Using suggestion models based on curated reputable crowdsourcing as opposed to simply volume of popularity is a way around this.
Here's a contrived example; say I'm looking for something that goes well with say, Lonnie Smiths' 1980 "In the Park" (http://www.youtube.com/watch?v=uBPSf-VoDZk) and say, Jon Lucien's Listen Love (http://www.youtube.com/watch?v=6Bm7c0z_0ws) and let's toss Fela Anikulapo Kutis' Witchcraft in there (http://www.youtube.com/watch?v=31cGWpe8_L0).
There is no software in the world that can curate music like that based on predictive modeling that I've seen.
I've been trying to make one for years that can span decades, nations, and genres like that.
I've found that if I aggregate a selective circle of people than I can get there. But that's the only way I've found so far.
Ok, USE CASE 2.
What's Justin Bieber or Keith Richards currently listening to? What about a talent scout at BMI? What if you could follow and tap into those people's personal playlists like twitter allows you to tap into what's on their plates for dinner and other personal things?
And I mean right now. Shania Twain hits play and it sends that to the server. Thousands of users world-wide tapping into her personal radio station immediately get a non-time-shifted instant feedback.
I think it would be compelling and addictive.
It's the classic celeb allure that's been twitter's mainstream pot of gold.
http://news.cnet.com/8301-31001_3-57332246-261/grooveshark-e...
http://www.digitalmusicnews.com/stories/101311cc
http://www.digitalmusicnews.com/permalink/2012/120405grooves...
The last record I bought (a single by The Greys) I listened to on recommendation of a coworker, who's in a band that I like, who had played with some people who are playing a show with The Greys. I also like going to a local record store (Soundscapes) and just buying what the people who work there recommend. Nobody is currently capturing these signals; Facebook is obviously the one company that has access to perhaps a sufficiency of this data, but music recommendations are seemingly pretty small beer in the Zuckerbergian scheme of things.
The lack of a decent music (or film/TV) recommendation system looks a lot like a business opportunity, but I think that it's fundamentally not a business. Recommendations can drive business to a retail operation, but how do you sell recommendations qua recommendations?
obviously no product is going to fit everybody's use case. but if you look at the number of links to music videos on youtube that are passed around every day, it's pretty clear that music on twitter fits a lot of people's use case.
Problem solved!
IIRC Pandora has a catalog size of about 1 million vs ~20 million for Spotify and yet they have been able to do a superb job in song recommendations. I find Spotify radio to be pretty terrible at this.
To me the word "discovery" means finding unknown works, not simply another way to listen to an artist who is already well established.
Pandora for discovery is a nonstarter, at least for me.