It seems that Spotify doesn’t accurately count plays for tracks under 30 seconds. I had a look through some of the albums I know of with short tracks and here’s what I found:

• Humanz - Gorillaz / Most popular track: 214m plays vs 11 second interlude: 1,185 plays
• Absolution - Muse / Most popular track: 259m plays vs 22 second intro: 32,174 plays
• California - Blink 182 / Most popular track: 158m plays vs 16 second track: 42,616 plays
• 3 Feet High and Rising - De La Soul / Most popular track: 12m plays vs 17 second track: 1,117 plays
• The Eminem Show - Eminem / Most popular track: 1.9b plays vs 22 second track: 42,050 plays

In fact I found the MGMT album 11​•​11​•​11 has a 26 second long track with no reported plays.

So while it’s probably true that people are listening to fewer full albums than they used to in the days of vinyl or CD, it seems more likely to do with Spotify only paying royalties on tracks over 30 seconds and the discussion surrounding that.

Oh I see, that makes sense. Sorry that I interpreted it differently based on the post!

I don’t mean to be rude since this seems like a good script and it’s really cool that you were able to get ChatGPT to help. I just find the idea of people writing code they don’t understand using an LLM, publishing it, and then expecting someone else to check that it’s safe a little bit troubling. Credit to you, you’ve been transparent about that being the case but it got me thinking about other developers who might not be so upfront.

This seems really powerful; I’m curious, how helpful were its suggestions?

Mlem on iOS seems to have that feature under Settings > Filters at least!