves You Porgy," she performed an interpretation of the same underlying music-object; the song transcends any individual version of it. With the advent of records, performances also became commodified as recordings (not without its critics of course), leading to a proliferation of music-objects. While commodification has many valid functions and purposes, it also converts music-process into something that can be accumulated, exploited, or protected from exploitation. What we are seeing now is the final boss in the commodification of music: music-product becomes so severed from music-process, that it allows the exploitation of music-product at the financial expense of music-process. No need to invest in time, space or energy learning music—just press this little button with the magic stars and wait a few seconds.
IV. THIS MOVEMENT TASTES FUNNY
What will happen when Teflon-music starts gunking up our airwaves? In Alan Watts' parable, the God who dreams lucid dreams of plenty soon gets bored, and starts introducing risk and danger to "spice it up [...] have a little adventure." When our earbuds are saturated with smooth music, we as listeners will seek the very imperfections and unpredictability that make music that much more interesting to experience.
In March 2024, I attended the opening night of Bartók's "Duke Bluebeard's Castle" at ENO which, due a last-minute illness, had a dramatic re-staging : Judith's role was split into two, with Jennifer Johnston singing from the side of the stage, and associate director Crispin Lord walking the role in a kind of performance art. The audience was ecstatic, having seen an asymmetric, spicy version of this piece that no one could have anticipated nor repeat again.
We may already be developing such an ear for the edges, as we see shards experimental music entering a range of mainstream spaces. On Monday, London experimental composer and singer-songwriter Daniel Blumberg won an Oscar for Best Soundtrack for "The Brutalist," a score featuring a disorienting and emotional blend of musique concrète , prepared instruments, live instruments and electronics. In his speech he thanked "the artists who played on this score [...] a group of hard-working, radical musicians who have been making uncompromising music for many years [...] and my friends at Cafe OTO." Mica Levi, a fellow London experimental music veteran who themselves has received multiple Oscar nominations in recent years, won a Soundtrack Award at the Cannes last year for their hyperpop-Penderecki score for "The Zone of Interest." In ten years, we have gone from "La La Land" and "Soul" sweeping the board, to awarding exciting, experimental music that grew out of grassroots scenes in London and elsewhere.
A couple of days prior to the Oscars, Charli XCX swept the Brit Awards, dedicating her success to "someone without whom none of us would be here tonight: the late avant-pop producer and pioneer of hyperpop, SOPHIE. Ben Nobuto's Proms commission, a glitch and splice version of the "Hallelujah" chorus, premiered last summer at the Royal Albert Hall, the LSO and Nonclassical put on sold-out nights at the Hackney Empire. Social media and digital platforms have accelerated the spread of microgenres that celebrate experimental DIY aesthetics as a way to grab short attention spans and go viral.
At the same time, our relationship to music is shifting, and there are shifts away from passive consumption of music-object and a revival of music-process. After spending hours trying to unsuccessfully convince ChatGPT to generate an image of a wine glass full to the brim, we will most likely walk out and pour ourselves an actual glass of wine.
We will see people seeking more communal and participatory music experiences: new-year rituals broadcast live on radio, cult-like alter-egos, listening parties in the dark, artists like David Sappa or Lia Mice making their own instruments, Wukir Suryadi creating ritualistic performances or a sketchy bloke in balaclava rocking up on his bike to Dai Hall in Huddersfield on a new music night with Neil Luck on the lineup, staying for the whole thing and commenting at the end that "it could have been spookier."
There's a wealth of Reddit threads where people ask for and share suggestions for experimental music, share experiences, learning and a sense of community; pre-pandemic, community choirs saw an all-time high with over 40,000 active groups across the UK. Perhaps some artists might stop releasing music on digital altogether and choose to create unique experiences only for those present. Let AI have music-object and make it cheap—we'll stay rich as long as we have music-process.
V. NOW WHAT?
However exciting all this might sound, it's by no means guaranteed. Napster, testament to the communal power of wanting to discover and share music with fans across borders, language, time, age, class, was put back in its box by the copyright fat cats—not because record companies wanted to protect artists, but because they wanted exclusive rights in exploiting them.
There is a future in which music becomes more of a living, breathing thing, something that enriches our social, political and civic lives, where music-making and music-experiencing becomes an integral part of how we express ourselves and connect with others; where we own the means of that process and AI is another tool that enhances this process. But it is not an inevitable one. There is also another future, one in which small grassroots music venues, crucial organ of the ecosystem that nurtured SOPHIE, Mica Levi, Ezra Collective, Daniel Blumberg and endless others, are a relic of the past; funding cuts have all but eliminated arts provision in schools and communities, university music courses no longer exist. AI has widened systemic inequalities and access to opportunities to make and experience music in our own individual and collective terms have evaporated. AI-tech companies successfully lobbied governments to invest in them as a way of investing in culture when they promise "a future where anyone can make great music," a branding sleight-of-hand that takes real money away from the common benefit and puts it in the pockets of companies like Suno which retain ownership and the means of music production (and profits thereof) while citizens become passive consumers.
The seeds of both of these futures exist today. By the time the future arrives for us to see which one has taken hold, it will be too late to change it. Will we listen to the Lorax?Whatever interest you may have in new music, however edgy your taste is or might expand to be: now is the time to notice, celebrate, support, advocate for, and be part of a living musical culture. Today it's AI, tomorrow it will be something else. All we can do is use these disruptor forces to reflect on which seeds we'd rather see grow and become better gardeners.
— Laonikos Psimikakis Chalkokondylis, 6 March 2025, VAN Magazine