Comment by piker

It works well until it doesn't. I've been a Spotify user for at least 15 years (wild to say that). I've discovered a ton of bands I've become a long term fan of.

But also their algorithmic playlists have gotten worse and worse. They overfit user data. They all recommend the same singles over and over and over. I've found they also don't make sense, recommending music that doesn't even belong in the playlist.

I've switched to user created playlists more often than the algo ones, although a lot of user playlists are just saved copies of Spotify generated recommendations.

There are many different "algorithms". YouTube has an algorithm that gets a very high hit rate because it plays music you've already proven that you like. That's good and bad.

I listen to the "Deep Cuts" recommender on Plexamp a lot, it uncovers a lot of good music that I haven't heard before out of my large collection, I've got no idea how it works.

One funny thing about it is that I do not play it through the speakers at home because my family tends to find anything selected by a recommender really provocative, not necessarily in a bad way, but the last thing I want to do is answer questions about music I didn't select myself. I mean I can play through a Charlie XCX or Tyler the Creator or rock-opera phase Kinks album and nobody says anything about it but if they do I have a good story to tell about it, but it's too annoying to get hassled about some UB40 track I never heard before that's between a BTO and Olivia Newton-John track.

(Generally I find that recommender selected content is "provocative", like if anyone is looking over my shoulder, they are very unlikely to see what I am actually interested in and working on but instead they want to ask me questions about things on my screen that I'm either just a little warm or totally cold to)

spotify has brought me some gems. more hits then misses, generally.

definitely engagement bait and they lean hard on popular (both pop, and just well loved / common) music.

no idea if the "payolla" system is still in effect like with radio but it feels like it sometimes.

Not sure about rating them against each other but I've had plenty of success with YouTube Music. Once it learns your taste well enough it starts giving you some real deep cuts. I've found songs I like with it that had less than ten thousand views on several occasions. Notably it has never played AI-generated music for me which is a big plus too. Google can be awful but their algorithm undeniably works.

It really depends, but in general I quite like it. I still sit through my Spotify weekly every week, moving the songs that I like into playlists. I regularly find new artists I like, thankfully I don't see much ai slop yet - the bands I find have tour dates and proper web presence. I've been to see some. But I wonder if it's a function of my niches or what?

I don't tend to let songs autoplay outside of playlists unless I'm commuting, and it usually just picks artists and songs I already know anyways.

Occasionally if I want to find more like an artists I go onto the song radio on Spotify but just read the song titles and artist names, it's not worth setting aside a couple hours for the possibility of two gems.

If I had one complaint or thing I wish I could change about my listening habits it's that I wish I spent more time listening to albums in order. I think something is lost by skipping straight to the best tracks - the dessert as it were, you've got to take the time so it's all the sweeter when it comes.