Brian Arthur has many fans amongst the kinds of economist, once considered outlandish but increasingly rather mainstream, who think about the economy as a non-linear dynamic system in which the concept of equilibrium makes no sense. Arthur has also written interestingly about technology previously, especially the path dependence of a series of successive technology choices. His latest book The Nature of Technology was reviewed for The Enlightened Economist by Richard Bronk of the LSE. I've just caught up with it myself and concluded that it has strengths and weaknesses.
The main strength is an utterly believable account of how the process of technological innovation takes place, at the level of the specific invention, the cluster of related innovations which enable it to diffuse through the economy, and the macro-system of technologies. Building on Schumpeter and economic historians such as Paul David, Arthur explains the combinatorial dynamics whereby any new technology is the fruit of older technologies being put together to serve a new purpose. “Technologies somehow must come into being as fresh combinations of what already exists.” (p19) He also sets this out in accessible language, a big plus. What's more, Arthur includes in the technology system the institutions and social structures such as money, laws, business organisations which enable the use of specific technologies, so he has in effect a theory of economic growth of a rather specific kind. It has something in common with Will Baumol's vision of The Free Market Innovation Machine.
I also like, although some readers might find it strange, Arthur's conclusion that technology is alive, in the way that a coral reef is alive. It is evolving, albeit through different processes than the evolution of what we normally consider to be living things. He points to the way it is even starting to reproduce itself, with nano-machines replicating themselves for example, or computer networks starting to learn from experience in interaction with their environment, and not in a top-down controlled way. (p189, pp206-7). Maybe this is too far-fetched but it's something I've often pondered myself when thinking about the proliferation of digital connections and wondering what that implies if human consciousness is itself the emergent product of a massive number of neural connections.
However, for all this provocative strangeness, this was a plodding book as well. Arthur spends too much time setting out definitions, too little on the examples which might have brought them to life. I think it's worth a read but ended up being a little disappointed at how unexciting it had been.