3 ms·
I'd believe this if we had a music service today with a "recommendation engine" that was worth anything. IMHO it's a complete crapshoot on all the ones I've use
by chadgeidel 11y ago
I'd believe this if we had a music service today with a "recommendation engine" that was worth anything. IMHO it's a complete crapshoot on all the ones I've used. I still think Pandora's is "best" but that's because it's actually recommended music I haven't heard. Most recommendations are of the form "you liked song A from band Y - try song B from band Y".
As an aside - this feels like a "submarine" post (http://paulgraham.com/submarine.html http://paulgraham.com/submarine.html). I can't tell what they are selling though.
- vinceguidry 11y agoI came here to say this. Algorithms are so horribly bad at predicting what I even like, how could they ever get 'what I want to hear right now' anywhere close? I could easily make such an algorithm that would work well for myself. Tag songs with a mood, have different modes that switch what kind of mix I want. Add in the ability to, in the player, designate certain songs to not play in the currently playing mix, and I think I could iterate my way towards never having to skip a song again. It's on my list of things to build 'one day'. But that's a far cry from being able to generalize over what people that aren't me would be like. It's a hobby of mine to ask random people how they consume music. We'd essentially need the power of a full-on programming language to make object frameworks with customizable predicate logic for each individual person that they would have to learn and implement, to improve upon the skip button. The domain is that hairy. Mind reading is not for the faint of heart.
- mej10 11y agoSpotify's Discover Weekly has been really good -- for n = 7.
- chadgeidel 11y agoHmm, I listened to it for a few weeks and it was basically the same playlist every Monday. I'll pick it up again - thanks for the recommendation.