Every so often (maybe once a year) I look through my music collection and then go “Oh, I forgot about this band that I used to love” and then look them up and see if anything new from them is out. If the band has broken up I also look up the band members to see what bands they are now in or people they have worked with etc. The Six degrees of separation thing works quite well.

I only discover music through human curators I trust.

Currently it’s via the YouTube channel “worldhaspostrock”, though the word “new” might not be appropriate because as of late they’ve just been posting decade-old albums.

I get recommendations from real people, not algorithms.

I go to live shows and music festivsls and buy music at the merch table.

It’s much easier to find organic art in a community, outside the art industries.

Sure, unless you set up a lot of subscriptions it won’t just come on a consumption conveyor belt, and some people might want or expect that, but it’s much better to be in a place where you’re actually interacting with artists.

Sites like Gnoosic and Music Map. Browsing Soulseek/Nicotine shares. Wikipedia to see who artists have collaborated with or any side projects they’re on.

Bandcamp

Record labels. For any band you like, go to their label’s website, browse the other artists on their roster or peruse the label’s past releases. The bands will likely be similar in style/sound, or at least worth checking out.

I literally look for live performances on YouTube and then look for similar recommendations or, if the artist did it, their list of recommended channels on their own channel.

I work listening to ambient, which is a terrible minefield for AI, but having discovered Martin Stûrtzer led me to discover a bunch of new no AI instrumental and electronic (mostly modular analog synth) music artists.

Listen to the billions of songs made before 2016 lol. You got a lot of catching up to do.

New does not equal good.

Beyond that, follow trustworthy artists. The only way to fight the slop is with community and trust.

Bandcamp, but also get to know the artist behind the music. Maybe follow their socials and see if they have other friends that publish music.

Oh wow, that’s something that’s been bothering me lately a lot. Not just avoiding Ai generated music but developing my own musical tastes and not having the algorythm-served taste that I mistook for my own.

Here’s what I have learned so far:

• I have switched from listening music on youtube to listening music on my Digital Audio Player aka good old mp3. It forces you to put your own music there instead of just letting it play.

• I have installed the RSS reader and follow many German and English music magazine/websites and look there for what I could like

• I have also started asking other people like friends and colleauges for recommendations. I have received many good recommendations, but I don’t have to like everything.

I’m fully aware that it’s not a great solution, because it’s a technological step back instead of evolution, but the way music on youtube with ads worked lately has really started to annoy me.

Blogs and small web stuff can be pretty useful.

Curious how you are using Spotify.

I learn about new music by playing Spotify song radio based on a song I like, and then adding the good songs I discover to a playlist.

Any standouts I listen to the whole album and research the artist.

Its not as good as when I was in high school and cross pollinating with the other music nerds bit it is ok.

I personally am a big fan of Rate Your Music. By cataloging and rating my music opinions there I’m able to get interesting recommendations, but moreso I’m able to actively navigate and explore music by genre, year, year-range, influence, even descriptor tag (breakup or migration, for example).

Better is engaging with it as a pseudo-community; finding others with similar tastes and seeing what else they like, or finding review authors whose writing resonates with you and following them.

As for how this pertains to AI, well you have sort of some protection in the crowd effect; obvious AI is likely to get either called out as such or buried in low review scoring. Users of RYM aren’t infallible of course, but in much the same way that the average lemmy user is already a comparative poweruser when compared to the average reddit user just due to the narrowing of field and additional barrier to entry serving as a filter for those less invested in decentralized internet spaces, the average RYM user is more music savvy and discerning as a listener than the average Spotify listener*️⃣, most of whom just want something on in the background while they do their errands. It’s not going to be a foolproof solution, but at least RYM feels more to me like navigating music through the lens of listeners rather than marketers and algorithms.

(*️⃣This is not a value assessment, I’m not saying RYM users are better listeners or inherently have better taste than anyone, just that they’re on average more invested in active music listening as a hobby.)

• Bandcamp
• see who tours together. If you like A and they open for B, you might like B
• indie record labels
• see who’s playing at local venues.