the thing is, spotify banned usage of their features in stuff like AI models which would help out greatly with recommendation engines
2024-10-01 27516, 2024
duckfromdisc[m]
so if someone else made a freely available dataset or model that gave some sort of features and it was open source/public domain/etc, we could contribute audio features linked up with something like musicbrainz similar to how acousticbrainz did it... just with more useful models
2024-10-01 27527, 2024
duckfromdisc[m]
it just seems a bit unfair that only spotify can/will do this stuff is all
2024-10-01 27544, 2024
BobSwift[m] joined the channel
2024-10-01 27545, 2024
BobSwift[m]
<duckfromdisc[m]> "MeB?" <- MetaBrainz. The parent to all of the various *Brainz projects.
