Google Music aims “to provide music to make the things we do every day, better. Waking up, working out, commuting: everything can be better with the right tunes. Music will now use contextual information—day of the week, time of day, and device, for starters—to figure out what you’re doing and what you might want to listen to. Then, you’ll get a playlist full of songs perfectly tuned not just for you, but for right now.” (quote from from “Google Music Crunches Your Data to Craft Perfect Playlists“)
Really? Good luck with that.
Obviously I haven’t tried this, but I can make a few straightforward critiques.
First: no, you do not have permission to track me this way. Not. A. Chance.
Second: even if you have a bunch of context and history, I defy you to know what I want to listen to right now. If I don’t know what I want, how can an algorithm know?
What does “better” mean anyway? There are an infinite number of things I might hear next, and some sub-infinity of them will be “good” in different ways. What in the heck is the algorithm optimizing here?
Third: this is a preposterous and idiotic thing to want to try to do. What problem do they think they are solving?
David Pierce reports in Wired, Google Music “designed to help people who don’t know what to listen to, or who keep listening to the same 12 songs every day because discovering good new music is too much effort. “They have no idea how to DJ for themselves,” [Google project manager Elias] Roman says, so Google is doing that for them.”
Even if this situation is a real problem (which I doubt), the cure is worse than the disease. Turning my choice over to an opaque and exploitative algorithm can’t be anything but bad for you, even if it works, which is won’t.
Instead of only 12 songs you like, you get dozens of songs the algorithm says you should like.
Just say no.