Because it seems to be that this is the way the world actually works

Interview with Prof. Melanie Mitchell, Davis Professor for Complexity at the Santa Fe Institute, Palo Alto, on why we should learn about complex systems science.

Foto Minesh Bacrania Photography, Text von Johannes Wiek und Sebastian Benkhofer

We keep hearing that we live in an increasingly complex world. But what does that mean?

Typically, when people say we are living in a more complex world they are thinking about all of the interconnections in society, in economies, or even in biology, that we’re becoming more and more aware of. And the interconnections are really what make a system complex. A complex system can be thought of as any system that has a large number of components. This can be individuals in the economy, or neurons in the brain, or insects in a social insect colony… It’s a very broad idea. These components have interactions with each other, and the interactions are such that you can’t really separate the components and add up their affects. We have the saying “the whole is more than the sum of the parts” – and that’s really a summary of what this idea is: That you can’t just take the parts and sum them up and predict what’s going to happen with the system. And that’s true in any complex system.
A more technical term for that is nonlinearity. That the system cannot be modeled in a linear form which is essentially summing up the components.

Can you give us an example of what complex systems science is trying to study?

A great example is the human genome project. It was touted as a way that we’re going to understand what genes give rise to behaviors or diseases or other kinds of biological phenomena.
But even when all of the human genome was sequenced and we knew what the genes were, we still didn‘t know how to predict or to cure diseases or how to predict biological phenomena in general. Because it’s not just the individual genes. It’s their interactions in a big network. And the same can be said of studying economic systems, of studying the brain and studying social systems like cities and so on. Complex systems re­search tries to focus on what tools we have for understanding very interconnected systems that have nonlinear inter­actions and therefore have hard to predict global behavior. And what these systems have in common. What does a brain have in common with an economy? What does a city have in common with social insect colony? These are the kinds of things complex systems scientists think about. And we’ve seen for instance with the latest COVID-19 pandemic how the interconnection of society makes things very hard to predict and to figure out how to intercede to improve the outcome. And that’s quintessential complex systems science.

What methods does complex system science use to observe and analyse these different kinds of systems? And what findings can you report?

One example is the science of networks or network analysis. In this sub area we look at the structure of interactions rather than at the individual components. There are different statistics that can show us things that just averaging the results of the system don’t show. One example is from COVID-19.
Everybody has heard of the R-naught or R 0 which is the infectivity, or how many people an infected person is likely also to infect. That R-naught is a useful statistic, but it has a problem: It is an average. I can say the United States has a certain R-naught, or Germany has a certain R-naught. But that’s an average over a very vast number of people. Rather than looking at an average we should look at hubs in the network of COVID-19 spreaders and where they’re located, how they are connected, and what their interac­tions look like. And it is a relatively simple network science concept that looking at average quantities is not as useful as looking at quantities like hubs of networks and their connectivity patterns. But that is why complexity science, which has not been mainstream in the last decades, has been gaining ground in fields other than epidemic research.

What other areas of complex system science boast the greatest breakthroughs?

I think economics is one example. Economic systems have often been modeled in a rather linear kind of way. When people were looking at the economy before the financial crisis of 2008, they were looking at models that were based on gaussian distributions of risk. But it turns out that a lot of economic phenomena like risk distribution of bubbles or crashes don’t follow a gaussian but something that we call a power law distribution. And these power laws have something called the long tail, which means that unlikely events like huge crashes or other kinds of failure modes happen more fre­quently than you would predict if you’re using gaussian.
And because real economic phenomena don’t follow these gaus­sian rules there has been the rise of what is called complexity economics that a lot of people are getting very interested in – even if it probably is still not in the mainstream.

Why should students be interested in this field of research?

One big reason is that it seems to be that this is the way the world actually works. If they are trying to understand natural or social phenomena this is a more realistic way of thinking than looking at more linear models. And I think looking at these systems from this more interdisciplinary perspective is also much more enlightening. If you’re an economist, you typically take classes in economics and learn about the history of economics and so on. But it broadens the perspective if you think about the economy as something similar to other kinds of complex systems, something that has a lot of nonlinearities and has a network structure and adapts.

One of the most prominent Authors in this field of risk analysis is Nassim Nicolas Taleb, author of the bestsellers The Black Swan (2007), Antifragile (2012), and Skin in the Game (2018):

For the foundation of thinking of the economy as a complex adaptive system see: W. Brian Arthur: “Complexity and the Economy” (Science, 1999)

The Sante Fe Institute’s associates network includes researchers from more than 80 institutions in 20 countries. Students and academic initiatives of our university are invited to make contact and to approach the respective researchers with questions or to invite them into seminars via Zoom.

Melanie Mitchell is considered one of the best mediators of complexity thinking. Her FREE ONLINE COURSE “INTRODUCTION TO COMPLEXITY” has been taken by over 25,000 students and is one of Course Central’s “top fifty online courses of all time.”

Melanie Mitchell: “ARTIFICIAL INTELLIGENCE – A Guide for Thinking Humans” (Oktober 2019), Farrar Strauss & Giroux

Melanie Mitchell: “COMPLEXITY – A Guided Tour” (April 2011), Oxford University Press