How do you decide what music to listen to? Since streaming is now America’s dominant form of music consumption, we’re guessing Spotify or Apple Music are helping make this choice.
Both these companies and their competitors rely on algorithms to manage an increasingly large variety of music playlists. So…a robot is picking your music? Well, yes and no. Even a tech titan like Apple recognizes that machine learning has its limitations: most streaming companies employ music experts who fine-tune playlists combining computer data with essential context and personal taste.
Although the dominant mode of music listening has changed drastically since just the turn of the century—when terrestrial radio was simply called “radio” and personal music collections were still common—people continue to need guides for discovering music. Some industry leaders including Eric Schmidt have argued that AI can now perform this DJ/curator function as well as humans.
Even Diplo, one of the world’s highest paid DJs ($28.5 million in 2017), relies on a data-driven listening platform for music discovery: “I’ll just put on YouTube and let it autoplay videos—that’s how I hear new stuff, because there’s just too much music to listen to.”
If a globetrotting trendsetter like Diplo feels overwhelmed by the sheer abundance of music and outsources his listening to automated discovery tools, has machine learning already replaced humans in music curation? Not quite. Diplo may be leveraging algorithms in his music exposure, but he’s exercising his own expert judgment in choosing what music to actually digest and disseminate to his millions of fans.
What does all this mean for the average listener? Or for a brand looking to soundtrack their experience in a meaningful and curated manner? Some in the industry agree with Mr. Schmidt: that the answer to today’s curation challenge is big data. But machine learning lacks situational context, as well as the distinctly human variable of taste. These are critical components that Diplo and other music experts use to filter which songs really connect and which—for a variety of nuanced, subjective reasons—miss the mark emotionally.
Arguing for the importance of human curators, music writer Jim McDermott makes another great point: “New genres often appear as a reaction to social conditions (think punk and rap) and often are initially disparaged by the mainstream. This [new] music is born like a virus and spread by small groups of passionate people until it is popularized. Everything about these genres contradicts the concept of ‘what actual listeners are most likely to like next.'"
Missing these kinds of contextual clues can put algorithm-curated music at odds with what listeners actually want to hear. Case in point, Spotify recently launched an “equalizer” (an equality analyzer) built to assess the gender breakdown of one’s listening habits and suggest something more inclusive. While the idea is appealing from a distance, in practice things didn’t work out quite so well. The Needle Drop founder Anthony Fantano recorded his experience on YouTube, highlighting how the new playlist misses the mark. Instead of adding music by female artists that mirror a listener’s stylistic tastes, the equalizer was more likely to add an assortment of mainstream pop hits.
Whether it’s a superstar DJ like Diplo or one of Spotify’s behind-the-scenes music professionals, most of today’s listeners choose—whether knowingly or not—playlists tailor-made by humans applying algorithms. The same is true for brands that use music successfully in their customer outreach. It’s telling that Spotify’s most popular and visible playlists are driven by talented individuals, such as the recently departed RapCaviar curator. At least for the foreseeable future, human curators with specialized expertise are best at song selection…using a unique combination of data, context, and personal taste.
Photo Credit: EliasMusicLibrary