I’m sure I’m not the only one who gets tired of my own Spotify playlists. I’m fatigued even with those daily recommendations of the app. That’s why I was excited when a friend told me about Spotify DJ.
Spotify DJ is an AI-powered feature from Spotify that was launched early last year. You can simply search for it in the app. As the name implies, it acts like a DJ. You don’t need to have a song as a starting point since it will analyze what you’ve listened to and take it from there. It plays music, talks in between sets, and introduces new music to you. It breaks the monotony of song after song and lets your ears rest a bit every 5-10 songs. It’s all good for the average listener but for those with a bit more discerning taste – like I am – the DJ feature is a let down.
See, I’m a metalhead. Metallica, Slipknot, Pantera, Meshuggah, and Deftones are part of my weekly rotation. It only changed when I moved to Bulacan since I now share most of my trips with my wife and I started listening to podcasts. Another change in my listening habit was because of my band, Breē. It kinda forced me to listen to some bands to ‘get’ the music that my bandmates want to produce, as well as listening to our peers in the local scene. Still, when I’m alone which is still a significant amount of time, metal and rock is the way to be.
So imagine my surprise when the DJ played songs from Top Hits Philippines, or Today’s Top Hits. It then switched to some OPM from the early 2000s that are too light for my taste, like Orange and Lemons. It’s not until 3 hours later when it started playing the likes of Blink 182 and Good Charlotte for some pop-punk, and a measly 5-song set of metal classics. After that, it’s back to pop mode for the Spotify DJ.
It tried to introduce me to new songs that it thinks will suit my taste in between sets. To be fair, the artists were good, but it still missed my real listening habit. I don’t need these EDM, bass heavy, hook-loving pop songs. Give me some riffs and double bass kicks with shredding guitar solos sir! Instead, apart from its recommendations, all it gave me are what’s popular or trending from TikTok.
Being an AI service though, I can only train it to be the way I want to. I’ll skip the songs I don’t like, which will be a lot, and let the songs I like to finish playing. That way, it will learn what kind of songs I’m really into. Although the quick shortcut to that would be, tapping into my most streamed songs from the previous years. I wonder why they won’t do that as a starting point instead.