Simply Complexity (2 page)

However, as I finish the book and offer it up to potential readers, I realize that the above wish-list can essentially be reduced to just one item: I would wish that you enjoy reading this book, and that it might provide you with fresh thoughts and insights for dealing with the complex world in which we live, and which our children will inherit.

There are some practicalities concerning the book’s content and layout which I would like to explain. The language, examples and analogies are kept simple since the focus of the book is to explain
what
Complexity Science is all about, and
why
it is so important for us all. I therefore avoid delving into too much detail in the main text. Instead, the Appendix describes how to access the technical research papers upon which the discussions in the book are based, and gives a list of Internet websites containing additional information about Complexity research around the world. Having said this, I won’t pull any punches in the sense that I tackle all the topics which I believe to be relevant.
Part 1
of the book takes us through the theoretical underpinnings of Complexity, while
Part 2
delves into its real-world applications. Some of the territory is only just beginning to be explored, with very few answers available for the questions being posed. From the perspective of other scientific revolutions throughout history this might seem to be par for the course. However we are not talking about history here – instead, we are looking at work which is emerging at the forefront of a new discipline. For this reason we will be highlighting where such research is heading, rather than where it has been.

But why should you believe what I write about Complexity? This is a crucially important question given that Complexity Science is still being developed and its potential applications explored. Unfortunately many accounts of Complexity in the
popular press are second-hand, i.e. they are typically written by people who have done little, if any, research on Complexity themselves and are instead reporting on their interpretation of other people’s work. Given the relatively immature nature of the field, I believe that such indirect interpretations are potentially dangerous. For this reason, I will base the book’s content around my own research group’s experience in Complexity. This has various advantages: (i) it reflects my own understanding of the Complexity field; (ii) it represents what I believe to be the most relevant and important topics; (iii) it will hopefully give the reader a sense of what it is like to be at the “pit-face” in such a challenging area of research; and (iv) it ensures that any reader can challenge me directly on any claims that I make, and can demand an informed answer. To facilitate this process of public scrutiny, a complete list of the relevant scientific research reports is presented in the latter part of the Appendix. I also encourage any readers who wish to email me with questions, to do so at
[email protected]

Finally I would like to thank most warmly the following highly talented scientists with whom I am fortunate to enjoy ongoing interactions on Complexity: Pak Ming Hui, Luis Quiroga, Ferney Rodriguez, Mike Spagat, Jorge Restrepo, Elvira Maria Restrepo, Roberto Zarama, Derek Abbott, Chiu Fan Lee, Tim Jarrett, Alexandra Olaya Castro, David Smith, Sean Gourley, Sehyo Charley Choe, Douglas Ashton, Mark McDonald, Omer Suleman, Nachi Gupta, Nick Jones, Ben Burnett, Alex Dixon, Tom Cox, Juan Pablo Calderon, Juan Camilo Bohorquez, Dan Reinstein, Mark Rondeau, Paul Summers, Stacy Williams, Dan Fenn, Richard Ecob, Adrian Flitney, Matt Berryman, Mark Fricker, Philip Maini, Sam Howison, Tim Halpin-Healy, David Wolpert and Kagan Tumer. In particular, I would like to specifically mention Felix Reed-Tsochas and Janet Efsthatiou, who are also my fellow co-directors in Oxford University’s inter-departmental complex systems research group. Many of the above-named scientists have played a fundamental role in the research discussed in this book – I have indicated their contributions explicitly where appropriate. I am also very grateful to Marsha Filion at Oneworld Publications, for her constructive comments on how to finalize the manuscript – and to
my mother and father for gently encouraging me to get a move on and finally finish it.

I would like to express my deepest gratitude to Elvira Maria, Daniela, Nicholas and Dylan. Thank you for putting up with a very complex husband/father while this book was being prepared, and thank you for putting last Christmas on hold.

Oxford, U.K.
2007

Take a look in many dictionaries, and you will find Complexity defined along the lines of “The behavior shown by a Complex System”. Then look up “Complex System”, and you will probably see “A system whose behavior exhibits Complexity”. So what’s going on? Well, unfortunately, Complexity is not easy to define. Worse still, it can mean different things to different people. Even among scientists, there is no unique definition of Complexity. Instead, the scientific notion of Complexity – and hence of a Complex System – has traditionally been conveyed using particular examples of real-world systems which scientists believe to be complex.

This book will take the “complex” out of Complexity, by going to the heart of what connects together all real-world Complex Systems. We will uncover the magic ingredients which make something complex as opposed to just being complicated, and show how Complexity is deeply engrained in our own everyday lives. We will also see why Complexity is set to revolutionize our understanding of science, and help resolve some of the most challenging problems facing society as a whole.

Complexity can be summed up by the phrase “Two’s company, three is a crowd”. In other words, Complexity Science can be seen as the
study of the phenomena which emerge from a collection of
interacting objects
– and a crowd is a perfect example of such an
emergent phenomenon
, since it is a phenomenon which emerges from a collection of interacting people. We only have to look at world history to realize that it is riddled with major events which have been driven by human crowd behavior. Everyday examples of crowds include collections of commuters, financial market traders, human cells, or insurgents – and the associated crowd-like phenomena which emerge are traffic jams, market crashes, cancer tumors, and guerilla wars. Even extreme weather conditions such as floods, heatwaves, hurricanes, and droughts can be seen as a sort of crowd effect, since they emerge from the collective behavior of “packets” of water and air in the form of oceans, clouds, winds and air moisture. And if we add to this the collective actions of humans – in particular, the environmental changes caused by human activity – we conjure up the controversial emergent phenomenon known as “global warming”.

At the heart of most real-world examples of Complexity, is the situation in which a collection of objects are competing for some kind of limited resource – for example, food, space, energy, power, or wealth. In such situations, the emergence of a crowd can have very important practical consequences. For example, in a financial market, or the housing market, the spontaneous formation of a crowd of people who wish to sell – and hence are effectively competing for buyers – can lead to a market crash in which the price falls dramatically in a short time. A related crowd phenomenon occurs among commuters who are competing for space on a particular road at the same time. This leads to a traffic jam, which is the traffic equivalent of a market crash. Other examples include Internet overloads and power blackouts, in which subscribers simultaneously decide to access and hence exhaust the available resources of a particular computer system or power network. Even wars and terrorism can be viewed as the collective, violent actions of different groups of people who are fighting for control of the same resources, e.g. land or political power.

The Holy Grail of Complexity Science is to understand, predict and control such emergent phenomena – in particular, potentially catastrophic crowd-like effects such as market crashes, traffic jams, epidemics, illnesses such as cancer, human conflicts, and environmental change. Are they predictable in any way, or do they just appear out of nowhere without warning? Can they be controlled, manipulated or even avoided?

What is remarkable about such emergent phenomena, is that they can arise in the absence of any central controller or coordinator. Just think about the level of coordination and communication which some central controller would actually require in order to be able to recreate a particular traffic jam. In other words, imagine the number of cell-phone calls he would have to make to ensure that all the drivers were on the same road at the same time, and in one particular pattern. It simply couldn’t be done in a reliable way. This represents a universal feature of Complex Systems: emergent phenomena can arise without the need for an “invisible hand”. Instead, the collection of objects is able to self-organize itself in such a way that the phenomenon appears all by itself – as if by magic.

The sheer power and momentum of these emergent phenomena can also be quite remarkable. We all know how easy it is to be swept up in the ebbs and flows of mob mentality – whether intentionally or unintentionally. Recent decades such as the 1970s delivered cultural tsunamis in terms of fashions and hairstyles: just think flared trousers and platform shoes. In the 1990s, we had the infamous dot-com boom with company employees agreeing to be paid in stock options rather than hard cash – only to find themselves penniless when the bubble burst around April 2000. And who hasn’t had the experience of wandering along a busy street in the middle of a crowd of people, only to find yourself separated from your companions and going in a direction you don’t actually want to go? We each seem to have an innate urge to join in with a crowd – but it may not be the best decision from our individual perspective. Just think of selling or buying a house or car. You will get a far better price if you sell when everybody else is buying, and vice versa.

It is not just collections of people that show emergent phenomena. The animal, insect and fish kingdoms are awash with
examples of self-organization: from ant-trails and wasp swarms through to bird flocks and fish schools. In fact, biology is sitting on a treasure-chest of such collective phenomena – from the immune system’s collective response to invading viruses through to intercellular communication and signalling which drives many important biological processes. The fact that all these effects represent emergent phenomena explains why so many different disciplines are getting interested in Complexity.

Closer to everyone’s personal concerns – and indeed, worries – is the area of human health and medicine. This is a prime example of Complexity in action. Our immune system consists of a collection of defense mechanisms for dealing with invading viruses. However just like the traffic, the stock market and the Internet, the system can go wrong all by itself – for example, when the collective response of the immune system ends up attacking healthy tissue. Hence understanding the extent to which we can predict, manage and even control a Complex System has particular importance from the perspective of human health. Indeed it may even lead to new forms of treatment whereby the collective responses of the body are harnessed to deal with a specific problem in a particular organ, rather than relying on one particular targeted therapy. A cancer tumor is a particularly horrific example of a crowd effect gone wrong. Instead of staying in check, cells begin to multiply uncontrollably – and just as with other Complex System phenomena such as traffic jams, it becomes very hard to know what to do to reduce the size of the tumor without causing some even more damaging secondary effects. For example, any treatment which involves damaging the tumor may indirectly lead to the survival of the fittest, most malignant cells.

Interest in Complexity is not confined to natural objects such as people, animals or cells. The ability of a collection of objects to produce emergent phenomena without the need for some central controller, has attracted the attention of researchers at NASA. In particular, Kagan Tumer and David Wolpert have been leading a research team at Ames Research Laboratory in Mountain View, California which is looking at emergent phenomena in collections of machines. The machines in question could be robots, satellites, or even micro-spacecraft. For example, NASA are investigating the
possibility that a collection of relatively simple robots can be used to explore the surface of a planet in a fast and efficient manner – as opposed to using one large and far more complicated machine. They have a good reason for doing this. If one robot in this collection were to malfunction, there would still be plenty more available. By contrast, a single malfunction in the large machine could lead to the immediate termination of a very costly mission. This also explains NASA’s interest in exploring the properties of collections of simple satellites, as opposed to one large sophisticated one – and also collections of micro-spacecraft as opposed to one much larger one.

But there is another, far more intriguing reason that NASA is interested in such research. Most NASA missions are likely to involve sending machines to distant planets – and it is hard to maintain reliable communication channels over such distances. It would therefore be wonderful if NASA engineers could just sit back, relax and let the machines on the planet sort it out for themselves. This would of course land the machines with the same difficulty as we have when trying to arrange a lunch-date by phone with a group of friends. Judging from what typically happens with the lunch-date problem, you might think that one of the machines would simply end up acting as the local coordinator, checking oneby-one the position and availability of each machine and then coordinating their actions. This sounds like it should work fine – however, the collection of machines would then be reduced to having the same vulnerability as a single sophisticated machine. If the local coordinator malfunctions, the mission is once again over. Instead, the “killer application” aspect of such a collection of machines, and hence the interest in such Complex Systems within NASA, is that it is not necessary for the machines to have local coordination in order for them to do a good job. It turns out that a suitably chosen collection of such objects can work
better
as a group if they are not being coordinated by some single controller, but are instead competing for some limited resource – which is actually NASA’s case, since there will typically be relatively few loose rocks available for picking up within a given area of a planet’s surface.