The age of complexity
How to look at the world’s big problems through the lens of complexity.
by Francisco Rodrigues, University of São Paulo.
On 23 January 2000, during an interview with the San Jose Mercury News, physicist Stephen Hawking, then a professor at Cambridge University, was asked what would be the most important area of research in the 21st century. Physics had reigned supreme in the 19th century, with thermodynamics and electromagnetism driving the industrial revolution. In the 20th century, physics continued its success with quantum mechanics, but electronics brought the computer revolution and genetics unravelled the human genome. Today, these areas continue to develop, but is there anything new? Hawking replied: ‘I believe the next century will be the century of complexity’. He saw the 21st century as the century of complex systems. But what are complex systems?
‘Complexity is a fundamental characteristic of the universe.’ — Stephen Hawking, British theoretical physicist.
Before we answer what a complex system is, let’s start with a question: what do the economy, the internet, the brain, food chains and cities have in common? At first glance, perhaps nothing we can think of, but these are examples of complex systems. They are all made up of interconnected parts that together produce emergent phenomena. For example, in our brains, neurons connect through synapses and exchange neurotransmitters, resulting in the emergence of our memory and cognitive functions. Similarly, viruses and bacteria spread between people causing pandemics; information circulates in social networks often generating misinformation; and countries exchange goods and services sustaining the global economy. Emergence occurs in a variety of systems, from anthills to cellular organisms. In a hive, no single bee ‘plans’ or understands the complete functioning of the nest, but together they build complex structures to efficiently store food and care for the larvae. The ‘functioning of the hive’ is an emergent phenomenon, the result of cooperation between bees.
The complexity of life and our society arises from this phenomenon of emergence, where we say that the whole is greater than the sum of the parts (see our previous post on the subject). In other words, we can’t understand the brain by looking at a single neuron, nor can we understand society by analysing a single individual. We need to analyse the connections of the whole system. We need complexity.
Now that we understand what a complex system is, why did Stephen Hawking suggest that the 21st century would be the century of complexity? Basically, because complexity theory provides essential tools for tackling major global challenges. Problems such as poverty, disease, epidemics, global warming, disinformation, mental health, financial crises, species extinction, war and corruption require an approach that takes into account the interactions between the different elements of these systems.
Understanding climate change, for example, requires an analysis of how different regions of the planet are connected and how climate variables affect each other. Changes in the intensity of the El Niño phenomenon impact not only the climate in the Amazon, but the entire planet, due to the network of connections that link different regions of the planet. Similarly, in 2019 we saw how a single case of coronavirus infection in China triggered a global pandemic affecting billions of people worldwide. In 2008, the bankruptcy of Lehman Brothers, one of the most prestigious banks in the United States, triggered a global economic crisis. In other words, the links between elements of the system were fundamental to the spread of crises or epidemics.
Complexity has been around since the beginning of time, so why has it only recently become so important? The answer is simple. We needed two basic ingredients to study it: data and computers. By the end of the 1990s, computers were getting faster, especially personal computers such as the Intel x86 versions. In addition, data became available with the revolution in search engines, such as that launched by Google in 1998. With data and computers, complexity could finally be studied in a quantitative way. It was the age of complexity.
To study complexity, we need a different philosophical approach to that of traditional science, called holism. Holism holds that “the whole is greater than the sum of its parts”. To understand complex systems, we need to understand how the elements are connected. We need a complex network.
This new approach departs from traditional disciplines such as physics, chemistry and biology, where the idea is to reduce systems to parts and study those parts separately. The methodology used in these fields, called reductionism, has led to the great success of modern science, revolutionising fields as diverse as physics and biology. However, reductionism has created very specialised fields of science, so that even scientists in the same field often cannot understand each other’s work.
On the other hand, the study of complex systems usually involves more than one field of research and requires collaboration between specialists. For example, in studies of brain networks, it is common for computer scientists to work with neuroscientists to understand problems that cut across both fields. Thus, unlike traditional specialised disciplines, the study of complexity connects different fields.
Because complexity involves holism, we need to analyse how the elements of a system are connected. In other words, a first component of studying a complex system is its network of interactions. For example, the p53 gene is a tumour suppressor that encodes a protein that plays a fundamental role in cell cycle control, DNA repair and the induction of apoptosis (cell death). Research at the beginning of this century showed that just as the disruption of a highly interconnected node on the Internet can have major consequences, so too can the disruption of p53 have serious consequences [Vogelstein et al.] A simple fault in one gene can trigger a cascade of other faults that lead to cancer.
Similarly, in 2003, a software failure in the alarm system in the control room of a FirstEnergy substation in Akron, US, caused a widespread blackout in parts of the Northeast and Midwest, and in Ontario, Canada, resulting in a loss of more than $6 billion. In other words, if a single element of a complex system is affected, the whole system can be affected. That’s why understanding the network of interactions is so important.
“Reductionism is insufficient to explain the complexity of living systems”. — Ilya Prigogine, winner of the 1977 Nobel Prize in Chemistry.
The connections in a network not only affect the robustness of the system, but also play a crucial role in the spread of information. If one person is infected with a virus or bacteria, the pathogen can spread to their contacts, causing epidemics or even a pandemics. The same principle applies to the spread of fake news. Those who create this type of content can start a process of spreading it to their direct contacts, who in turn pass the information on to their own contacts, gradually affecting a large part of the network.
A striking example is the 2016 case of Pizzagate, an unsubstantiated conspiracy theory alleging the existence of a child trafficking ring operating out of the Comet Ping Pong pizzeria in Washington, D.C., allegedly run by members of the Democratic Party, including Hillary Clinton. This theory was promoted by members of the radical right, conservative journalists and other critics of Clinton, especially at a time when she was already under pressure over her use of a private email server, an unrelated matter. The misinformation was spread widely on platforms such as 4chan, Reddit and Twitter. Hashtags such as #Pizzagate quickly became trending topics, accumulating millions of mentions in just a few days. The structure of the network of contacts was essential for this spread to reach a global scale, demonstrating how connections directly influence the speed and extent of dissemination processes, whether biological or informational.
A second ingredient in the study of complexity is the modelling of dynamic processes using computers. Basically, the idea is to recreate the real world in a computer. This is already happening in games such as The Sims, which simulates people’s daily lives; Microsoft Flight Simulator, which recreates the real world with geographical precision, using satellite data and artificial intelligence to represent landscapes, airports and weather conditions in real time; or Cities: Skylines, which simulates urbanism, infrastructure, economics and environmental challenges with a realistic approach.
But unlike games, scientific computer modelling is about creating algorithms that reproduce our reality using real-world data. For example, we can use SARS-CoV-2 transmission data to simulate the spread of an epidemic and determine the best way to vaccinate a population. Or we can map the connections between neurons in an in vitro experiment and simulate the synchronisation between them to understand epilepsy, which is characterised by abnormal electrical discharges in the brain. Computer models allow us to see how the network of interactions affects the functioning of the system. This computational approach allows us to study problems as diverse as mental and developmental disorders, the interaction between genes to understand genetic diseases, and the effects of species extinction.
By combining two essential elements — connections and computer simulations — it is possible to study a wide variety of systems in different fields of knowledge. This approach reflects the high degree of interdisciplinarity inherent in complexity theory. The same tools used to study the brain can be adapted to analyse the global economy. Complexity therefore offers a new way of doing science, promoting integration between scientists and between different fields of research. For example, understanding the complexity of life goes beyond the exclusive domain of biology. It requires the integration of physics, to understand random processes such as diffusion; mathematics, to build theoretical models; computer science, to perform simulations; genetics, to explore the mechanisms of cell division; and even philosophy, to reflect on the ethical and epistemological implications of discoveries. These fields are interconnected, just as the world itself is increasingly interconnected.
Many research centres have been set up in some of the world’s most prestigious universities to study complexity. At the University of São Paulo, at the Institute of Mathematics and Computer Science, we have been studying complexity for over a decade. Basically, we are interested in the major challenges facing humanity. For example, we have developed methods to improve the forecasting of dengue and Zika epidemics, with the aim of being able to predict the number of cases in the following year so that governments can prepare for new waves (e.g. Lober et al.). We are also interested in developing new computational methods for diagnosing mental and developmental disorders, to enable diagnosis based on medical data (e.g. Alves et al.), as we do today for diabetes.
In collaboration with the Central Bank of Brazil, we are interested in analysing the dynamics of economic crises and ways to avoid them (e.g. Alexandre et al.). Using climate data, we have built models to predict the impact of El Niño in different regions of Brazil in order to prepare for droughts and higher than expected rainfall. By analysing social network data, we are interested in understanding the polarisation that has become increasingly present in our society (e.g. Interian and Rodrigues). In other words, we want to use data and computers to develop methods to improve the quality of life in our society and to anticipate catastrophic events in order to avoid them or even mitigate their effects. (The publications are organised by field in this link).
Despite significant progress, the study of complexity still faces major challenges. These include modelling temporal and multi-level interactions, which are fundamental to adequately describing the dynamics of complex systems. For example, our social relationships change over the course of a day, and we may travel by different means, such as car, bus, metro or plane. In addition, current models often ignore the intrinsic heterogeneity of these systems, such as the variable capacity of routers on the Internet or species-specific characteristics in ecosystems. Interactions between different processes — such as the impact of messages on social networks on the spread of epidemics (e.g. Ventura et al.) — are also poorly understood. There is still much to be explored if models are to get closer to the real world and be able to explain the complexity around us. The study of complex systems is still in its infancy, and many discoveries are yet to be made. The age of complexity has just begun.
If you’re curious about my research, check out this link: https://sites.icmc.usp.br/francisco.
Find out more:
- Emergency: how order emerges from chaos, Francisco Rodrigues.
- Complexity: A Guided Tour, Melanie Mitchell.
- Complexity: The Emerging Science at the Edge of Order and Chaos, Mitchell Waldrop.
