Don't Be a Fragilista

One thing about a collaboration space is: you hear things not intended for you.  This can be good or bad.  About a couple months ago, I overheard Andy Brown make mention of the term "antifragile." I had never heard it before so I looked it up.  To my surprise, it is not yet a legitimate word, but rather the title of a book.  It just so happens that the author if the book, Nassim Nicholas Taleb, also wrote The Black Swan, which was a very eye-opening book for me.

So, I bought the book and started reading it.  Let me be clear -- this is NOT a book review.  This is an "idea review."  Once again, Taleb has found an important idea that seems to have fallen through the cracks of common understanding.  To fix this, he has coined the term "antifragile."  It is an important concept.

What is the definition of fragile?  I think of this as something that breaks easily under stress.  What is the opposite of fragile?  Most people would use words like robust or sturdy.  However, these words imply something that stays the same when under stress.  That's not really the opposite of fragile, which suggests that the subject of fragility will degrade under stress.

Therefore, antifragile things must get BETTER under stress.  Taleb gives us some examples and the one he likes best is nature.  When the environment stresses a species, it evolves and becomes stronger for it.  How does this tie back to what we do?  Well, we build systems (services, software, etc.).  How will they do under stress?  Will they just be robust and sturdy?  Or, can we make them antifragile?  Is it possible to build software and services that benefit from change?

It is the ultimate challenge.  One point about the species example -- although the species may get stronger from adversity, many individuals may suffer and even die.  In other words, individual sacrifices are often necessary for a system to be antifragile.  What sacrifices must we make for antifragility?  I don't have the answers, but one thing I do know: disruption is necessary.

Leaders often feel it is their job to manage the system to avoid disruption.  Taleb would call them "fragilistas."  These are people that remove the normal disruptions of an antifragile system only to cause it to be fragile in that, when it finally breaks, it breaks in a big and sometime irrecoverable way (think Alan Greenspan and the 2008 economic meltdown).  People who try to remove disruptive forces from a system always mean well.  This concept is very similar to one introduced in one of my all-time favorite business books -- Surfing the Edge of Chaos.  The idea is not to eliminate disruptions, but rather to build systems and processes that are responsive to a changing environment.

In the business world, responsiveness translates to increased profits.  At a medical school, it ultimately means better education for our future doctors.  What might we do to be more antifragile?  In the software delivery team, we can build solutions in such a way that data driven feedback suggesting that something in the process is not working optimally triggers changes that the process stakeholders can control without outside intervention.  In other words, we build things that allow users to adapt it to a continually changing environment.

Once upon a time, robust and stable applications were good enough.  Not anymore.  Now we need to build antifragile applications that allow users to benefit from disruptions.  This will likely require a rethink of the architecture that underlies what we do.  The same will be true for all of our services.  Architectures that benefit from disruption may not be new to the natural world, but they are brand new to human designed environments.  We've spent most of our history trying to eliminate disruptions, now we must learn to embrace them.