May 31, 2007


How much does the Internet weigh?

How heavy is information? Most of us know that computers represent all types of information—e-mails, documents, video clips, Web pages, everything—as streams of binary digits, 1s and 0s. These digits are mathematical entities, but they are also tangible ones: They are embodied and manipulated as voltages in electronic circuits. Therefore, every bit of data must have some mass, albeit minuscule. This prompted DISCOVER to ask the question: How much would all the data sent through the Internet on an average day weigh?

link (via Andrew Sullivan)

Worth reading not just to see the math but also for its elegant conclusion.

posted by John Buaas in internet, science, technology 

comments

3 Responses to “How much does the Internet weigh?”

  1. Andrew Simone on May 31st, 2007 at 10:22 pm

    Best article I have read in months. Thanks, John B.

  2. Tom on June 2nd, 2007 at 9:11 am

    I’m not buying it.

    What he measured is not the weight of the information, but the weight of the medium carrying the information. The same electrons arranged in a random configuration wouldn’t weigh any less would they?

    Besides, he just considered the electrons involved in encoding the information when it’s in a memory chip. What about when it’s on disk? How many electrons are involved then? What about when it’s on the wire? I don’t know what the numbers are, but I feel confident they’re different. Does that mean the information has different weights depending on the medium in which it is encoded? It makes no sense to me to call the measurement the weight of the information if it changes with the medium.

    Suppose the Internet ran on vacuum tubes, which would use many more electrons to record a bit than solid state capacitors on integrated chips do. Would that mean that the information weighed more? That doesn’t make sense. If information has weight, it should weigh the same no matter how it’s encoded.

    We can also imagine storage devices in the future that will be more efficient than current capacitors, using fewer electrons to record each bit. Will that make the information weigh less?

    Finally, consider the same information stored in a human brain and in a computer. Take for example, the word “dog”. Based on the argument in the article, we can compute the weight of the word “dog” when stored in an electronic memory. I don’t know how many electrons would be involved in storing the same data in my brain, but I bet it wouldn’t be the same number. If it’s the same information, it should weigh the same regardless of medium. The notion of information having weight seems meaningless to me.

  3. Michael Grant Smith on June 2nd, 2007 at 10:30 am

    Tom, you make some interesting points. Here are some random comments in response to them.

    The Internet does not run on vacuum tubes, but is in fact itself a series of tubes, the tubular kind. There are trucks involved, too:

    http://blog.wired.com/27bstroke6/2006/06/your_own_person.html

    If I, in my own opinion, question the existence of “dog”, do the electrons that store the information about “dog” actually exist, or are they simply faith-based electrons? Are those electrons heavier or lighter than secular electrons (which I call selectrons)?

    If Michael is in the woods with no one around to hear him and he speaks, is he still wrong?

    How many electrons can dance on the head of a pin?

    Which is heavier, a pound of lead or a pound of electrons?

    Thank you.