mead cohen berger shevtsova garfinkle michta grygiel blankenhorn
The Information Revolution: Alive and Well

Science never ceases to amaze. George Church and Sri Kosuri of Harvard’s Wyss Institute have managed to pack about a thousand times more data into DNA than ever before, reports ExtremeTech. The bioengineer and geneticist recently encoded about 5.5 petabits (about 700,000 gigabytes, to use a more familiar unit) of data in a single gram of DNA. The data they used was 70 billion copies of Church’s most recent book, which included not only text but photos and some computer code, showing the versatility of the medium.

Church and Kosuri’s ability to utilize this method may revolutionize data storage:

Looking forward, they foresee a world where biological storage would allow us to record anything and everything without reservation. Today, we wouldn’t dream of blanketing every square meter of Earth with cameras, and recording every moment for all eternity/human posterity — we simply don’t have the storage capacity. There is a reason that backed up data is usually only kept for a few weeks or months — it just isn’t feasible to have warehouses full of hard drives, which could fail at any time. If the entirety of human knowledge — every book, uttered word, and funny cat video — can be stored in a few hundred kilos of DNA, though… well, it might just be possible to record everything (hello, police state!)

As for right now, encoding information into DNA is a relatively lengthy and novel process and so probably far from any practical applications. But it’s a clear sign that the information revolution is nowhere near being over. Not only has Moore’s Law held true for far longer than anyone expected, but the increasing ability to store denser amounts of data in smaller and smaller media suggest that computers are going to continue to improve for a very long time to come.

Features Icon
show comments
  • C. Philips

    “The data they used … included not only text but photos and some computer code, showing the versatility of the medium.”

    Storing computer code and pictures does not show any additional versatility. DNA stores bit sequences, and there are already standard methods to represent pictures at bit sequences (just as for text), so storing text is no different from storing pictures.

  • Jim.

    Can you search it? How do you pay people to sift through that much data?

    One of my friends is the librarian at a major Midwestern newspaper. They’re having trouble finding funding to digitize their archives.

    If people aren’t willing to bother keeping pre-sorted information about important events available for posterity, who in the world is going to pay to interact with random camera fodder in any meaningful way?

  • f1b0nacc1


    You can search ANYTHING if you commit the time and resources to do it. I am a database architect, and the fastest growing part of my business is data mining of just this sort of data. Businesses in particular are spending very large sums on building the infrastructure for big products of this sort.

    The first part of your question though is the more important one, i.e. can you search DNA stored data. The answer is ‘yes and no’, you can search anything, but you cannot search it fast, and for masses of data such as this, fast is crucial. DNA is a write-centric storage solution…useful for archiving but little else at this time. This may change, but it is unlikely to be the case in the near to mid future.

© The American Interest LLC 2005-2016 About Us Masthead Submissions Advertise Customer Service