This past week we made a trip from Buffalo to Long Island and back again (about 16 hours of driving total). On the way back my husband and I were listening to music from my iPhone. (I currently have 9.5 GB of music stored on my 30 GB phone—in case you wanted to know).
Sick of the same 10 playlists, we started looking for “Genius”-worth songs. For those of you not family with iTunes Genius: when listening to a song you enjoy, you can press an icon at the top of the screen (the Genius icon) and it will create a customized playlist using your current songs on the device.
A game ensued to find the “perfect” song that would create the “perfect” genius playlist. (Meaning a playlist of songs we both liked). This got me thinking… how does Genius work? How does Neflix work? What makes Pandora awesome?
The answer, in essence, is The Algorithm. But what exactly is an algorithm? It was at this point I wished I had paid better attention in math class.
According to Wikipedia, “In mathematics and computer science, an algorithm is a step-by-step procedure for calculations. Algorithms are used for calculation, data processing and automated reasoning. More precisely, an algorithm is an effective method expressed as a finite list of well-defined instructions for calculating a function.”
Algorithms are everywhere in the technologies we use today. I’ve noticed that I like some more than others. For example, I love Pandora. I love the “discovery” of new music and artists. After doing some research I found out that Pandora does not look at what others are listening to or what they “Like.”
Pandora has no concept of genre, user connections or ratings. It doesn't care what other people who like a certain artist also like. When you create a radio station on Pandora, it uses a pretty radical approach to delivering your personalized selections. Having analyzed the musical structures present in the songs you like, it plays other songs that possess similar musical traits.
According to a post by Julia Layton on Howstuffworks.com: "Pandora relies on a Music Genome that consists of 400 musical attributes covering the qualities of melody, harmony, rhythm, form, composition, and lyrics.”
iTunes, however, works very differently. According to an article in Technology Review written by Christopher Mims, “Basically, your library of tracks is compared to all the other Genius users’ libraries of tracks. Apple then runs a set of previously secret algorithms, which…[are] recommendation algorithms similar to those used by other services like Netflix when it suggests movies for a user to watch now or add to his queue, to generate statistics for each song.”
The problem I encounter with Netflix is that it is unaware who is viewing the recommendations and that it is used by multiple users. So, in my case Breaking Bad shows up next to My Little Pony. The algorithm is all messed up by the fact that both shows are viewed on Netflix. Consequently, I rarely get a great recommendation.
I think this is only the beginning. At CES this year Samsung announced Video Discovery, “a service that provides consumers with a faster and easier way to find and enjoy live TV and video content. Complete with personalized recommendations and the ability to search across multiple content providers.” Wouldn’t it be nice to sit down and see recommendations from Live TV, Netflix, Hulu Plus, and YouTube all in one place? Not only will the service be available on Samsung’s smart TVs, but also from its mobile devices.
And how will they make this happen? Algorithms!
How do you feel about someone, or some “thing” paying attention to what you’re viewing or listening to in order to make better recommendations? Is your time being better used? The internet has been watching us for years. (Even notice that Google “reads” your email and then makes recommendations by what it advertises? Creepy!)
I, for one, am fascinated by this. I’ll continue to watch and learn about the way algorithms work. Who knows what shows, music, and more I’ll discover.