Last summer I left Spotify and wrote about it with a rather surprising headline: “Why did I quit Spotify?” My reason remains. The software became clunky, the ads were unrelenting, and Sabrina's carpenter songs were inevitable. I wanted to find a better music streaming service. It doesn't make me happy to report that I had rejoined a few weeks ago.
The algorithm got me. I don't just mean that I got it I, how the Tiktok algorithm glues to the screen. Spotify's algorithms got me how my old friends got me, my strange love for yachtlocks and my continued obsession with French touch music along the way. It took me months to dig into the proverbs of Apple Music to realize that there are things Spotify doesn't get from other streaming services. My 15 years of listening to music and artificially intelligent software will strengthen those habits.
This is why algorithms tend to be considered villains these days. They are the technology behind your page for Tiktok. This brings you to some weird videos that you can't stop watching, as well as Amazon recommendations that seem to know what prescription you're taking. Meanwhile, Facebook's algorithms have been radicalizing Americans for at least a decade, and Instagram's algorithm feed is destroying mental health across generations. The meaning of Spotify's algorithm can be argued to be quaint in comparison.
Spotify's algorithm has given me a way of getting me and a strange love for the yacht lock.
But when I stopped Spotify and let it go, I made something happen. Similar to the algorithm feed, the centrality of how you consume information gives you more control over how these algorithms shape your preferences and behaviors.
If Algorithm works for you – just like Spotify is for me – don't mind submitting to that fun and useful product.
Music has always been important to me and over the years I began to feel like I had to wager Spotify to find songs I really loved. When Spotify launched in 2011, it was basically a large library of all music, but over the years it has introduced increasingly algorithmic recommendations and playlists that it promises to suit my taste. I still needed work to find something good.
This work makes Spotify's algorithm irreplaceable to me. It has been a decade and a half of my listening history, and over the years I have learned that habit and tinkered with it to meet my needs. I spent months trying to replicate this experience to Apple Music, but the algorithm struggled to surprise me.
All music streaming algorithms work with two basic principles: content-based filtering and collaboration filtering. Content-based filtering attempts to identify specific aspects of the song itself in order to arrange the next song, such as artist, genre, or mood. Collaborative filtering refers to recommendations made based on others listening to a particular song and what else they are listening to. If two people are listening to the same five songs, you have a chance like these six songs. It's all mathematics and sometimes there are anomalies that please you.
“Some of the serendipity you get is a kind of error turned into a virtue,” Glenn McDonald, former Spotify data alchemist and creator of all noise, told me. “So you're surprised, and sometimes those surprises are fun.”
Spotify's recommendations are not only tend to be comfortable, as there is a lot of data about me. Spotify has a history of 675 million people, and its interests could stack my interest in countless ways. Over the years I have developed a set of habits that will help me hone these recommendations – creating playlists, rejecting recommendations I don't like, exploring artist catalogs and, perhaps most importantly, digging into other people's playlists.
This is what I call lean forward listening. It's easy to click on Discover every week, but you can listen to the whole thing like Leanback and Radio Show and move on to the next playlist, but the more you put in to curate the experience, the more the algorithm works next time. At the very least, you can find your way into playlists that the algorithm didn't create.
How to resist algorithm rules
Whether they're like them or not, the algorithm recommendations go anywhere. Because companies like Spotify keep people hooked on their products when they work. Because companies like Amazon can manipulate people's actions through algorithmic recommendations. With proper product recommendations, someone can buy something they are not planning to buy. (We all did that.)
This current situation seems dystopian in many ways. The algorithm recommendation was the rage of decades ago, which I found useful, rather than creepy. Netflix pioneered the concept of providing customized film recommendations in the late 1990s, which is why it deserves a lot of credibility. However, by the early 2010s it was difficult to convey the differences between personalized recommendations and targeted ads. Nowadays, everything you see online is actually somewhat personalized, from the New York Times front page to a list of restaurants with your favorite food delivery app.
When you're talking about Spotify's music or Doordash's burrito restaurants, you can probably learn to live together. “We've seen a lot of people who have had a lot of trouble with us,” said Meredith Broussard, professor of data journalism at New York University. “As we all know, misinformation and misinformation are very popular, but not good.”
Role algorithms designed to increase engagement are the topic of book length that is played in the spread of misinformation. For now, I'll repeat that you don't need to go back and flood Facebook, Google, or X with algorithmically generated information. You can learn more about how these platforms use algorithms and how they can lead them to your interests.
If you're tired of X's algorithms, and if you're supplying right-wing propaganda, try BlueSky. This allows you to choose from a variety of algorithms for your feed. And if Netflix or other streaming services get old, try nuke the history of the view and try again. Spotify provides a detailed list of how to recommend and tweak content. Amazon also has tools designed to improve recommendations. (I tried all of these, including Amazon tools, which is very boring, but still might help.)
On big platforms like Google, Facebook, Instagram and Tiktok, it's a bit difficult for the algorithm to tend to be towards the black box end of the spectrum. Still, knowing how algorithms work and taking an active role to make them better can improve the experience on almost any platform. The algorithm is only responsible for if you allow them.
In some cases, you may like it if the algorithm is in charge. This is how I feel in Spotify in general, but I'm always fixing it and leading it. This is also the way I generally feel on Amazon, and I'm trying to buy only the basics. When I left Instagram a while ago, I decided that the algorithm was a little too much. Maybe I'll give it another try if I get bored one day.
This version of the story was also featured in a user-friendly newsletter. Sign up here Don't miss the next one!
