MOD: Modeling the Dynamics of Technological Evolution

Santa Fe Institute, Santa Fe NM

Investigators

Abstract

The aim of this study is to develop an empirically based, quantitative model of the dynamics of technological evolution. The goal of this model is to explain how technologies are related in a dynamical network and why certain technologies improve faster than others. The project focuses on examining both incremental improvements and radical new discoveries that are based on fundamental scientific advances. This problem is approached from diverse points of view, with an interdisciplinary team including economists, engineers, physicists, biologists, and a sociologist. While model construction is the main focus of this project, several tools are developed in the process, taking advantage of the researchers' experience in complex systems analysis. These tools include developing best-practice guidelines for making technological performance forecasts. Work on portfolio theory, and specifically in designing portfolios under increasing returns in the energy sector, should also provide a useful decision-making tool for public and private actors investing in low carbon energy technologies. This study consists of an empirical component, theory and model construction, and a simulation component. In the empirical component of this project several large data sets are constructed and analyzed. The project begins with testing of Wright's law, which states that the cost of manufacturing a given technology decreases as a power law when plotted against the cumulative number of units produced. Alternative performance curves are explored, including replacing cost with other measures of performance, and replacing cumulative number of units with other measures of prevalence. The study focuses on both radical new discoveries and incremental improvements, by relating patent and scientific literature data to performance curves. The empirical work includes broad studies of as many technologies as possible, as well as a few in-depth studies of technologies where there is a more detailed view of all the factors that influence technological improvement. These detailed case studies largely deal with energy technologies, an area with societal relevance. The theoretical component improves on existing models of technological evolution. The operating assumption is that technologies must be studied as part of an ecology of related entities, including all internal component technologies and all external technologies that influence these components. The focus here is on making more realistic models of these network effects. Connections to innovation in biology and the role of selection are explored. The simulation component consists of the construction of toy models of technological innovation and a study related open problems in portfolio theory. Wright's Law implies increasing returns, so consequently technologies that are developed early accumulate an advantage over technologies that are developed later. What is the optimal approach to investing when technologies follow Wright's Law? In this context, the portfolio problem defines a highly nonlinear stochastic dynamical system whose properties are far from obvious. The case studies focus on the low-carbon energy sector. Understanding the likely outcomes of public investments in research and development and of the influence of market transformation programs on technological innovation is critical for effective public policy formulation. The focus here is on energy related policy, and in particular on the problem of public investments to create low-carbon energy technologies. The results of this work will be shared with the international private investment and public policy communities through the London Accord project. The results will also be shared with students from a variety of institutions.