I received my first computer in the summer of 2008. It was a wonderful experience. Unfortunately, by some defect in manufacturing, no doubt, it came without any music on it. “Damnit,” I thought, “now I have to make a library.”

I had never made a proper music library before. As my computing activities were relegated to the family computer, I chose not to curate a library to which I would have only limited access. I was faced with a herculean task: create a brilliant music library from nothing. I knew that at college what I had in my library would shape peoples’ perception of me. I recall one story a friend told me about one boozy party. She had been hooking up with a smokin` (her emphasis) guy when a song by “Death Cab for Cutie” came on. While gasping for air, my friend noted that she really liked the song. The guy looked at her, his face contorted in disgust, and promptly departed without a word.

Since the summer of ’08, I have amassed what I consider a fairly decent library. It includes 6,657 tracks and has a runtime of 21.9 days. When I look through it, all I see are the gaps. I have next to nothing from the ‘70s. I have next to no classic rock. I have almost no jazz music. I have, however, been told that 6,657 tracks is a pretty substantial library. So, I did the math.

In the past year and a half, I have acquired 12.2 songs a day, which corresponds to one hour of music a day. I recognize that that is absurd. More surprising, however, is that I believe that the rate at which I acquire music is increasing. When I started building the library, I put a lot of music in the library very quickly. After that, the rate at which I acquired music collapsed to something extraordinarily small. Since then, I believe that the daily rate at which I acquire music has been steadily increasing.

For the sake of conjecture, let’s make the simplifying assumption that it is only within the last week that I started acquiring an hour of music a day. If we assume that the growth of my library is roughly exponential, then in another year and a half I will be acquiring about four hours of music a day. And, in a short three and a half years, I will be acquiring 24 hours of music a day. From that point on, even if I listened to new music all day with no repeats, I would not be able to hear all the music that I was acquiring. If I live another 60 years, then I will be getting 1,600 hours of music a day streamed directly to my deathbed.

I will bequeath to my offspring an estimated 2,700 years of music, which is approximately the amount of information we have stored in U.S. academic research libraries. When idle, computers use about 22 watts of electricity. It would take 8,500 contemporary laptops to hold my music library, which would use about 185 kilowatts of electricity. If this database was left idle for one year, it would consume 1,500,000 kilowatt hours of electricity, which is roughly enough energy to support 1,500 American households yearly energy needs.

I will not, in all likelihood, amass that large of a library. But, we can assume that the size of music libraries on a global scale will grow in this fashion. As more people gain access to computers, the canon of recorded music grows, our libraries get backed-up in more locations, and physical media fades away, the size of the global music collection will grow exponentially and so too will the energy and storage requirements needed to support it. At some point, we will need to detonate one hydrogen bomb every 14 minutes to maintain our libraries. The energy needed to support our information storage will outstrip the available energy supply on the planet.

A Note on Citations: All the facts and figures used in the preparation of this piece are one google search away. The mathematics was performed on the back of a napkin.