May 2024

Learn About The Music You Like
I have been using other higher-quality services to explore and compare
Editorial By J. Martins

  On arrival in Shanghai, China, to attend a conference of the local audio industry association, I was promptly reminded of how powerful the digital curtain is in the region, being unable to use most of the online services that we rely upon for work.

But in this geopolitically divided world, this also reminded me about the dangers of relying too much on the online music streaming services provided by a limited number of companies.

In particular about Spotify, the largely dominant service currently blocked in China and which is today the TikTok of music services. It creates an addictive bubble for each individual based on their unfettered predilections, resulting in a vicious cycle from which it's very difficult to escape. For the musically enlightened, Spotify recommendations might seem relatively anodyne since we all know what we like, correct? We all have our list of favorite albums, and the artists we want to listen to most often, and obviously we want to listen to new music from those artists, and we want eventually to discover similar music, and similar artists to the albums and artists we like.

That's precisely how Apple Music works, and that's why every time I go there, I have a feeling of going nowhere. Yes, much of the music I know and care about is there for me to listen again, but inevitably all attempts to browse new music take me to places I don't want to go, or take me nowhere exciting.

In contrast, it is clear that Spotify's algorithm will inevitably conduct us in certain directions based on what we said we like (that's the early stages of the algorithm), but progressively based on what we effectively listen to — including those moments when we just follow along endless algorithmically generated playlists. Without noticing, gradually we are diverted from listening to music from our earlier preferences list, to an endless list of artists we don't know anything about, and songs that we have no idea where they originated and where they were made.

I don't like that, and obviously I tend to go back and explore the things that I was "subjected" to listen to. From Spotify, I discovered that I was often listening to faceless people making music to feed the algorithm, because that's how they manage to get "plays." Not surprisingly, Spotify was among the first services to be pointed out for using artificial intelligence-powered music tools, and allowing AI-generated music to be fed to its servers.

In principle, much like I don't care if a musician uses shoelaces on a bucket to generate bass notes, or if they use sampling in their music, I would have nothing against those same artists exploring AI-based music tools. But I want to know who they are as an artist, what music they create and where they play live — in essence, that they are real people with whom I am able to generate an emotional connection. That was always for me a foundation for music enjoyment.

What we are increasingly discovering in the music being fed to services such as Spotify, which is all about algorithmic addiction, is that AI-based music can fit this model perfectly and simply generate artificial "pleasance" that is meant to extend and perpetuate the vicious cycle.

I have long left Spotify for my daily music listening, but I return there to check on how it is evolving on a regular basis, and also to check the music metadata between services — which is another gigantic problem these days, where music on my favorite playlists disappears at an astonishing pace (that is a whole article for another time).

I have been using other higher-quality services to explore and compare, although the metadata issues remain a problem in all of them. After years of use, I can say that Tidal certainly steered me toward listening predominantly to music styles and artists that I never thought I would be listening to, but simultaneously doing a decent job of keeping my favorite albums and articles very close — Tidal sends me reminders about albums I have saved to my library and which I have not listened to for a while.

On my way to China, I was reading in the Financial Times about how a large percentage of (predominantly young) TikTok users have been effectively "influenced" to support populist parties with implausible messages as a result of finely tuned real-time recommendation algorithms. I'm afraid the algorithmic music experience is having a similar unhealthy effect.

J. Martins
Editor-in-Chief

Subscribe To audioXpress Magazine
Click here to subscribe to audioXpress magazine.