Track 5: Dancing On My Own — Robyn
If I had to pick one country as my musical home, it would be Sweden. ABBA was my first favorite artist, in no small part due to the fact they came first in my parents’ iTunes alphabetically-ordered collection. The first band I remember listening to on my bulky iPod is Ace of Base, a classic of 90s Europop.
Robyn released “Dancing On My Own” in the spring of 2010, which became an instant classic in the queer circles in which I moved during my university years. She graced billboards when I visited Stockholm a few months later, parroting the subway announcements to practice my Swedish (“Nästa: Kungsträdgården!”).
Little did I know that Stockholm was also home to a company that would come to dominate the music industry in the decade to come. Spotify had been founded four years earlier and had yet to cross the Atlantic. As music journalist Liz Pelly chronicles in her book Mood Machine, it viewed its main competitor as file-sharing website The Pirate Bay, another Swedish export and the bane of the music industry at the time.
Spotify’s founders were more interested in advertising than music. It’s no surprise that its recommendation algorithm relied on tracking its users’ behavior in minute detail, just like the advertising-funded web giants that developed what Shoshanna Zuboff calls “surveillance capitalism.” While early algorithms like Ringo asked users for explicit feedback, giving them some measure of autonomy in the process, now there is no escaping Spotify’s watchful eye.
The Atlantic. Spotify Doesn’t Know Who You Are By Nancy Walecki
“we can’t really tell an algorithm who we are; we have to show it.”
“You were much more in control of how you represented yourself under those [earlier] systems.”
The biggest flag was when users stopped listening. Anthropologist Nick Seaver details in his book Computing Taste how streaming platforms wanted to maximize engagement, or the “hang-around factor” in industry-speak. They wanted to keep listeners hooked, measured by “captivation metrics.”
This is where I hit a wall with understanding algorithms by focusing on the code itself. Even if I could infiltrate Spotify and look at lines and lines of jealously-guarded code, I would be totally lost. There isn’t just one algorithm, but many different ones slicing and dicing behaviour into bits of data that are then fed into equations, and algorithms to coordinate other algorithms. Instead, maybe the best way to understand algorithms is to look at how they changed our relationship to music.
