Musings on Entrepreneurship and Innovation
Sunday, September 24, 2006Physical Technologies and the Fitness Landscape
This is the first in a series of posts on my understanding of an evolutionary model of innovation. My goal in this post is to introduce the concept of fitness landscape as a useful way to visualize the relationship between a prospective innovation and its usefulness.
Given my day job, I'm most interested in prospective innovations that might be called Physical Technologies (PTs). So what do I mean by PT?
More formally, technology is defined here as the arrangement of people, ideas, and objects for the accomplishment of a particular goal.
Andy Hargadon, How Breakthroughs Happen
I define technology quite simply as a means to fulfill a human purpose...As a means to fulfill a purpose, a technology may be a method or process or device...Technologies are put together or combined from component parts or assemblies."
Brian Arthur, The Logic of Invention
Physical Technologies (PTs) are methods and designs for transforming matter, energy, and information from one state into another in pursuit of a goal or goals.
Eric Beinhocker, The Origin of Wealth
So, for our purposes, I'll define a PT as simply a device that fulfills a human purpose.
The concept of fitness has sometimes been misunderstood and abused.
This preservation of favourable individual differences and variations, and the destruction of those which are injurious, I have called Natural Selection, or the Survival of the Fittest.
Charles Darwin, The Origin of Species
In our context, preservation - the replication of a PT over time - is a key element of fitness. The other is that fitness is contextual. That is, the fitness of a PT may change as the environment changes.
Given the preceding definitions, a fitness landscape is the graphical representation of varying levels of fitness across the design space of Physical Technologies designed to accomplish a specified purpose. So, in the following 2-dimensional representation, the fitness of B is greater than that of A (illustration by C.O. Wilke, 2001):
In this example, PTs A and C represent local peaks of fitness, while PT B is a global peak. In other words, a design change to the PT indicated by the red dot toward A would increase fitness, but not as high if the red dot design were to move toward B.
The following, 3-dimensional fitness map (from The Moving Peaks Benchmark) helps to illustrate two further points. First, most of the design space yields low fitness. Second, the fitness landscape changes as the environment changes. Such changes can be both endogenous and exogenous. That means that at PT that lies on a fitness peak today can find itself in a fitness valley tomorrow. For instance, automobiles powered by internal combustion engines have proved to be extremely useful over the last century and, as a consequence, have proliferated across the globe. However, increases in the price of oil, concerns regarding political stability, and alarm over the potential impacts of human-generated greenhouse gases could represent environmental changes that may result in the diminished fitness in such PTs.
In such an alpine-like, roughly correlated landscape, short jumps across the design space are likely to result in a modest change in fitness. That means, a short jump is unlikely to allow a design to move off a local peak onto a global peak. A long jump, on the other hand, offers the potential to move to a higher peak, but is much more likely to result in a significant diminishment of fitness. That is probably why Amar Bhide (The Origin and Evolution of New Businesses) has found that successful incumbents with high opportunity costs tend to explore incremental changes in PT design, while aspiring contenders, who start in a low opportunity cost fitness valley, are willing to depart more radically from current designs.
In my next post, I'll explain how deductive tinkering and the evolutionary algorithm help would-be innovators effectively explore the PT fitness landscape.
