The Evolution of Technology
To economist W. Brian Arthur, the value of innovation depends on harnessing the natural progression of shared knowledge.
Who determines the path of new technologies: the engineers in the R&D lab, the decision makers of a company (or at a university or government sponsor), or the consumers and citizens who put that technology to use? For many people, W. Brian Arthur’s theories about complexity and innovation have provided the most effective answer. In the 1980s, as an economist based at Stanford University and the Santa Fe Institute (to which he would move full time in 1996), Arthur began to develop agent-based models of stock market behavior. These showed the extent to which irrational factors such as investors’ subjective expectations could bias the behavior of the whole system.
In the mid-1980s, as the full impact of personal computers on business began to be felt, Arthur put forth the idea that increasing returns — the expanding value of a technology as more people use it — could lead economies in unexpected directions. For example, Microsoft’s Windows operating system, deliberately set up as an open-architecture platform, was increasingly valuable because of the large array of hardware and software components that worked with it. The greater the number who used it, the more companies would fit their products with it, and thus the greater the number who would use it. Such a system would be vulnerable only to another product with an even stronger network effect (such as Google’s search engine, eventually). Arthur’s concepts provided an apt strategic explanation of the ways in which innovative products, surrounded by networks of avid users, came to dominate niches in the technology industry. (See Joel Kurtzman, “An Interview with W. Brian Arthur,” s+b, Second Quarter 1998.)
Arthur’s latest work, The Nature of Technology: What It Is and How It Evolves (Free Press, 2009), delves more deeply into the meaning of innovation over time. He observes that new technologies are put together from existing technologies. Thus, the collective body of technological endeavor evolves by creating new elements from within itself. It forms a system that is constantly changing in ways that nobody can quite predict. This evolutionary process provides the impetus in turn for large-scale changes in science, the economy, and much of human culture.
“We put together pieces of metal alloy and fossil fuel so that we hurtle through the sky at close to the speed of sound; we organize tiny signals from the spins of atomic nuclei to make images of the neural circuits inside our brains; we organize biological objects — enzymes — to snip tiny slivers of molecules from DNA and paste them into bacterial cells,” Arthur writes. “Two or three centuries ago we could not have imagined these powers.”
Recently, s+b spoke to Arthur at his home in Palo Alto about the evolution of technology and how people and companies can harness innovation for their own benefit and the benefit of the world around them.
S+B: You wrote in your new book that an economy is an organizing system for technology. What did you mean by that?
ARTHUR: If you consider what an economy consists of — organizations, laws, markets, banking systems, and so on — you realize that human beings have created an enormous system of means or arrangements to meet our needs. And then when you look closely at all of these arrangements, which have become enormously complicated, incredibly interlinked, hyper-communicative, and very much dependent on each other, you realize that they are made up of a huge panoply of technologies. I find this actually quite marvelous — that one of our primary accomplishments as human beings is to get ourselves organized to meet our needs, and we’ve done it in a brilliant way that’s evolved over centuries.
S+B: By “technologies,” do you mean just machines, or a broader definition?
ARTHUR: There are very recognizable physical technologies, such as radar or gene sequencing or chemical engineering, but I wanted to open up the whole discussion of innovation to a wider class of things beyond the obvious. Because if you view technology as the centerpiece of the arrangements of an economy, of these means to a purpose, then questions must be raised. Are shoelaces technologies? Surely, they are means to a purpose. I’d have to say yes. What about traffic lights, tax laws, university administrations? Are those all means to purposes? Certainly. What about venture capital financing? In a very real sense, in the sense that they were developed by humans as new ways to meet some need, these are all evolving technologies and innovation.
The economy thus isn’t something that contains a few technologies. The economy is its technologies and is an expression of its technologies — in the same way that an ecosystem emerges from its species. And if an economy consists of its technologies, then it automatically changes the moment you change any of the technologies. If a part of the legal system or venture capital financing is altered, changes reverberate throughout the entire economy, and not just in the parts of the economy most directly impacted by what has occurred. Any time a new species comes into the ecosystem, the ecosystem changes.
S+B: How can decision makers use this concept of economies and innovation?
ARTHUR: Most people think of innovation as a specific action done differently than anyone did it before. So, for instance, if you come up with a better way to sequence genes, that’s innovation.
But once you start to see the economy as an evolving system of technologies, you begin to realize that an awful lot of innovation has to do with understanding the new possibilities that exist. For example, starting around the 1960s, the possibility of digitization came along. Someone could take, say, a business process and redesign it inside a computer and get something very different.
The movie business initially used computers just for accounting and bookkeeping and the like. But in the 1980s, moviemakers began to re-express their existing graphic special effects through the new digital languages. In other words, the movie business combined some of its old processes with some of the new possibilities to create completely new technologies. Then it started to use those technologies for new types of filmmaking, and for other things. That’s not the same as just adopting an existing set of tools and techniques; that’s a much more alive and dynamic activity. Call it innovation.
For decision makers, the payoff of this idea occurs when they see some big new technology about to come into the economy and change everything. If you can adopt and adapt this change in some new way, you can be in a very privileged position. Just look at messaging capabilities. In the 19th century, we embedded message sending in the economy within electronics, through technologies such as the telegraph and telephone. Later, we embedded message sending within photonics, through fiber optics. Then computer hardware and software led to the ability to send messages almost instantaneously and for everything to communicate with everything — and this has defined business processes in the economy during the last few decades.
Fifteen years ago, I would have walked into an airport with a piece of paper and somebody at a desk would have looked at the ticket and done certain things to it to show that I was allowed to board the plane. Now, I come in and maybe I put a credit card or a frequent flyer card under a machine and instantaneously there’s a whole conversation from that machine to other devices that say, “Yes, this guy has made a reservation,” and God knows what other things they’re checking on. A whole set of conversations is triggered. It’s all because computers, photonics, and electronics made it possible.
S+B: Does understanding the dynamics of this kind of evolution allow you to foresee innovations that otherwise would not be foreseeable?
ARTHUR: It does give me certain qualitative ideas for the future, and one of them is that there will be absolutely no end to the process of evolution. New technologies become building blocks that get combined to make new technologies. Then whole families of new phenomena get discovered from time to time and we busy ourselves with those phenomena. For example, in the 1700s and 1800s, chemicals were investigated. In the 1800s and 1900s, electricity and electronics were deeply investigated. Since about the 1930s on, we’ve been working on quantum phenomena, and that gave us the laser and the transistor and magnetic resonance imaging. Looking ahead, the key question to ask is, “What new families of phenomena are we investigating in science now, and how will those express themselves in new technologies?”
I’ll give you an example. In 1953, scientists uncovered the structure of DNA. Within 10 years, they’d broken the code for how DNA made proteins and they had begun to understand an awful lot about molecular biology. And within the next 10 or 15 years, they knew how cells create proteins and so on. And then, from the ’70s on, that became translated into technologies like recombinant DNA, which gives us the ability to go into the human genetic system and switch certain genes on and off or insert new genes. So if you want to understand the future, understand what families of phenomena are being uncovered, investigated, and mined at the moment and start to imagine how those are going to translate into useful things or means to purposes.
S+B: Would you agree that when you put these ideas together, they add up to a theory of progress?
ARTHUR: Let’s call it a theory of progression. Basically, what people have done throughout history is to find more and more sophisticated ways to meet their own needs — whether involving clothing, shelter, health care, efficiency, entertainment, or anything else. If we’re lucky, we find new technologies that meet our needs better than they were met before. Our medical needs — say, to bring healthy children into the world or to live longer lives — are far, far better met now than they were 100 years ago. We could say that human beings are getting more and more elaborately organized. Usually that’s seen as a sign of progress. Sometimes, though, I’m skeptical.
S+B: Is the pace of innovation accelerating or is technology just becoming more diverse and complex?
ARTHUR: It’s clearly more sophisticated nowadays. Imagine if we were cave people and had about 20 words to express ourselves, such as “good, bad, yes, no,” and so on. There are only certain things that you can communicate with those words. But we have constantly created new words — new technologies — to expand our old vocabulary. And in so doing we have created new vocabularies for innovation. The result is that we can express sophisticated thoughts because we have lots of words, expressions, phrases, and combinations of them to work with. And in that sense, modern technology has gotten very sophisticated.
But I don’t think human beings are any faster at inventing these days than they were 100 years ago. Maybe more of us are working with technologies — there are more scientists and engineers. Maybe we’re better organized than before. But above all, what has changed is that we have much more to invent with. That creates a certain speed-up in innovation, but I’m not expecting this rate to accelerate to infinity any time.
S+B: Would you characterize innovation as a craft or a science?
ARTHUR: Craft. I’m sitting here in Silicon Valley and I’m saying, “What’s special about this place? It was all prune orchards 40 years ago. Why should it be the hub of innovation now?” It’s because people in this little part of the world understood quite well how to work with certain phenomena in electronics, and then in computation, and then more recently in genetics. They understood the craft, not just the science; they understood innately how to manipulate a technology to produce a newer, better technology. One argument against this would say: “Anybody could have understood that. They could have made recombinant DNA in the ’70s in Estonia. They’ve got good scientists. So did the Soviet Union, or Germany, or Sweden, or Japan.”
But it’s not sufficient to have good scientists or good technical journals or even good universities, any more than it is sufficient to take a recipe book off the shelf to be able to cook Chicken Cordon Bleu. What’s more important is the implicit knowledge of what temperatures to use, and just how much to cook something.
The more advanced the technology, the more craft is required in innovation. Innovation is about shared knowledge: of how to deal with phenomena, of parameter values and what to do when things go wrong — knowing what new pathways to try and what things have already been tried so you don’t have to waste your time on them. And in the computer industry, much of that knowledge resides locally in Silicon Valley.
If you can nurture that kind of craft, it generates innovations like crazy. Akron, Ohio, used to be the headquarters of tire companies. The companies pulled out in the 1980s, but the place had this incredible embedded knowledge about polymer chemistry. And when the tire companies left, people there found that they could innovate wildly in polymer chemistry, doing all kinds of things that nobody had ever thought of. Now the area is known as “Polymer Valley.” You can purposely build up such a craft locally, but it is hard to do. Once you have it going, it’s an enormous asset and shouldn’t be squandered.
- Art Kleiner is editor-in-chief of strategy+business and the author of The Age of Heretics (2nd ed., Jossey-Bass, 2008).