Live link to the playlist below in the post.

I have mixed feelings about the ‘Spotify Algorithm’. It is on one hand the premiere way to sell music globally in 2019, and it replaced tens of thousands of geeks in record stores who were recommending music to you. I think it has certain biases, too, and it buries some of the music I work with. But it is also very damn good, it keeps me recommending music that I like.

Machine learning algorithms learn, and it you teach them regularly, they will really figure out what to recommend you. The ‘Release Radar’ is recommending you new songs based on what you have ‘liked’ and listened to recently. If you find music that you ‘like’ again, it will go into that direction next week. It’s learning capabilities are enhanced by another feature: you can ‘dislike’ a recommended song or artists. The more often you take the time to listen to your ‘Release Radar’, and the more feedback you give to it, the better new music you will find every week.

Let me show you my list called ‘Made in 2019’. They are very subjective to me, and I am sure it has a lot of personal biases, but it is a very diverse list. I think that the oldest artist is Geraint Watkins, he’s 68 and still rocking. The youngest are probably the Chastity Belt members, but Michelle Zauner aka. Japanese Breakfast and Benjamin Francis Leftwich are only 30. There is rockabilly, indie, psychedelia, krautrock, all sorts of things on the list, and most of them would be unknown to me without the algorithm.

If you like a song or two from this, put it on a new playlist, and check your Release Radar every week. You’ll be suprised how many new songs are created every single week all over the world.

Co-founder of Reprex, a reproducible research company and member of the Dutch AI Coalition. Data scientist behind the screen, analogue photographer in the sun.