| FWIW... I use Spotify quite a lot, and when I first open the app, I'm immediately presented with pre-curated playlists that seem to cater what I might need at that time. On a Sunday morning, I get 'Sweet soul Sunday' or 'Easy Love Songs'. At 9am on Monday it's 'Concentration Music' or 'Background Synth Songs'. Around 2pm, you start to see 'Workout Mix'. Then you have 'Daily Mix 1', 'Daily Mix 2' and so on which are automatically created playlists, based on what I've listened to. They themselves are generally categorized into rough timbres or styles. I think Spotify actually employs 'agents' that curate and search for new music for a given style or scene. I think Apple does too. I very rarely search on Spotify. I've never created my own playlist. If it's relevant, I would consider myself well versed in music. I studied it during my younger years, and whilst I don't these days - it's a passionate hobby of mine. I just wanted to mention this, as you've mentioned things you want to do, that Spotify already does well, in my opinion. I can't speak for the other services. That being said, maybe you could do it better. So I don't mean to discourage you. If you could build something exactly like Spotify, that allows me to filter out songs with explicit content I would snap that up in a heartbeat. LOL. |
The main difference in my service is that we don't do any manual curation using "agents," since the public editorial playlists can actually be quite generic, and not as personalized as creating a mix just for you, based on your tastes.
So, we've basically built an AI tech that can accurately understand music like humans.
Using this, our service can better serve you the right music for your moment, since it can understand your tastes much better (by understanding the music you like/listen to) & serve you the type of music you want, on-demand.
For example, you can tell the app to play some "relaxing music," and we'll instantly serve with you with music that's relaxing (which we didn't have to manually curate for you; it''s done on the fly by our A.I.).
As a result, you're not limited to any editorial selections, and the music you get once won't be the same as what someone else gets, etc.
Also, as another example, if you don't like to listen to aggressive, explit-type tracks, we can automatically learn that & improve the recommendations you get instantly, since our music-understanding AI can pickup the fact that you're not really into explicit-type tracks.
I've explained a bit more on how my service is a little different than Spotify in the comment I've posted above.
I would love to hear your thoughts on this!
Once again, thanks a lot!