Courtesy of Spotify

October 15, 2019

YANG | Code Local Music Into Streaming Algorithms

Print More


p class=”p1″>For students leaving for Fall Break, music is most likely the go-to temporal escape from the hassle of the trek out of Ithaca. Whether it is amplified through the speaker on your car or your earphones, chances are the music is streamed but not owned. And chances are you conveniently “borrowed” your favorite tracks, which are merely ephemeral on your playlist, from playlists curated by recommendation algorithms. While streaming platforms merely exist in the virtual space, their impact is intertwined with the human geography of the physical world. Your very own algorithms travel with you as you physically move across the country. Yet when you are finally home for the break, your playlists and your recommendations stay pretty much the same. When was the last time you discovered a local artist that started out from your neighborhood? Perhaps only when there is a free concert with free alcohol. Your playlists are most likely dominated by big-name artists, a disappointment to your angsty teen-self merely a few years ago.

In the algorithm age, the mapping of music follows such trajectory of globalization and homogenization. Spotify and Apple Music have fundamentally changed our experience of discovering new music. The dominance of playlist culture has turned listeners to the lean-back passivity of discovering music through algorithm-curated playlists. While streaming platforms position their personalized recommendations as a tech-savvy model to refine our listening experience, such a notion is merely illusional. These personalized algorithms are never coded to surprise or to diversify in the first place, but rather, it perpetuates the stagnant cycle of discovering the same thing you have been enjoying. The echo chamber accentuates a specific algorithm-friendly type of sound and expels the alternative further into oblivion. By no means am I attempting to say that our digital culture is blatantly horrendous. The open platforms did get rid of the reliance on inner-circle human connections and has fostered a friendly environment for emerging artists. Yet, it is evident that the human nuances of music discovery got lost in the quantification and coding of databases. These human touches in music are fundamentally geographical. We used to discover music through magazines, radio, blogs and, a lot of the time, just through word-of-mouth. These are all mediums that are human-curated and put emphasis on geographical networks, yet now algorithm loom over everything and expel humans from the focal points of music discovery.

Back in the day, if you map out the networks among artists, musicians, engineers, managers, agents, critics and DJs, they were geographically concentrated. Iconic genres were mostly born in such concentration with the focal points of record stores and venus as physical traces of such connections. Bands like Interpol and The Strokes resemble the early 2000s alternative rock resurgence all emerged from the local music scene in New York City. Psychedelic rock bands in the 1960s emerged from the counterculture community in San Francisco and defined the touted and revered San Francisco Sound. In the pre-algorithm era, music discovery was largely influenced by such human geography and fostered an organic discovery process for listeners. Yet on Spotify and Apple Music, such information of human traces is made obscure and devalued in the process of machine learning.

So, how does the mapping look like when we attempt to conceive local music scenes as intertwined social networks in the algorithm age? Globalization. The world is turning into a more connected place for the circulation of music. At the same time, the inner circle of a vibrant music scene is getting smaller with fewer people as the connecting dots to support and diversify the music scene. It is gradually becoming a reality in which there is no “local” scene to begin with. Algorithm-based recommendations do their best to replicate the geographical concentration of urban music scenes. Big-name artists reach an even wider audience and tour around the globe. Meanwhile, emerging artists struggle to break through as they are virtually excluded from the playlists with their “abysmal” numbers. When numbers dictate your shelf space in the digital sphere, emerging artists are silenced and marginalized by the algorithms in the sense that they are geographically isolated in the virtual world.

While things may look bleak for musicians aiming for that breakthrough moment, our music landscape is far from hopeless. Magical things did happen during the phase between the post-Internet and the pre-algorithm era. Chillwave was born amid the advent of online communities. Bedroom artists like Toro Y Moi, Washed Out and Neon Indian found themselves lumped together under the term “chillwave,” a phrase coined by bloggers drawing parallels in their sound. Despite the fact that their bedrooms are thousands of miles apart, the formation of such genre was robustly fostered by the virtual network without the reliance on a physical cluster.

Perhaps a modern take on the radical mapping of music in the algorithm age would be the re-evaluation of missed opportunities for algorithms to foster virtual music scenes that surpass geographical limits. A radical revitalization of music scenes would, inevitably, be global and digital. To revive your local music scene, it is essential to put your local music on the map of the streaming platforms for exposure. If our algorithms can code such human nuances into the machine learning process, perhaps our music landscape in the algorithm age could connect people better than anything else.

Stephen Yang is a sophomore in the College of Agriculture and Life Sciences. He can be reached at [email protected]. Rewiring Technoculture runs alternate Mondays this semester.