Genius

One of the features Apple pushed at yesterday’s event was Genius—a web-based super-brain that automatically determines which of your songs go together, and puts them into a playlist.

In and of itself, Genius works quite well. But what interested me was how it works. Steve described it in his keynote as “genius algorithms” on the iTunes servers, that are constantly evolving, as they receive more information from iTunes and iPod users around the world.

You send information about your music library to iTunes, where this ‘brain’ analyzes it, runs it through the algorithm, and spits it back out to you. This is all done anonymously, but the more information this mega-brain of information feeds on, the smarter it becomes.

Brilliant, right?

Then I took a look at some of the information Genius is uploading: track names, play counts, ratings, and playlists. Seems like pretty standard stuff. But when you look at the power this gives the iTunes Genius, it’s easy to see why Apple is pushing it so hard.

Apple will now have a database of millions of tracks, along with their average ratings and playcounts. Depending on how complex the algorithms are, they can determine which music-listeners prefer certain music, and can collectively make more accurate music recommendations. Millions of playlists being sent to Apple will no doubt aid in how the iTunes Genius pairs your music as well.

While this idea sounds great in theory, you need something that no one other than Apple has—a huge music user-base. Sixteen months ago, Steve Jobs touted that iTunes had received over 300 million downloads. It’s safe to say that today, iTunes is installed on hundreds of millions of computers, and is the number one digital music player in the world.

The thought that Apple has a super-evolving mega-brain that is intelligently sucking up detailed information about people’s musical tastes from tens of millions of iTunes libraries is mind-boggling.

For one, imagine the improvements to the iTunes Store. Apple could easily add an average rating to every single track and album, based on millions of samples. Even more powerful, they can show only average ratings based on users with similar musical tastes as yours, further targeting your specific preferences. Popularity of a track can be determined by the number of times it’s been played, and music can be grouped together into similar groups and playlists.

Even more powerful from Apple’s perspective is the ability to give sell this information back to artists, bands, and record labels. All of a sudden, the music industry has a one-stop place where they can go to find out information about their target demographic—which tracks are popular among their audience, and what other similar artists and bands they enjoy listening to.

Apple was right by calling it Genius. They don’t want it just to be an algorithm that your music is run through. They want Genius to become the ever-evolving brain that knows everyone’s musical tastes.

And of course, in the end, it will simply be used to sell you more music.