Information theory: Knowledge and know-how

Journal name:
Date published:
Published online

Philip Ball examines a study unpicking the broad ramifications of information flows.

Why Information Grows: The Evolution of Order, from Atoms to Economies

César Hidalgo Basic: 2015 ISBN: 9780465048991

Buy this book: US UK Japan

Ivan Alvarado/Reuters/Corbis

Lithium is produced from brine in Chile's Atacama Desert.

The Atacama desert in Chile is the world's biggest single source of lithium. But Chile does not make its own lithium batteries; for that, the metal is exported to countries as distant as South Korea and China. Why?

“Our world is marked by great international differences in countries' ability to crystallize imagination,” writes César Hidalgo in Why Information Grows: that is, in the ability to turn the inventiveness of the human mind into concrete, saleable products that embody information.

Hidalgo's book is largely concerned with explaining why those differences exist, and what the economic consequences are. In doing so, Hidalgo, a Chilean physicist at the Media Lab of the Massachusetts Institute of Technology in Cambridge, aims to embed his ideas in a broad view of how information economies work both in society and in nature, from Silicon Valley entrepreneurism to genetics.

Information can now be accessed so easily — you can find out about pretty much anything from anywhere, or so it sometimes seems — that it might naively be expected to reduce discrepancies in the local know-how needed to “crystallize imagination”. But differences in the character and performance of national economies persist in ways that are not simply the result of inertia; nor can they be explained solely by different labour costs. The core of Why Information Grows is about the acquisition (or not) of the capability to put information to use. Many of Chile's recent exports to South Korea were copper (another resource that Chile has in abundance). The reverse trade was largely in vehicles and their parts. The trade balance lies in Chile's favour financially, but Hidalgo points out that this masks a deficit in the flow of embodied information. That does not, of course, mean that Chilean automotive engineers are ignorant. Rather, the know-how to make cars reflects a whole suite of considerations about capacity: a trained workforce, access to materials, distribution channels, reputation, trading partners and so on. To truly understand these macroeconomic transactions, Hidalgo argues, we need to delve into the infrastructure and networks that make knowledge productive and allow know-how to accumulate.

Part of this understanding has already been developed by economists, in particular Ronald Coase, whose 1930s work on transaction costs and the origin and growth of firms won him the 1991 economics Nobel. More recently, social capital — connections, and the trust they engender — has been recognized as a key factor in the health of economies. Hidalgo's particular contribution here is to weave these ideas into a theory that provides a conceptual picture of a society's collective accumulation of productive information “in firms and networks of firms”. He argues that these entities are not so much the result of that accumulation as the means by which our societies have evolved to make it possible.

What emerges is a new measure of economic complexity that acknowledges not just gross domestic product, but also diversity in the types of product that a nation produces. This index, Hidalgo says, offers a metric that predicts long-term economic growth. His thesis here is dense and not always lucidly explained, but it contains some innovative thinking about what drives growth that could help us to navigate the turbulence of the ever more interconnected global economy.

Hidalgo wants to take his ideas about information and its productive exploitation beyond economics, especially into biology and ecology. In a nutshell, the task is to restore notions of meaning to information theory. The seminal text The Mathematical Theory of Communication (Univ. Illinois Press, 1949) by Claude Shannon and Warren Weaver explicitly expunged the idea that information in itself has meaning, as Hidalgo reminds us. But we often persist in thinking otherwise, not least in the delusions that we can do science by mining big data without a guiding theory, or that genomes contain 'instructions' for making an organism.

The distinction between knowledge and know-how, which seems evident enough in economics, can be profitably made in the natural sciences too. A gene sequence is information; know-how refers to the ability to make something from it, and this cannot reside in the information itself. “DNA has no knowhow and cannot unpack itself; it is a slave to the machinery needed to unpack it,” says Hidalgo. This seems obvious, but it is worth restating, if only to remind us that this unpacking process (and not further sequencing per se) is now the most important unanswered question in genomics.

Whether all this really justifies Hidalgo's claim to explain “what information is, where it comes from, and why it grows” is another matter. For one thing, with quantum mechanics being reframed as an information theory, it is unclear whether any classical arguments can give the whole picture. But that does not detract from the stimulating new perspectives on offer. Hidalgo has identified a fertile seam, and all his book really needs (apart from some prose-tightening) is for the title to be turned into a question.

Additional data