It's the chaos, stupid!
I've just finished J. Doyne Farmer's Making Sense of Chaos, and it's one of the most refreshing takes on economics I've read in years. Through the compelling story of his academic journey and unconventional entrepreneurial experiments, Farmer reveals an economy that behaves less like a predictable machine and more like a living, evolving ecosystem, outlining a blueprint for how economics could finally leverage the computational power of our time.
Introducing the reader to the world of complex systems, Farmer questions the assumptions still underlying most macroeconomic decisions: equilibrium, rational expectations, and the "representative agent"—elegant on paper, but often blind to the messy, adaptive, networked reality of the world we live in. He replaces them with an ambitious computational paradigm built on agent-based modeling, evolutionary economics, and network science.
As I moved through the chapters, I was struck by how the uncertainty of business cycles can be illuminated by disciplines that seem, at first glance, rather far removed—like meteorology, where the Lorenz attractor and other mathematical curiosities were born, or even the biology of metabolic systems. Farmer shows how these fields, each dealing with their own forms of complexity, offer powerful metaphors and even practical tools for understanding how economies evolve, adapt, and occasionally spiral into turbulence.
With the exponential growth of computational power, we finally have the opportunity to bring to life the early intuitions behind agent-based modeling—which is, after all, millions of agents running around like a very stressed version of SimCity. And on its side, generative AI now offers a palette of robust models to simulate scenarios or even fill the gaps in our inevitably incomplete datasets.
Taken together, this opens exciting perspectives for anyone willing to better understand, model, and ultimately navigate the complexity of major economic transitions or market movements—typically to anticipate emerging trends arising from millions of micro-decisions operating within policy-constrained systems.
