A Conversation with J. Doyne Farmer: Author of Making Sense of Chaos: A Better Economics for a Better World

John Dalton, VP of Research at FCAT, had a chance to catch up with Professor Farmer about his new book, and where he thinks complexity economics is headed.

1. Complexity is a pretty daunting topic. How would you describe a complex system?

   When we study complex systems, we’re studying emergent behaviors that are qualitatively different at the high-level compared to the behavior of the individual interacting components. My favorite example is the brain. A neuron is a complicated cell, but if I just look at a neuron on its own, I wouldn’t have any idea that you could hook up 80 billion of them and get a conscious human being. The economy is a wonderful example of a complex system, and so is the climate. As I point out in the book, we’ve learned a lot about how to simulate economic activity from advances in climate modeling.

2. If I’m not mistaken, this is your first book. Why this book and why now?

   I just felt like it would be good to communicate directly with the public rather than talking only to economists. I’ve had the opportunity to talk with people globally — from Abu Dhabi to Norway — and some of those who have read the book got a hold of me with interest to act upon what they’ve learned. People seem motivated to dig into these deeper ideas and do something with them.

3. You and your team used complexity economics to predict the impact of the COVID-19 pandemic on the economy, and your model outperformed other forecasts by a longshot. Given that kind of success, I couldn’t help but wonder — why is complexity economics still in its infancy when the idea itself has been around for a while and has this kind of data backing it?

   Some of the ideas go back to the Sixties, when Herbert Simon argued that people don’t make myopic decisions but instead use heuristics or rules-of-thumb, which he called “bounded rationality” — a core idea in complexity economics. However, Simon didn’t have the tools we have today. We have computers that are a billion times faster, allowing us to simulate systems from the ground up, as they are, and, not as some theory says they should be.

   As an additional factor, economics is a rather conservative field. Graduate students are sometimes told that if they do agent-based modeling, they’ll ruin their career because it’s not the standard process, and so, it's been a fringe activity largely pursued outside of economics departments — except for a few avant-garde departments in Europe. But I think we are at a kind of a tipping point.

   Economics is a bit stuck because behavioral economics isn't always effectively integrated into economic theory, and people are aware of that. Within macroeconomics in particular, people are hitting a wall because those models are so hard to work with, but because of all this, I sense a softening of perspective.

4. It’s worth noting that the book isn’t just about economic modeling. You also argue that within the realm of agent-based modeling, we might one day become a “conscious” civilization. What do you mean by that?

   In this sense, a conscious world is one that we're able to model ourselves. What is consciousness? It's an awareness that I, as an individual, exist. I'm different from other conscious beings. We're now able to start modeling ourselves in a way that gives us better guidance, so in many respects, we’re already pretty good at this.

   Modeling the climate of our natural world is actually a great example. We're aware of a problem that's really 50 or 100 years out on the horizon, and we're starting to take actions to deal with it now. So, I’m hoping that by modeling ourselves, we can understand how to improve systems like the economy or democracy, as well as better understand the cause and effect of our collective decisions.

   With these tools and a shift in perspective, I think we're seeing some promise of being able to make civilization work better than it is now.

John Dalton is VP of Research at FCAT, where he studies emerging interfaces (augmented reality, virtual reality, speech, gesture, biometrics), socioeconomic trends, and deep technologies likesynthetic biology and robotics..