Last week I read about a research team at Harvard lead by George Church that encoded Church’s next book in DNA. As the write-up in the Harvard Medical School web page, http://hms.harvard.edu/content/writing-book-dna, said:
“Although George Church’s next book doesn’t hit the shelves until Oct. 2, it has already passed an enviable benchmark: 70 billion copies—roughly triple the sum of the top 100 books of all time.”
A Wall Street Journal write-up by Robert Lee Hotz, http://online.wsj.com/article/SB10000872396390444233104577593291643488120.html?mod=e2tw, quoted Church as pointing out that:
“A device the size of your thumb could store as much information as the whole Internet”
The articles go on to talk about the issues involved with the achievement, for example currently it is only possible to read the information sequentially and both reading and writing is slow, all of which in the end are engineering issues which will be solved over time.
The interesting issue to me is how it is increasingly useful to use biological metaphors to drive thinking about systems design and architecture and in more recent times of course using actual biology itself. And while we use the words, I am not convinced we have thought through all of the implications.
Much of our casual language uses such terms. We talk about viruses and worms in discussing cybersecurity threats. However, almost all the time when we design security we return to static design creating deeper moats or thicker security walls to stop entry of bad guys.
We worry about trying to stick a never-sufficient number of fingers in the security holes that occur at the end-points of networks, while we discuss BYOD (bring your own device) to make such goals even more complicated and unlikely to be successful.
At a George Washington University class I lectured to on cyber-security, http://www.ourownlittlecorner.com/2011/05/30/my-gwu-discussion-%E2%80%93-part-3-%E2%80%93-what-to-do-about-cybersecurity/, I pointed out that if you were going to attack viruses perhaps one approach would be to look at Stigmergic systems (I currently look for opportunities at social events to mention Stigmergic systems).
An example of stigmergic behavior is exhibited by Ants who use chemical trails to cause loosely coupled but highly organized behavior. This kind of dynamic change may be necessary to combat the increasingly sophisticated viruses that are being created. The question would then be how to design systems that can exhibit stigmeric behavior.
As an aside during the eleven years I worked at Sun I remember that when Scott McNealy, who may be one of the most brilliant people I ever was fortunate to work for, used to say that his goal for Sun was to have a loosely coupled but highly aligned company I would note that he always got it half-right.
The human body has viruses and bacteria continuously present, yet most of the time it remains pretty healthy and functional (though as I get older, perhaps not so much). What lessons can we derive from that when thinking about cybersecurity and systems design?
Looking at the bigger picture, the Internet itself is approaching or perhaps has already begun to achieve an Asimovian result of being like a large brain. Though I believe unlike a brain it has too many single choke points which need to be designed around and expanded upon.
Perhaps all future software development projects will need to include at least one senior architect whose qualifications included a biological or ecological background.