Pop Culture

Do you really want computer programs to decide the soundtrack of your life?

Streaming services pick songs for listeners based on algorithms, which deprives one of the joy of discovering new artists, genres

With almost any and every song easily available online these days, the happy problem plaguing music fans is how to sieve through the sea of songs to find the ones that you want to listen to.

Streaming services such as Apple Music and Spotify contain algorithms that automatically decide your playlists for you based on your past choices or on the selections of other listeners whose picks overlap yours.

Last year, Apple Music launched a feature, My New Music Mix, which plays newly released songs that you might like.

Spotify has Daily Mix, a series of automated playlists that mixes your favourite songs with new ones that you might like.

But do you really want to let a computer tell you what songs to listen to? It's not the same as sitting in a self-driving vehicle - any music fan will tell you the songs we embrace go towards the formation of our identity and, if we are being dramatic, our soul.

Yet we seem to be heading for a future in which software trumps our intuition.

Automatic playlists are convenient, for sure, if one is after a muzak list for a gym session.

But for someone raised on the pre-Internet ways of listening to music, the soundtrack of our lives should come from discovering new songs, artists and genres through old-fashioned ways such as reading reviews, talking to other music fans and going to record stores.

I spent a lot of my formative years discussing, arguing about and sharing music with friends and fellow music fans - saving up pocket money to buy cassettes and CDs, borrowing and duplicating one another's collections to add to our own and making cassette-tape mixes of our favourite songs.

In bookstores, I usually homed in on the music magazines first, from popular rags such as Rolling Stone, Spin or NME to local ones such as BigO magazine, an early champion of Singapore-made music.

I also devoured The Straits Times columns by former music writers Chris Ho and Paul Zach to keep updated on what's good and new out there.

The opening of mega music stores Tower Records and HMV in the 1990s was like manna from heaven - they had massive selections and listening stations that were never available at mom- and-pop heartland music stores.

Knowing there was so much good music out there just waiting to be discovered was mind-blowing and I was passionate about expanding my music repertoire.

It was a hit-and-miss affair, of course. I kept reading about how Kurt Cobain loved veteran punk/noise band Flipper, so I, a huge Cobain fan, bought a copy of the band's 1982 debut album, Album - Generic Flipper. But I didn't like it.

There weren't any streaming services back then and the radio didn't play their music. But taking that risk was part of what made music discovery so exciting.

The thing about letting algorithms decide new music for us is that there is a risk of the listener being sucked into an echo chamber and listening to only similar- sounding songs and acts.

I like the progressive and psychedelic undertones in the new album by metal band Mastodon, but I am also digging the intricacies found in rapper Kendrick Lamar's new hip-hop masterpiece. Would a computer program have predicted that I would be into these releases by two acts from disparate genres?

Music also triggers memories of when we first got into them. When I listen to Nirvana's Nevermind, it always takes me back to my late secondary school years; and Radiohead's OK Computer harks back to my national service days.

Years from now, how might you think of Ed Sheeran's Shape Of You? Hopefully, you won't be associating the song with a random click of your computer mouse? That'd be incredibly soul- deadening.

Some might argue also that contemporary radio playlists are somewhat similar to computerised algorithms. That's probably correct in a sense, since commercial, advertiser-driven radio stations have playlists that are decided by programme directors who focus on what's popular.

But there are also radio personalities such as the late iconic BBC jock John Peel, who famously championed influential bands like The Clash and The Smiths based on their music, not commercial appeal.

The echo chamber effect of algorithm-based music suggestions could be taken to an extreme.

An article in online news website Quartz over the weekend detailed how composers and companies are using neuroscience and software to compose music in tune with your brainwaves.

A playlist synchronised to your brain activity could possibly help you deal with anxiety, increase your productivity or help you sleep better.

It's exciting to think about such developments and the possibilities are endless.

But musicians, composers and artists spend years passionately honing their craft, creating music that moves listeners physically and emotionally.

While it can be argued that there are many songs out there that are so similar it feels like they share the same template, songwriting has never been an exact science.

Technology can make a lot of things in our lives easier and better, but I sure hope music never loses that irreplaceable human touch.

A version of this article appeared in the print edition of The Straits Times on April 19, 2017, with the headline 'No fun letting a computer tell me what songs to listen to'. Print Edition | Subscribe