Our guest spot for this unit is Professor Luis Bettencourt.

Luis is a resident professor at the Santa Fe Institute. His background

is in theoretical physics, but like many of the Santa Fe Institute faculty, he’s worked

in a lot of different fields, ranging from computational neuroscience

to epidemiology. Much of his current work is focused

on studying cities as complex systems. So, welcome Luis.

Thank you. So Luis, in this course we’ve covered

metabolic scaling. We’ve talked to Geoffrey West.

And now we’re looking at how complex systems science is being applied to

studying cities. But let me ask -- why do you think it’s useful

to view cities as complex systems? Well I think

most people that start even just a little bit about cities know that

they are complex systems in this way we’ve just described, in that

they obviously look like they vary a lot. They’re made of many different

pieces. They look all very different. There are poor parts of the city, rich parts

of the city, parts of the city that are dedicated to some kinds of businesses

or other. So there’s a lot of diversity. However,

in whichever way you measure them, that makes the city look

very diverse. They also share with some other complex systems

the fact that cities by and large are open ended, so we’ve have cities

certainly for now over ten thousand years, and there’s something

recognizable about cities old and new, but they also have evolved obviously in terms

of technology, in terms of what it is that people do, and so

because of all this, I think cities are not just

something you can study from just their economic consequences

their social consequences, the way they are, they live in space and occupy

space, or structure space, so all of these things need to come together

social life, economic and material life, energy,

but also the way the city exists in space and in time, because cities

increasingly now with large suburbanized cities we cities that exist also

in time, where people come together during the day and then disappear again at night

people go back home. They really have the hallmark of something that

evolves, not in terms of biological evolution,

but in terms of things that we still struggle in the social sciences to define

what evolution is, and also, perhaps most importantly,

creates a new diversity, new ideas, new

spatial forms that reflect what we can do as a species when we

come together in large numbers in terms of our social life, and so

the reason I like cities is that all of these things come together into the same problem. Okay.

Well, it’s interesting that you use the term evolution. We’ve talked about evolution

in a lot of different contexts in this course, and

you say cities evolved maybe not in a biological sense, but in some sense

and people in the past have certainly

characterized cities as being sort of like organisms

so I guess one question is, are cities alive or

does it even make sense to talk about it in those terms?

They are not like an organism, but they evolve, where an organism per se does not evolve.

Other people, for example, compare cities to ecosystems famously

Jane Jacobs did this in The Economy of Cities. In the sense an ecosystem

has organisms and species that

reflect certain properties. Functional

properties, perhaps, functions that those creatures have in their niche

but it’s the interaction of all these different functions that creates the ecosystem

and in a city it’s a little like that, though with all the things they need

to change changed, in that the city only really exists as you have

division of labor between people, division of functions, and their integration back together

so that out of scale, this increasing diversity

and the productivity increases that division of labor and coordination

of labor and colocation permits. You can have a system that evolves

in that sense that at least superficially is like biological evolution that create new

things. So in that sense, the city is probably

in terms of human systems, the ultimate complex system, but it is a system

that is very interesting, when we compare different complex systems, in that

there’s a certain amount of

the creation of new things, and the selection of those

things, both of which occur also of course in biology, they become

endogenous, they become part of the city itself, that’s why the city evolves as a whole

just like, in sense, an ecosystem, although by very different means.

Whereas an organism only has a very limited way of creating new

information, which of course through reproduction, which is mostly random,

and then the selection is known at the level of the environment. Okay.

Let me ask a slightly different question. We looked a lot

at scaling, different scaling laws and

particularly in the context of metabolic scaling so far, and I know you’ve done

a lot of work on scaling laws in cities, so let me ask first, what do you

see as the main new thing that you

discovered that hasn’t been known before? What we

know now is essentially, in a nutshell, that cities

are basically social reactors. They’re sort of open-ended places where a lot of

people can come together and interact with each other and when things

go well, and necessary organizations

and technologies come together, there are places where we can do lots of good things

together, in terms of having new ideas, producing things economically

more productive, and so forth, but of course we can also do a lot of bad things

so what I mean by that is that there are places where a lot of people can

interact, react with each other, and create things that we can only

create socially, there’s sort of a famous quote by

E. O. Wilson, and he’s very interested in ants, and he calls a set of ants

an ant colony a superorganism in the sense that even

in genetic terms, they of course are very related, and that’s part of the explanation for their

sociality, but he says that an ant alone is a disappointment

you only find what ants can do when you understand how

the colony works. But to a much larger extent

you could say that a human alone is a little bit of a disappointment, that it’s when

you actually find ways of integrating a lot of what humans can do

together that you understand

what we can do in terms of our creativity and our originality and

how we deal also with resources, and that’s what cities are, cities are

technologies to do just that. When you look at organisms

versus cities, you find this interesting distinction that’s subtle but it’s

important, which is that by and large the way we understand organisms

to some extent also river networks, which may be something that you covered in your lectures, I’m not sure

is that these are systems that exist in some sense to preserve

information or certainly organisms, information

is already there in the genome, but then they want to do that using

the least amount of energy possible, and they form these networks that reach all

parts of the organism that allow you to do that. In a city

as I just described, it’s a little different, because in some sense you want to

enable this reactivity of people and organizations, so you want

to have people move around -- your cells in your body don’t move around and conspire

to do new things. In fact when they try to do that, you kill them.

You have many mechanisms to kill them. You don’t want anything inside the organism

Yes, exactly. So what happens in cities is the opposite in that you need

to promote all of these interactions, it’s out of the recombination that

these heterogeneous elements that have different information, that new ideas

new organizational forms come about. In order for a city to promote

that, it still needs to occupy, in this case, cities live in space, so

you need to have networks that allow you to cover all of space, and that’s the sense in which these networks

bear some resemblance to those in biology for an organism

that also need to irrigate all parts of space, but in these networks

what you have is by and large information, people

interacting with each other and creating new things, and that form of transport

actually takes more energy, so it’s less efficient than the organism, but

also creates an output that is

greater, so this is what we call superlinear, so you find essentially there’s

superlinear output in terms of products or social reactions

like ideas, innovation, wealth, and so forth, but you also

find a superlinear amount of energy, not so much in the network, which

actually tends to be spatially smaller, just like in an organism a little bit,

but because in that smaller network you need to move more things around

faster, and that ends up creating more dissipation. So when you say

superlinear, you mean when you look at

some functionality with respect to the size of the

city, right? Yes, so in some sense what you find is that

superlinear just means that if you measure some property of the city,

let’s say GDP or it’s number of crimes,

and if you look at cities that you think are comparable,

usually inside the same nation, but across different sizes, so you have small

cities, medium cities, large cities, and you ask how that property varies

with the number of people in the city, with the population, you find that

typically social and economic things increase faster than

population, so its per capita rates go up, and this is why

large cities are more expensive, but also people earn more money, often they are

more dangerous, and so forth. This is basically this characteristic

which is essentially an acceleration on the rate of these social and economic

interactions, this reactivity.

So the reactor is basically getting a little hotter, in some sense

much like a star. I like to use sometimes the analogy of a star. It’s

a lot like a star, in that we have more mass, the reactions in the center

are faster, and the star is hotter and everything is faster.

In that respect, it’s a lot like a star, although quantitatively it’s a little different in that the physics

and the social dynamics are different.

What do you get out of predicting all of this? What kinds of applications are there of all this measurement?

So the first thing is just to understand what kind of

complex system a city is, because

well, as you know, and probably as people watching this know,

at least since Plato and Aristotle, people have been debating what kind of system a city is

and how should we deal with it, in terms of policy or development infrastructure,

etc. Should we want more cities, less cities, what kind of cities, what can

actually someone like the mayor do to make the city better? If you think of a city

primarily as a social network that lives in space and time,

and if you think then of the infrastructure as a way to

enable its socioeconomic activity, then you

understand that there’s a value to infrastructure, for example, in terms of

allowing a more efficient use of people’s time or

costs, in terms of doing what people do well,

which is their socioeconomic activity, so there's

a sense in which one can start in some sense putting a value to what

infrastructure actually does.

You see this a lot

in the history of urban planning, the last sixty years, you have cities that tend to be

very dense, and that’s not good,

because people cannot move around and meet people that they might actually want to and get organized

and so those cities benefit essentially from having better infrastructure

that allows movement not only of people, but also of things

and information faster, but you also have then

in the United States more recently, cities that become too large in terms of infrastructure

where the cost of just moving across, at least in time, if not

in other forms, is quite high, so

those cities are just too diaphanous, and they

cease to be easy places of human interaction, and so the theory

that I’ve developed, basically tells you what are the

right compromises in terms of

social interactivity and the characteristics

of infrastructure, and the structuring of space. The other answer

is that this also allows us to put quantitative expectations as to what a city

with certain characteristics like size and land area might

do in terms of its wealth, or in terms of its innovation, in terms of its

crime

Some of this is contextual, because it depends a little bit on the history of that urban system

but some of it depends on these general characteristics of cities, so that allows us to say a little bit

how well a city’s doing relative to its peers, if you wish.

One last question. So what’s the most exciting thing you’re working on

now? So

that’s a hard question, because you prefaced this by saying I work in too many

things. I’m excited about them all. I didn’t say that.

So what’s missing when you look at cities and human society more generally

is that even though we now have a

reasonable understanding of the properties of cities versus size

we have a pretty poor understanding

of the properties of cities in time.

That goes to may others. You mean how they evolve?

How they evolve over time? How they evolve

over time, and this is of course all entangled with

deep, deep questions like how is the US

or any other country going to grow economically in this

coming year, or is our society going to be more peaceful

or less peaceful, because basically what you find is that the properties

of cities, which vary across sizes, are then integrated in terms of an urban system

of cities of many sizes, and there are these flows of people and information

that occur between all of these places, and so

it’s important that when you have an idea in New York,

this idea is applied, for example, is applied

everywhere that it can be useful. So when you develop computing in Silicon Valley

the wealth of creating that computing is not in that idea per se,

but it’s in the application of that idea everywhere, including

to farming and agriculture, and mining and primary activities.

So that cascading of ideas and organizations all the way

through a society is what most people think is

going on when you think about economic progress, economic growth, but also

growth of other things, such as improvements in public health,

and so forth. This is important in developed countries, but

it’s especially important when you go to developing countries, where

many of the properties of cities and human societies that we haven’t developed, countries are not

there yet. And so there’s a question, for example, in terms of infrastructure and services

So there’s a question about what is the path which you can create these

properties of human settlements

in order to enable a self-sustained way of creating

again, social interactions, ideas, organizations, that allow this system

to go on and evolve, and grow, at least economically.

So that’s a very important question which has to do with economic growth and

other things, and the other question is the question of creating actually a statistical theory

of what cities are, both for organisms and for cities

and for most other complex systems, for the brain and for many other

systems, we have a very poor idea of

their statistical character. For example, all of these scaling laws only give you an idea of what

how a system behaves on the average, given its size

or any other characteristics. But when you look at a particular

creature, a particular city, it’s never quite that number

it’s something around that. So the question is what creates these deviations

all these quantities behave a lot like

fluctuations in the stock market, at least in terms of their correlations

multiple different processes. So they’re hinting to

what happens in a lot of social organizations, not just

in cities, as being the result of processes there are long chains

of causation there, more like ants, so a city gets rich because

it attracted people, it developed some infrastructure, that

allowed them to do what they do, and discovered some technology

and had finance, and more people came

it’s a long history of things that need to happen, they’re all necessary

in order to get to a certain place. There are interesting

theories, they’re all qualitative, in

the social sciences, that try to express this, and they are

good candidates for explaining some patterns like crime, or wealth

things like cumulative advantage and disadvantage, and so

I think I’m close to having a way of describing that statistically and therefore

end up with something that can actually make predications where we might expect

we might know how wrong we may be and why. Well that sounds really

interesting. We will look forward to hearing more about that later.

Okay, well, thank you so much. You’re welcome, it’s been a pleasure.