ChatGPT is Knock Knock Knockin’ on the doorstep of Spotify
Songwriter James Blake’s most recent album, Wind Down, plays in my ears as I meet Oleg Stavitsky, the co-founder of Berlin-based audio tech company Endel. As the sun turns to rain, the melancholy ambient piano tracks echo my mood. Maybe it’s no coincidence, Stavitsky says, pointing to the album’s credits where Endel is listed alongside Blake as co-creator of the music.
While Wind Down bears the name and face of Blake, and was mixed from its ingredients – it provided individual “stem” tracks with drum beats and melodies – Endel’s technology generated the product final. Its sound engine, trained on thousands of internal stems, creates personalized “soundscapes” for listeners by adjusting to externalities such as listeners’ heart rate, temperature, or time of day. Stavitsky cites Brian Eno’s “generative music” as inspiration, with humans constructing a framework that machines can then arrange and rearrange.
If Music AI’s Turing test is in good taste, Blake-Endel’s album doesn’t pass mine. I prefer soundscapes that are a little less cold. But I’m not Endel’s target audience. “Functional” music – whalesong, white noise, anything designed to play in the background – garners 10 billion streams a month, according to Stavitsky, double last year’s total and contributing between 7 and 10% of the entire streaming market. Real humans listen to machines: Endel says it draws more than 2 million monthly listeners across all streaming platforms, has a playlist partnership with Amazon.com and released an “AI Lullaby” featuring the artist Canadian Electronic Grimes.
All of this is serious enough to rattle record companies, who are rightly beginning to wonder if functional music is the end of a dangerous corner. For now, Endel’s technology creates music to strict specifications, such as adhering to the C major scale, and aims to provide soundtracks for tasks such as rocking babies and adults to sleep. . But how long before ChatGPT or something like that can create James Blake or Grimes or Beatles music from scratch? Benoit Carre, an AI-assisted music composer, says there’s no “big red button” yet for generating ready-made songs, but he’s already ticking off what artificial intelligence tools can do : Create snippets of songs in different genres, mimic the styles of individual lyricists, and adopt the vocal timbres of particular singers.
After sleepwalking into the last great disruption of MP3 file sharing two decades ago, labels are reacting with sound and fury to what would normally be called muzak. Universal Music Group NV, after recently lambasting “substandard functional content” (which presumably does not include Wind Down, released on a UMG-owned label) has reportedly called on streaming platforms to crack down on artificial intelligence services who retrieve artists’ back catalogues. to train their machines. Shareholders are nervous: When analysts at Exane BNP Paribas downgraded UMG earlier this month citing the potential for AI disruption, the stock lost 2 billion euros ($2.2 billion , about Rs 17,962 crore) market value in a single day.
While AI is a socially disruptive technology that needs safeguards, as my colleague Parmy Olson has written, there is also something more selfish and performative about this “war on white noise.” UMG is less concerned with the future of humanity than with protecting an already starkly unequal music streaming model. If functional music occupies a prominent place on platforms like Spotify Technology SA, it is because it serves as leverage in negotiations with music labels, whose collective market share is under pressure.
It’s also highly likely that of all the artists threatened by AI, iconic pop stars — the top 1% who account for 90% of streams — are the most enduring. UMG is working with streaming platform Deezer SA on a new “artist-centric” payment model to drive the music people really listen to first. And Endel’s Stavitsky knows that humans have star power: his ambition is to convince labels to let his technology tap into the catalogs of artists like Taylor Swift or The Weeknd to produce sound versions of existing albums. It could strengthen, not disrupt, the rock aristocracy.
The real problem is with those at the bottom of the food chain. “It’s going to get a lot harder to cut through the noise,” says Stavitsky. Even those who optimistically view AI as a tool for artists, rather than a threat, are worried. Denis Ladegaillerie, director of Paris-based music company Believe SA, says AI could help musicians in the same way the punk generation’s “three chords are all you need” sparked a democratic revolution. in songwriting. But he also says equality and diversity will need even more protection in a global music market where curation algorithms are already encouraging winner-take-all listening habits. “There is a real problem here for regulators,” he says.
The disruptive future of music is therefore likely to look a lot like its past: noisy and uneven. Record companies are not entirely wrong to ask streaming platforms to clean up in favor of more “human” music. But it’s also a good time to imagine fairer ways to distribute the spoils of streaming and bring out new human artists. If whales are on the verge of becoming a musically endangered species, what hope is there for the rest of us?
© 2023 Bloomberg LP
Leave a Reply