Aperture Book of the Month - August 2025

J. Doyne Farmer, 'Making Sense of Chaos: A Better Economics for a Better World'. Allen Lane, 2024

Given our focus on helping organisations navigate complexity and uncertainty with our systems-based approach to strategy design, we were very much looking forward to reading J Doyne Farmer’s 2024 publication, Making Sense of Chaos. The book was generally very well received as an important contribution to widening awareness of complexity and the application of chaos theory, with particular regard to the serious flaws of conventional economic theory.

Farmer describes how the economy is not a system in equilibrium that can be modelled with simple linear equations, but rather a complex, adaptive system, where the influence of countless interacting agents gives rise to emergent, unpredictable behaviour.

“The standard workhorse macro-economic models that are currently used by central banks and treasury departments assume that we live in a world of risk rather than one of uncertainty.”

Farmer’s perspective challenges conventional economic theory and provides a more realistic framework for understanding the nature of market crashes, bubbles, business cycles and the like. His background as a founder of a quantitative trading firm and an academic at the Santa Fe Institute gives his arguments an empirical weight, as he recounts first-hand experiences of using these very principles to predict and navigate financial markets. Farmer is clear that while short-term prediction in such systems is possible, long-term forecasting remains a perilous, if not impossible, endeavour.

With that as our start point, we regret to say that because of problems of tone, structure, and theory, we found Making Sense of Chaos a surprising disappointment. Unfortunately, the book’s promise to expose the significant shortcomings of conventional economic wisdom is overshadowed by structural flaws and a conceptual narrowness that prevents it from really making ‘sense’ of complex systems. Specifically, the book suffers from a continuously self-congratulatory tone, significant repetition, and – most importantly – a notable failure to engage substantively with the implications of open systems thinking.

One of the most distracting elements of the book is its inherent self-congratulation. As a pioneer in the development of complexity science with institutions like the Santa Fe Institute, Farmer is uniquely positioned to narrate its history. Unfortunately, the narrative often becomes less an objective history of ideas and more a celebration of the intellectual milieu the author claims to have created. The constant recounting of the initial breakthroughs and the roles played by Farmer and his cohort, while perhaps historically relevant, frequently shifts the focus from the compelling scientific concepts themselves to the personal achievements of the researchers. This pervasive tone can make the reader feel like an observer of an exclusive academic success story rather than a participant in a grand intellectual exploration, thus dulling the impact of the approach being described.

"My background is an interesting one. I made a modest fortune with a friend by beating the game of roulette using the first wearable computer in the 1970s. This experience immediately demonstrated that the world is more predictable than people thought, but that you have to apply the principles of physics and chaos to succeed."

Structurally, the book struggles with persistent repetition. While it is necessary for a text on complexity to re-emphasize core concepts like phase space, sensitive dependence on initial conditions (the butterfly effect), and attractors, Making Sense of Chaos revisits these concepts with an unnecessary frequency that bogs down the momentum of the narrative. In various chapters, fundamental definitions and historical anecdotes are retold without adding new illustrative or contextual depth. This repetition not only makes the book feel longer than it needs to be but also suggests a lack of trust in the reader's retention, frustrating those who grasp the foundational ideas early on. The cyclical revisiting of material acts as a brake on the book’s progression, hindering the development of more advanced, application-focused discussions.

Perhaps the most significant theoretical weakness is the book's limited engagement with the philosophy and methodology of open systems thinking. Chaos theory, as it emerged from physics and mathematics, traditionally relies on defining and observing closed systems – systems where all variables and their interactions are, in theory, known or bounded.

This criticism is particularly relevant when considering the institutional context in which complexity science matured. The Santa Fe Institute, where Farmer and many of his peers did their seminal work, has itself faced long-standing criticism for this very bias. The groundbreaking models developed there – whether related to complexity economics, systems dynamics, or chaos theory – often relied on idealized, self-contained simulations.

While invaluable for establishing theoretical baselines, this approach fostered a conceptual environment where the messy, unpredictable nature of external real-world forces was often downplayed or modelled as simple noise, rather than being treated as an intrinsic, shaping feature of the system itself. This institutional focus arguably reinforces the book's narrow conceptualization of ‘chaos’ as purely an internal phenomenon.

Real-world systems that the book seeks to explain, however, such as economies, global climate, and socio-political conflicts, are fundamentally open. They are constantly subject to inputs and influence from an unbounded external environment that cannot be fully measured or predicted.

“We need models that can help navigate the big challenges that we face, problems like climate change, inequality, and financial instability. And we need more than just conceptual understanding - what we really want are models that can provide quantitative answers, based on testable predictions.”

Farmer’s analysis, despite touching upon these real-world systems, does not substantively wrestle with this critical distinction. There is a lack of rigorous, philosophical acknowledgment that applying closed-system mathematics to inherently open “We need models that can help navigate the big challenges that we face, problems like climate change, inequality, and financial instability. And we need more than just conceptual understanding - what we really want are models that can provide quantitative answers, based on testable predictions.”systems – where variables are constantly being introduced from the outside – imposes severe limitations on analytical frameworks. This omission diminishes the book’s critical value, as it fails to provide a holistic conceptual framework for addressing complexity outside of the historical, theoretical comfort zone of early chaos models. By focusing primarily on internal dynamics, the book overlooks the essential truth that for true open systems, the ‘chaos’ often originates less from internal non-linearity and more from unknown external influence.

We applaud Farmer’s basic thesis, that simple representative modelling as used in conventional economics is unhelpful and even dangerous. We agree with his that there has to be a better way and that we might look to biology and ecology to expand our thinking. We acknowledge that Famer has enjoyed a significant amount of success with his stochastic modelling approach to gambling and the stock market. Despite this, however, we’re left with the view that Farmer’s mindset is itself a closed system. It’s a great shame that his ‘alternative’ approach doesn’t go anywhere near far enough in recognising the potential contribution of open systems thinkers to sensemaking and decision support.

It's even more of a shame because Farmer introduces a host of really interesting characters and ideas. There’s much to learn from this book, from the early thinking of Herbert Simon and Frank Knight to the unforeseen drivers of systemic risk that shaped the 2008 crash, and from bounded rationality to the Nash Equilibrium and Wright’s Law. But our feeling is that Farmer has missed an opportunity to influence more strongly than he does. More importantly, whilst we accept that stochastic and agent-based modelling certainly have an important role to play in addressing the challenges of complexity and uncertainty, and that they are certainly more useful than conventional economic modelling, our contention is that replacing one model with another, albeit more sophisticated and nuanced, doesn’t provide the necessary mindset. Investment is required, but not just for producing more intuitive and realistic models.

Farmer concludes his book with a pitch for funding agent-based modelling to inform decisions about sustainable growth in the global economy, reducing inequality, avoiding financial crises, recovering from natural disasters, and reducing climate change. Unfortunately, although some – clearly including Farmer – would argue that agent-based modelling is the place to begin, it’s going to need a significant change of mindset to make an impact on any of those intractable challenges. It will help, but we have to report that for us, ‘Making Sense of Chaos’ doesn’t.