|
I would highly recommend that folks read Liz Pelly's Mood Machine, which is referenced here via the Harper's excerpt, not just because it shows artists' perspectives on the long tail model that tech pitched to them, but because it can be read as an indictment of the platform model that increasingly dictates how we engage with art and society at large. Pelly's account really gets to the heart of the problem: these platforms are neither neutral nor meritocracies and don't care about the content that they host. Sounds obvious, yes, but reading about the concrete consequences of this fact, namely the aesthetic flattening that results from conditioning the audience to listen to music passively, definitely got me to reexamine how I was using Spotify and consuming music in general. It really drove home the fact that some people treat art as a fungible object, and that these folks are the people deciding what music we're hearing unless we really make an effort to seek it out on our own. Is the model profitable? For some. Good for society? Perhaps not. EDIT: also want to concur with others here that the problem here isn't necessarily AI but how we're selecting what music we're listening to. In the book, Pelly specifically identifies channels like Chilled Cow as being part of the watering down of this genre, since they have a similar incentive to play music at as low a cost to the channel as possible versus playing the best music available to them. |
As much as I love the fact that teens these days are growing up with the same songs I grew up with as a teen, I also view it as a problem. The shared cultural experience that radio generated was powerful.
We're lost somewhat in quantity now and there's really nobody who's helping form and shape taste. An algorithm might find similar songs based on musical features, but the same sounding song over and over is boring. AI just makes more boring songs because it's largely looking to replicate popular song features as well. This can be passable for purely background music meant to fill space with non-distracting sound, but is terrible for active listening.
Radio was good at mixing in variety within the confines of the genre and audience expectations. Heck, many channels use to program to support the mood during the commute and work hours, and outside of those main audience times would allow the DJs to get a little wild sometimes. Growing up the only place you could catch early EDM was on weekend late-night broadcasts on the local alt-rock station.
It's not like radio of a sort does still exist - just download the GNOME Shortwave client and you can drown in channels there. It's just not powered by the marketing that supports Spotify.
edit - I think it's interesting that the comments I'm seeing below this so far are talking about recent radio. I should have been more clear. In the U.S. markets at least Radio "died" during a great consolidation wave in the earl 2000s when Clearchannel and a few other media companies slurped up all the local channels, switched their formats and started playing consolidated playlists.
It really did used to be the case that your local station DJs were local brands, each with their own curation of songs. Some stations would even have local music festivals and were big promoters of local talent. I spent many evenings calling up the local station to request songs to be slipped into the playlist, sometimes to promote somebody I knew and get them some airtime.