Why it is better to obtain music recommendations from people rather than algorithms
Sandeep Jayendran is the founder and CEO of Groovli, a social jukebox powered by music shared inside and outside of the users’ social circles.
How was Grooli born and why?
My teenage years were heavily defined by the types of music I paid attention to. Before the internet, the majority of the music that I discovered was through family, friends or random encounters. If it wasn’t for my friends from high school, I never would have been introduced to my favorite band, Tool. If it wasn’t for my one of my best friends at college, I never would have discovered my all-time favorite group, Bon Iver.
Most people underestimate the amount of music they’ve not heard. The fact is, there’s always more music that you’ve never heard than music you have. Nowadays, mainstream artists are pushed into the front of listeners’ consciousness by algorithm powered streaming sites. Accordingly, people’s tastes in music are becoming” normalized “and niche artists and music styles are gaining less attention. In this article I will outline the changes that algorithm powered music platforms have made to society as a whole, and then offer reasons why music recommendation is far better from people than from computers.
The good old days: music as a connector?
Since the 1950s, youth subcultures – with their distinct ways of dress, makeup, hairstyles and most importantly, shared passion for a certain genre of music, exploded around the globe. From rockers and hippies in the sixties and seventies, to goths and emos in the 90s, young people were easily identifiable as a part of a chosen creed. Today, however, few of these groups are seen – unless you count “Beliebers. “.
In an article for the Guardian, Alexis Petridis argues that today the general lack of popular youth subcultures can be blamed on the advent of the internet, and that modern teenagers are more interested in constructing an identity online than they are in making a public statement about their allegiances to a certain group or genre of music. In the same way as the internet has changed the way that young demonstrate their identities, it has also changed the way that they discover music and bond over it. Back then when music was shared via vinyl, then tape, then CD, it involved the physical passing of an item from one person to another. This music was then discussed, enjoyed together and, according to Stefan Koelsch, music psychologist at the Freie University Berlin, created a psychological bond among people. Koelsch found through research that when people discover that someone likes a song that they like, they tend to think better of them. Nowadays, as young people’s tastes in music become more generalized, due to the constant bombardment of “top 40s” music they receive on the TV, radio, and online, so does their ability to network with other people over a shared interest in a genre or artist. If everyone loves Pitbull, it doesn’t really make for an interesting topic of conversation. If you meet someone who shares your love of Balkan Jazz, then it probably does. The changes caused by streaming sites. Recent studies suggest that an increasingly digitalized lifestyle has reduced the way that newer generations consume content, to the extent that a 2014 study by Microsoft reported that most test subjects who took part in the study had a shorter attention span than goldfish. Richie Siegal argues that while in the past people may have listened to an album multiple times over the space of a few months before they grew tired of it, now people simply click on the next suggestion without a thought. He calls this “music exhaustion”
With a never-ending stream of digital media at our fingertips, our attention spans are at an all-time low while our constant craving for fresh content is at an all-time high, said Richie Siegel, Founder and Editor of online magazine Seersucker.
Music algorithms make us listen to music at an unnatural pace. As my grandmother said, “you should chew your food before you swallow it.” But with streaming sites, users don’t have time to digest the new music, instead clicking on a new song in seconds if they aren’t blown away by the tune immediately. Leading streaming sites such as Spotify use algorithms to recommend “new “music to users. While the algorithms are complex, they are based upon a simple idea: if millions of users ‘listening histories are analysed using big data, connections between certain songs can be made. This system, known as collaborative filtering, is based on the idea that if numerous users who listen to track A also listen to track B, C and D, then the chances are that these are probably similar tunes. Changing tastes. Unfortunately, this means that as young people’s tastes in music become more generalized, and they are unlikely to discover new and exciting new types of music that are outside of their comfort zones.
If they choose the same band that everyone else is listening to, then the algorithm is going to pick them another “big hit” selection too, creating a vicious circle of bland, mainstream music consumption in which only trending, heavily marketed artists get any air time. With individuals, music selection doesn’t work like this. People tend to make recommendations based on your existing tastes, character attributes, places you have visited or shared moments together. Accepting suggestions from a real person greatly increases the chance of hearing something that you have never heard before.
Social aspect of new music via people?
Back in my college days, my CD collection was a statement of my character. Many records had a backstory, whether it was memories of shared experiences with friends, a gift from an ex-girlfriend, or a CD that reminded me of driving in a car with my parents as a child. The music was intertwined with memories of the people who had shared it with me. According to Tomas Chamorro-Premuzic from Psychology Today, there are clear links between musical preference and personality. Consequently, music which is passed on to you via recommendations or gifts from other people ultimately moulds itself around your character, attaching itself to your moods and memories, good or bad. If used correctly, music algorithms offer an interesting tool to discover new content, but we should be wary of creating a new generation of youngsters without opinions on music, who only listen to the trending artists funnelled in their direction. Be it at a party, while travelling, via a friend on Facebook or in the street, allowing ourselves to be influenced by the music tastes of people that we meet, however brief or insignificant these encounters may be, opens the doors to infinite opportunities to discover new music which may well become significant for you and your character.
After all, the Beatles didn’t say “I get by with a little help from my Spotify.”
Take the experience here with Groovli