One of my favorite authors, Nassim Nicholas Taleb, describes things–organisms, objects, systems, et cetera–as being either fragile, robust, or antifragile based on how they react to disorder. This distinction helps us talk about everything from pebbles to companies, and I may use these terms to describe things like software.
The first of the three categories is the fragile. Fragile things generally don’t benefit from disorder (randomness, stress, change, volatility, variation, chaos, and the like). For example, computers don’t like variation in their environment. Hardware requires regulated temperatures, and user errors can have disastrous consequences on software systems. The best condition for a computer to operate in is a highly-regulated, predictable bubble. It is more often harmed by environmental randomness (say, a busted fan) than helped.
The second category describes the robust. Robust things are not affected much by disorder. A rock is more robust than a computer because it can withstand a greater number of physical and chemical shocks. Rocks can endure boiling temperatures, handle gallons of coffee spilled on them, and even resist users throwing keyboards at them. It takes thousands of years for the stress of wind and water to whittle a rock down to dust–rocks are resiliently robust.
The third and final category is the antifragile. Antifragile things can benefit from disorder. When stressed, antifragile things bounce back and return greater than when they started. For example, muscles don’t grow stronger unless they are subject to varying demands–they don’t improve by resting all the time. And while pushing muscles to their limits may cause temporary soreness, those same muscles gain long-term strength.
Antifragility has its limits, of course. With enough stress, almost anything will break. The important idea is that the antifragile has the possibility to improve in a chaotic environment, while the fragile tends to break.
It is also important to note that things are not rigidly placed into one category in the triad. There is no unit of measurement for fragility, so things can only be labeled as more or less fragile than other things. It is also possible to be antifragile in some ways yet fragile in others. A plant can benefit from variation in temperature, yet remain vulnerable to disease.
Now that we know about the triad of fragility, robustness, and antifragility, what do we do? I think that many of the systems we have in place today suffer when exposed to disorder. However, we can improve these systems. Because the world is full of randomness, a system that can harness unpredictability is more likely to survive than a fragile one. In fact, many systems already have antifragile tendencies. So by knowing the distinction between fragile, robust, and antifragile, we can actively work towards making ourselves and our systems benefit from disorder.
How do you think we can create things that gain from unpredictability?
Pingback: The Lindy Effect | Startup Helium
Pingback: Generalist or Specialist? | Startup Helium
Pingback: In Search of Black Swans | Startup Helium
Pingback: My Mom’s Recipe is Better Than Yours: Stochastic Tinkering | Startup Helium
Pingback: In Search of Black Swans | Sheldon's Software
Pingback: Generalist or Specialist? | Sheldon's Software