The Spotify Weirdness Edition
On composition, algos, and serendipity
Robin Sloan (RS) is an author, olive oil maker, and internet tinkerer. I could point to any number of amazing pieces he’s written around the internet, but there are two that stick out most for their effect on me personally: Stock and flow from his Snarkmarket days laid out a simple framework for thinking about the different types of content on the web and Writing with the machine encouraged me to spend some time playing with neural networks. Thanks Robin! - Noah (NRB)
Robin here. The other day, I was listening to Spotify’s recommendations, after the platform’s algorithmic radio station had picked up where my short playlist left off. An interesting track came on; its whompy brass reminded me faintly of Too Many Zooz.
Nice sound, I thought. I wonder what else this artist has produced …
The answer was: virtually nothing. “Danni Richardson” appeared to be one of those artists with just a handful of tracks available on Spotify, of which this one, titled Romilda Gebbia, was by far the standout hit, with 91,000 plays.
I don’t mind these one-hit wonders at all. For years, the artist M. T. Hadley offered just two perfect songs on Spotify. He seemed to ask: “Aren’t these enough?” They absolutely were.
Oh, well, I thought. Back to the algorithmic radio station.
Except: the next track was the same. I don’t mean that it was Romilda Gebbia again. This one was nominally different: Veneranda Caputa, by Brett Byrne. But its core was unmistakable, barely hidden beneath the veneer of different timings and different timbres.
Both tracks a mere 45 seconds long , I should add— more vibes than songs.
That’s very weird, I thought, still unaware of the spectral zone I was entering.
Because the next track was the same. And the next! AND THE NEXT!
I’ve collected these tracks in a playlist so you can listen for yourself, and hear the strange repetition. There are far more versions than the ten I’ve pinned here — definitely dozens, possibly hundreds — and Spotify’s algorithm will obediently deliver them to you in the Recommended section, down below the playlist.
In all of them, you can detect the underlying composition, its essential elements: call that Romilda Prime.
Why is this interesting?
This discovery is interesting because it offers a glimpse of something otherwise unseeable, misted across the vastness of a digital platform. A lot of things in 21st-century media are like this.
There are plenty of fascinating questions to ponder, about the productions and economics and intent of these variations, but/and I find myself returning to the way I found them: the spectacle of Spotify’s algorithm plunging into an absolute rut.
I’ve never seen the recommendations settle so quickly into such a deep groove; once the algorithm got the taste of Romilda Prime in its mouth, there was no getting it out. Is that a matter of chance? Maybe our mystery producer has spun this wheel many times, and it’s only with Romilda Prime that they hit the jackpot. They found a sound that is perfectly seductive, not to humans, necessarily, but to the one listener who counts most: the algorithm.
Like a lot of things in 21st-century media, I both love this and hate it.
I love it, because it’s so strange, so dizzying, and — credit where due–because our mystery producer is truly going with the grain of the medium, in a way that no one merely “making albums” does, at all. What could be more 21st century, more “liquid modernity”, than releasing your music as a haze of variations into the swirling currents of the algorithm?
I hate it, because of course it troubles every intuition I have about what it means to be an artist — the very idea of authorship! There’s a cynicism to the specific execution here, too. I can easily imagine another version of this project that uses interesting track titles and generative album art, so that groping around for the hidden “prime” might feel more like piecing together a puzzle, less like leafing through junk mail. (RS)
List of the Day:
In a way, this is simply the musical version of a phenomenon that’s common across the food delivery services. Here’s a picture snapped in my neighborhood:
My bandmate Jesse Solomon Clark and I are trying to go with the grain of the 21st century in a different way, connecting my AI tinkering to his compositional skill. (RS)
WITI x McKinsey:
An ongoing partnership where we highlight interesting McKinsey research, writing, and data.
How and where to work next. Selling, general, and administrative functions are integral to your organization’s success. What’s the best place for those people to work? Read here to see how the future of work will take shape.
Thanks for reading,
Noah (NRB) & Colin (CJN) & Robin (RS)
Why is this interesting? is a daily email from Noah Brier & Colin Nagy (and friends!) about interesting things. If you’ve enjoyed this edition, please consider forwarding it to a friend. If you’re reading it for the first time, consider subscribing (it’s free!).