 Hello everyone and welcome to the Circular Metabolism Podcast, the bi-weekly meeting where we have in-depth discussions with researchers, policy makers and practitioners to better understand the metabolism of our cities, or in other words, their resource use and pollution emissions, and how to reduce them in a systemic, socially just and context-specific way. I'm your host, Aristeed, from Metabolism of Cities, and on today's episode we will talk about the future of cities. You will all have probably heard that most of global population now lives in cities, and that the rate is going to increase very rapidly in the near future. But what if I told you that everyone on this planet will live in cities by 2100? Imagine all the implications this could have, and all the subsequent questions that will follow. For instance, where are those cities are going to be situated? How big these cities will be? Will they continue expanding on old cities? Will they be new cities? Well, to talk about this fascinating topic I have Michael Batty, which is an urban planner, geographer, spatial data scientist, and a professor at the Bartlett School of University College London. He has been the director and now chairman of the CASA, so Center for Advanced Spatial Analysis, and his research was and still is on computer models focused on city systems. He has been the author of numerous books, which I've read over the years, such as Fractal Cities, Cities and Complexity, the New Science of Cities, and the latest one, Inventing Future Cities. Just before kicking off this episode, I'd like to make a small request. So if you like this podcast and this episode, just share it around with your friends and colleagues. I think many of you will find this interesting. And if you're watching this on YouTube, please make sure to subscribe to our channel and support this podcast by sharing it along with all that being said. Hi, Michael. And could you please briefly introduce yourself and your research? Yes, thank you, Ariste. As you say, I've worked for a long time on what I loosely call the science of cities, which is to try and develop more formal approaches, but nevertheless, still practical formal approaches to thinking about how we can think about, understand and forecast, predict in some sense, the future of cities in this context. And a lot of the focus that I've been involved in draws ideas from the physical side. The whole social dimension of cities is missing in some sense. It's just that the emphasis in the kind of work that I'll be talking about is very much to do with how we can formalize in some sense, at least to begin to think and understand what goes on in cities to formalize the different relationships, the different processes. And for example, one of the kind of key features of this podcast series is the idea of circular metabolism. Of course, in this context, metabolism means the sort of processes that go on in a particular system or ecosystem in that sense. The other processes whereby you get growth and change in general, and the functioning of the system in this particular context, and inevitably metabolism is very much involved with the way energy is actually transformed and used within any system. Now, we do take that specific kind of focus to some extent in developing a science of cities. The other key issue, which again relates quite strongly to this, is that metabolism is really related very much to the kind of complexity of how these processes maintain themselves, how the processes which are involved in cities lead to new growth, regeneration and so on. And metabolism in this particular context is very much related to another theme that has emerged over the last probably 30 or 40 years, that's maybe more over 30 years, which is complexity theory in that sense. So the idea that systems that we deal with, cities in particular in this instance, are complex, they're increasingly complex as we invent new things in this particular way. And therefore we need quite powerful techniques and models to actually deal with this in that sense. But in a nutshell is really the position from which I actually come to talk about these ideas this afternoon. Yeah, I was wondering if you could perhaps take us back, because of course, urban, you were trained as an urban planner, and of course, back in the day, computer science and urban planning were not necessarily in the same room. And of course, CASA does this, but I was wondering, so I can imagine that geography back in the day used GIS, so some sorts of computers. But when did this use of computer models in urban planning or city making started to exist? Well, actually, the origins are a long time ago. In fact, almost as soon as computers got invented, to give some context about this, generally speaking, I and most of the people who've looked at the history of computing, and the little computer was invented in the middle of the 20th century, about 70 or 80 years ago, essentially, during World War II, World War II gave a massive boost to the whole notion of digital thinking in this sense. But of course, the computers were invented in those days, but very much scientific instruments in this particular context to enable people to do scientific computation, basically. And unsurprisingly, at the time, there were a lot of defense related applications of computing, the initial computers that were built in Britain and Germany and the United States were essentially financed because of the war effort. But as soon as computers are invented, some of the pioneers such as Alan Turing and John von Neumann, these were people who thought about computation by an abstract point of view. They began to make the point that the computer was a universal machine. And this was a very hard thing to take because it really meant that if it was a universal machine, it could only begin to operate or simulate a real system if everything could be reduced ultimately to zeros and ones, the binary code. And this was a very hard thing to take when all one had really was numerical calculations at that particular point in time. The idea that computers could actually produce pictures in the middle of the 20th century was very unusual, although there were glimpses of these possibilities. The idea of computers producing music and song was really very much on the back burner. But as soon as computers began to be developed, it was very clear that one could store enormous quantities of data. And so there was an immediate commercial application, IBM, for example, which had manufactured calculators in the 1920s and 30s, basically picked up just after the war on the development of these new mainframe machines, basically, and they became the biggest computer company in the world by the mid 1950s. And in the mid 1950s, there were a lot of business applications. So you had scientific applications and you had transactions processing for business in this particular context. And it was pretty natural that people thought about other sorts of scientific systems, such as economic forecasting and so on, in this particular context. Now, of course, in the United States during the 1950s, the other thing that was happening was a very massive move towards urban sprawl and the automobile and so on in this particular context. The interstate highway system was developed in the mid 1950s. And as part of that, the kind of complexity of all these things was picked up by people who began to think about building computer models. So the first wave of computer models in cities was in transportation in the 1950s. The Chicago area of transportation study commission cats basically in the mid 50s, basically were built a variety of different models, population, employment, obviously transportation, which was the focus in this sense, in that sense. And really by the 1960s, we had land use transportation models. And these models were very different from what we see today in the sense that you really there was no visualization if you wanted to draw a map of the output basically then had to be done by hand. A lot of the processing was was done very laboriously. So for example, punch card and punch tape took the data input and the programs and so on. And the mainframe machines actually, you know, these were fed to the main computers and then a day or so later, you would get the output. So it's a very, very different world in that sense. If you looked at the process of building a computer model back in the 1950s and 60s, then it would be very arcane compared to today you would be surrounded by punch tape and punch cards. The whole process would be based on trying to get the model to work without actually demonstrating any errors. In other words, if the model kept producing errors, then, you know, and it's only you were only able to get one run a day over 24 hours, it would take ages to get models to work. So there was a very different emphasis from today, whereas if we get errors when we program, basically, we simply dismiss them and move on and move on and find a way around them in that sense, there was really much less of that. And so consequently, the focus was very much on building the models right the first time. But the biggest single problem was that these models were data hungry, the time was tiny compared to what sort of computer time is now used with these models. But basically, the experience was very salutary, that in other words, there was a rise and fall of these sorts of models in the 1960s. And by the 1970s, lots of other things have come onto the agenda. So in the 1970s, the land use transport models really went onto the back burner. And what really emerged were other ways of using computers within cities. So geographic information systems, GIS, which you mentioned, which is really to do with the representation of things, that really began to take off. And that took off largely because graphics came onto the into the picture, mainframe machines got scaled down to personal computers. In personal computers, you always had a part of the memory was graphics in that sense. So gaming really began in a big way. And consequently, GIS really came out of the idea of automated cartography, and therefore, and then spatial analysis in that sense. So then there was that. And then, of course, in the 90s, we had interactive computing, the web and so on, and web pages, and the dissemination of ideas about cities, you know, public participation and so on. So you can think of every aspect of computing as somehow influencing what was going on in cities. Now, this does not mean to say that what was going on in this context was necessarily the right or the best thing in some sense. Many of the models were based on really rather poor theory. We've not really still got, we've still not really got good theory when it comes to understanding cities in that sense. That's, that's one of the purposes really of, you know, the sort of work I've been doing really, and then many of my colleagues too, to think about a science of cities so that we can actually take forward a bit more of a good theory. And of course, in a context where there are many different perspectives on cities that we can't ignore all of these different perspectives, we have to integrate them in some sense. And of course, the idea of circular metabolism is to some extent thinking about integration as such in that sense. So all of these things were really sort of elements that I've worked on over my career really in that context, starting with, you know, the big kind of lumbering dinosaur type models from the 1960s. They're not big, they're tiny compared to our models now, but they were big in terms of the amount of effort and scale that's had to be put into them. So that's really from where we came in this sense. Yeah, I mean, it's fascinating, especially as you say that sometimes we just deal with it with an issue or a challenge just because we have that tool, you know, when you have a hammer or problems or nails somehow. And I guess the city has suffered as well with subsequent theories that kind of sometimes say that the city is a machine or sometimes says that the city is an organism and stuff like that. And we kind of see that the integration of challenges come also based on our definition of a city, of our theory of a city and all of that. And I can imagine that very rapidly you focused on linking cities with complexity as well, which I imagine also emerged in similar times, I mean, complexity theory and and computers also were hand in hand in the 60s or 50s. Did you directly make the connection between complexity and cities? Well, not particularly me, but the what really happened in the mid 20th century that, you know, I read the 20th century in terms of the social and physical sciences as one where the physical sciences made great strides in the 19th century and then with the development of quantum and relativity in the early 20th century. And many people in the social sciences, particularly in the United States, began to think that the social sciences should also aspire to a scientific paradigm in that sense. And so consequently, economics was the one that began to change first and became essentially in the late 1930s, 20th century, fairly mathematically many senses, very mathematical subject in some senses still today. And so you got this transition towards a more harder social science. And that led in the 1950s, really, to what was loosely called the systems approach. In other words, if you had a problem, a system, which was not very well structured, not very well disciplined, the city would be a very good example, right? On like a physical system of atoms and molecules and things of that sort, which were relatively simple, but can be treated in a very complex way. Cities were a much more sort of vague sort of animals, basically. There's still systems, there was still structure. And a lot of areas like this were subject to the so-called systems approach. You define a system, you define the system's environment. That's the input and output between the system and the rest of the world. You look at the subsystems and there's a hierarchy of the subsystems and so on. This was the terminology of the systems approach. And of course, you can, having heard what I've just said, you can apply that to a whole range of different systems, from the economy to the city to a psychological system, you know, to physiology and so on. And so the systems approach was that it really downplayed the idea of dynamics or change, the change from one system to another, which was really all important in terms of metabolism, for example, changes absolutely essential to noting the way different processes take place within any particular system in that sense. So consequently, the first change in the system, the development of the system approach was the movement to begin to think about dynamics, basically. And there were lots of ideas in mathematics going on in the 1970s and the 80s to some extent, related to dynamics, related to change over time. Catastrophe theory got invented in the 60s. Chaos theory in the 70s. And this does not, the chaos theory or catastrophe does not mean randomness. It means sort of a way of thinking about how systems develop really with a degree of unpredictability, but not necessarily a random unpredictability. So in some senses, all of this mathematical machinery and physical machinery was put together. And a number of people, mainly in physics and and to some extent in biology, but physics and economics really, got together in the United States and they developed the idea of complexity theory. There are many, many contributions to complexity theory in this particular context, but mainly the development came at the Santa Fe Institute largely that became a focus in the United States for the development of complexity theory in terms of economic systems and then social systems, but also things like city systems and so on, as well as various sorts of physical systems from the physical sciences, basically. So complexity theory became a fairly comprehensive view about how you treated any kind of system, really, whether it be a physical system or a social system. And of course, in planning, or rather in cities, I should say, cities became very good candidates, if you like, for the study of these complex processes in that sense. It's a very good analogs, really, for the systems, for the complexity approach, generally, all the features about a complex system, how it evolves from the bottom up rather than top down. So there's a degree of uncoordination at the bottom up, but it leads to pattern and order as we develop the system. A very good example would be to look at the any kind of remotely sensed image of a large city, you know, the night lights photographs from that NASA, for example, and you would see that nobody planned it at that level. It's like an evolving biological system. And the analogy that cities were more like biologies, they were more like organisms, really, than they were like machines, really, came onto the agenda. You can loosely think of the systems approach as thinking of the world as being made up of machines, manufactured from the top down, whereas cities are really built from the bottom up, they evolve. Rather than being manufactured, they actually evolve in this particular context. So that's a very good way, I think, of noting the difference and the development of these things over the last 50 years in that way. Yeah. And especially, I think we had also, Joffrey Wester, lately on this podcast, talking about scaling laws and also how this bottom up approach of cities. And I think somewhere in this Inventing Future Cities book, you mentioned that cities are getting complex at a faster rate than our understanding is able to keep up with. And I think that that is quite not frustrating, but sometimes scary for us researchers, you know, how we are always chasing after this ever evolving element. And I'm wondering, could you perhaps elaborate a bit on this and how do we take a stand or how do we act on this? I think, you know, my own view is that we really need to make a distinction between the pre-modern world and the modern world. And the modern world really dates from, at least in my perception, it dates from really the 17th and 18th centuries, where there was a lot of development of first ideas of physical sciences, the Newtonian Revolution and so on. And then hard on the heels of that came the Industrial Revolution. And I suppose, in a sense, the Industrial Revolution marks the great divide, really, in that sense, that everything prior to the Industrial Revolution really was very, very stable in the long historical context. So, for example, death rates, for example, remain roughly the same for men and women, really for probably 10,000 years since the era of the Cultural Revolution, probably 100,000 years, basically, people lived to about 25 to 30, and that was it. And that was still very much the case, you know, in the Middle Ages and the Renaissance and so on. And then things began to improve in some sense that, arguably, it's not at all clear why things improved a bit, but there was a sort of stability or there was a sense in which we moved above the best subsistence level. And this led to the Industrial Revolution. And in many senses, that led to the invention of technologies. First of all, there was mechanics, and then there was electrics and electronics. And then there was the computer. And now we've just got a succession of technological revolutions, most of which really relate to the original mechanical and industrial and electrical revolutions as encapsulated in the computer, basically. So everything these days, in terms of a revolution, is digital in some sense. So in the present time, well, various people have talked about these waves of industrial revolution. So generally speaking, the third industrial revolution after mechanical and electrical, which led to digital, basically, is the world that to some extent we operate in. But everything from now on is part of that. Third and now fourth industrial revolution, which is biomedicine and digital and so on, the way computers and computation are infecting everything, basically, in that sense, organizations and so on. So in other words, what we've actually got is a series of ever faster revolutions with the technology. And the technologies don't knock themselves out. To some extent, some technologies become obsolete, but most technologies are absorbed into the growing complexity. And that's what leads this growing complexity, the fact that new things are being invented all the time. The complexity of the world is such that old things are not really thrown away. At least some elements maybe, but a lot of what's in old technologies is subsumed in you in this context. So we have wave after wave of technology and it becomes increasingly difficult. Basically, if you're changing, the technology is changing so rapidly, then even people who are trained in it and become experts become, it's not exactly obsolete, but they're not really aware of the new technologies, even though they may be working in the field. They don't have enough time to actually be able to absorb these new technologies. And that's particularly true in terms of computation. You can actually see that writ large across the way they use computers and technologies in science and social science at the present time. So this is what I mean by the increasing complexity of things. The other feature, of course, is that arguably the world is getting wealthier. It may not appear that way in some context, but generally speaking, falling death rates and modern meds and globalization generally in some aspects has led to increasing opportunity. And if you give people more opportunities, then they invent complex things, basically, or more complex things. And in that sense, this is what I mean by things getting more complex in this particular context. And in terms of understanding cities, many of these new technologies that are being invented are being embedded in some sense into cities. I mean, if you think of the simplest distinction between the way we looked at cities in the long term, what I like to call the, not the short term city, but the long term city, basically, in this sense, the short term city is the 24 hour city, basically, the smart city in that context. Now that's all emerged in the last 20, 30 years as computers to the buildings, basically. And then we're using computers to model cities, but we're also using computers to model cities that are actually composed of computers in that sense. So there's this mixing that's gone on, which is quite problematic in terms of getting a handle on it. And of course, the revolutions that we're about to or already beginning to go through, are where computers are being embedded into not into the buildings around us. That's certainly taking place in the transport systems, embedding them into ourselves, really, into the into the natural environment and then into the human environment, basically. Now, it's very hard to know where all this will take us in some sense, but there is no doubt that it's increasing the level of complexity everywhere in that sense. So that's one of the reasons why I say cities are increasing in complexity. It's not that all the things aren't. If you could generalize that argument and talk about society, society is massively increasing in complexity in this context. New organizations being added. We spoke before the podcast about the use of modern software, basically, all the time new software is being invented and people have to get with it to be able to use it in some sense. And there's only a limited amount of time in the day to actually do that. So this is adding to the kind of complexity of facing this mind-boggling complexity in this context. And it's difficult to know if the speed of the development of new ideas continues at the rate it has been doing. Then to some extent, we could argue that we're overwhelmed. I suppose the saving grace is that new people coming to what they see out there and learning from the bottom up, so to speak, are able to handle it better than people who are already part of the system in that sense. That's always been the case, you know, young and old generations and so on. So this is, I suppose, what I mean by increasing complexity. And so that means hopefully the newer generation will better understand the complexity of the city than the older generation. Because something else I have always encountered is that many times us new researchers kind of think that we've found something or we theorized something and then we read a book and we realized that this was there already 100 years ago. So how do we orchestrate the tool or harmonize the tool? Something that, you know, there is still some in-depth knowledge that exists for years and years and centuries and still at the same time there is this complex new bottom up elements that are made of, you know, new complex cities invent every day and how do we align these two in order to comprehend this city or to better create cities, to better transform cities and stuff like that? Well, this is very difficult in the sense that a lot of knowledge is built layer by layer on top of each other and you see this particularly in computing. Nobody really understands how the internet works or we don't have any neat laid out formalized theory of how these systems work in some sense. And typically things that are grown from the bottom up in this way are quite hard to theorize about in this sense. People coming to the phenomenon, any phenomenon without really knowing much about its history are able to accept much more than people who've actually lived through that history, I think. That's a very important issue. So, for example, if you are, I mean, one of the classic examples in computing is that networks are built of in layers. In fact, computers are built of as layers. Right at the bottom layer you have the circuitry that deals with the zeros and ones and, you know, the experts and physicists who deal with all of that are really people who are, you know, experts in the physics of light and materials and all sorts of things like that. Nothing really to do with applications of computers, in a sense, that we're talking about here. And so in some senses, all of that is taken for granted. I don't know whether there would be anybody in the world who really understood how everything actually worked. I suspect that, I suspect one or two people, you know, a handful of people do understand, I think, the way computers work in terms of their programming, right down to the basic guts of it. But even down at that level, it's now been so highly automated that it's very difficult to kind of discern tiny little. And people don't need to, because somehow when we build a world in layers, it may not be the right way of doing it, but when we build it in layers, we can accept, you know, all the layers that came before. So that's one issue in a sense. You made the point about things being rediscovered. That's a very interesting issue that are a lot of things that, you know, have been thought about in the past, basically. But also there's a lot of reinterpretation that goes on. What was important, you know, 30, 40 years ago in terms of thinking about how cities are structured is perhaps no longer important. So for example, when I was a graduate student, basically, things like transportation flow, transportation flow, the structure of cities where city centers were. And so these were all really quite important things that you could manipulate by models in a sense and make forecast for and plans were organized in that way. Today, there's much more concern for the social structure of cities, even in spatial analysis, the social structure of cities, much less concerned about some of these big questions about movement. And of course, what has happened is that cities themselves have changed. No longer do people think about moving in cities in quite the same way. There's lots of different sorts of movement patterns which are now important in the context of cities. And that's reflected in what people study. So for example, if you look at the number of people who are working with smart card data and transit systems, it's enormous, right? It's out of all proportion to the number of people actually traveling, right, in that sense. So because smart card data is easy to get from transit systems and so on and from mobile phones and so on. Lots and lots of people are gravitated towards that. It's not at all clear whether anything that we don't know already has actually been produced from all this work, basically. That's a big issue. I mean, I'm not a detailed expert on this, but there are lots and lots of issues. Back in the day, we were very always very concerned about the integrity of data, whether it was a representative sample. And it was almost a situation where you wouldn't work on a problem if you couldn't get what was a reasonably representative sample. Obviously, the data you could get would depend on what you would study in that sense. So in other words, if you wanted to study a particular phenomenon, but you couldn't get data on it, you probably meant that you didn't study it. Whereas today, because we can get loads and loads of this data, this massive amount of data, but it can be largely unrepresentative. I mean, classically, mobile phone data, social media data is unrepresentative in terms of the age groups, for example, of people who are associated with using those technologies generating that data. So from that point of view, there is less concern about these issues. It's not the form we were any more right than the latter. It wasn't what was done in the past is any better than what was done now. It was a difference of interpretation, a difference of what turned people on in this particular context. So for example, and we still work with these tools, that generally speaking, we build statistical models. Now, you can talk to people today and you can ask them what they're doing and say, oh, we're working on machine learning. What they mean is they're working on regression analysis, basically. I mean, regression analysis with a few fancy handles like applied to a little bit more structure in the data and so on. And in some senses, in a way, there's been a reinterpretation of what's important in some sense in terms of the very techniques themselves in this particular way. So I don't know whether there are many good discussions of this particular issue, but it is an important thing to get things into perspective in that sense. No one generational group is any writer or wronger than any other in this particular context. And so it's a very good indication of how cities are so complex that so many different perspectives on looking at them in this context. Of course, with the advent of big data and number of data sensors and all that, we love the fact that we can somehow try to dilute complexity into numbers and therefore use them to do something. We don't know what exactly, but at least they help us pass time and try to invent or to figure out some new relationships. And over there, the whole thesis of your latest book is that even if we model cities, we can really predict them. And so perhaps the easiest way to predict cities is by inventing them somehow, you say. Yeah, more or less. I think I say we can invent cities. It doesn't mean to say that we invent better cities. We can invent them and we do invent them because life goes on in that sense. But we're unable to make strong predictions for most things in cities. And it's taken a long time to realise that, I think, that partly it's due to the fact that unlike in the narrow physical sciences where we can make predictions which can be borne out in terms of the laboratory and so on, certain sorts of sciences, most sciences you can't make predictions either. But some you can and that they represent the sort of, you know, the peaks really of thinking in physics and things like that. Generally speaking, the difficulties we have over prediction are really that in a sense we need to think about the future. We need to predict but we know we can't. So we need to explore it. And some example of pushback onto this dialogue that these are tools that are not going to give us right answers. They give us answers that really dictate our thinking that can actually lead to new thinking in some sense. At least that's the theory behind it in some sense. So it's a sort of a retreat really from this very strong, positive model that one can make firm predictions in that way. I mean, one of the things that you mentioned was that you mentioned Jeff West and so on. And you mentioned the idea of reducing complexity to a whole series of basic things. I think that one of the kind of key features that has been developed by Jeff West and his colleagues such as Louis Bentoncourt and so on, they have resurrected the idea of scaling in terms of economics and also biology, allometry. The idea that Alfred Marshall talked about in the late 19th century, he was an economist at Cambridge who wrote very influential books on principles of economics. Marshall's agglomeration economy is basically really related to the fact that as things got bigger, there were qualitative changes in what we could do. And really in a sense, Jeff West's notion that the power law basically is a sort of signature, if you like, of the relationship between the size of objects and how we understand them in some senses is quite an important extension really. The idea that as a city gets bigger, you get qualitative changes. I mean, there's some very good examples of that. In big cities, you have subway systems, which can't really be afforded in small cities basically. And if you look at that over the whole spectrum from the very biggest cities, which are really mega cities like in the Pearl River Delta, perhaps the Netherlands, and southern Britain basically, you get these great agglomeration, the Northeastern Seaboard and so on. Very different style of life from the small town village, even Hamlet basically in a sense. So massive qualitative change. And that's reflected in this notion of the power law basically, which is just a measurement of size in a sense. And so it seems to me that looking for signatures like that is quite interesting because it unifies, it gives a degree of uniformity, or at least it gives a degree of relationship between the sizes of different things so that you can relate them directly and look at how the properties of them change in that sense. So power laws are good signatures, I think, of complex systems. Things like the fractal dimension, the dimension which really relates to the amount of structure you have in a pattern at different spatial scales. The fractal dimension is also another signature. It's a space filling signature for how space is actually filled basically. So there are one or two key issues of this kind that are being drawn out in terms of what we would expect to see in all cities, I suppose in a sense. The quest, I suppose, in a science of cities, like a science of everything, is to actually produce theories and models that are generalizable to a very large class of objects basically. And one of the problems I think in cities is that you can see that certain things are generalizable between cities. But you can see that lots of things are not generalizable between cities. So it's this tension between knowing of a city, knowing of certain stable characteristics in a city and contrasting them with the more volatile characteristics or biggest differences in a sense and beginning to construct a theory around that in some sense. But then, I mean, you go straight to the most difficult point which is then what is a city, what is the functions of a city? Because you mentioned or a lot of people also mentioned that cities are the place where economies of scales exist. And therefore, we pull together scales and money and agricultural surplus in order to make this happen. In other places, we say that a city should have or I think you said that cities is where we have a critical mass of people in order to have city-like functions. But then, you know, what are these functions as well? And that's the, I think the most intriguing question which we always boil down urban science or whatever to, okay, what is a city after all? Yeah, I mean, of course, in some senses, this is a question which is also quite a changeable volatile question because there's no doubt that cities through the ages have changed in terms of, I mean, they're much bigger now than they always were. At one level, you made the point that I make in the book that eventually we'll all be living in cities by the end of this century. But the clear issue is, we'll all be living in different sizes of cities. The size distribution is probably not going to alter very much that we're not all going to live in the biggest cities. Most of us are going to be living in smaller cities in that sense. And in other words, the experience in these different places is very different. Although they're united by certain, you know, quite well-grounded relationships such as power laws and so on, you know, hierarchical laws and etc. There's still quite profound differences between big and small in this particular context. And that's quite difficult to reconcile in some sense because a lot of the discussion about different sorts of cities at different sizes is highly qualitative in some senses. And it isn't even particularly uniform to say for the same city that one of the things that's not been done much is to look at the inside of cities and to think about how some of these issues about scaling change inside a city in this context. And that really introduces this notion about, you know, where does the city begin, where the city end and where does everything else begin? Because if we alter the definition of what is a city, then the biggest single question then is that lots of things begin to change. Now, in London, for example, in Britain, there's quite clearly London has a much bigger impact on other cities in the UK than you would find New York or Chicago do in other cities in the United States. That may be a scale thing, right, in a sense. But generally speaking, a lot of people in Britain believe, you know, geographers and so on believe that really in Britain there's only one big city and it's London. And really, if we start taking taking London away from everything else or everything else away from London, we don't get a good explanation of London in some sense. Only can we begin to reconcile some of these changes in scale, which you see in power laws, when we begin to relax our definition of what constitutes cities in Britain in this particular context. So and globalization, of course, is changing this all the time, basically, increasing number of people who generate the wealth in these big cities don't live in them to some extent. So it's very difficult to equate the wealth with where it's produced. I mean, in some good examples of scaling, in terms of scientific papers, you'd expect the biggest cities to have probably more than proportionately the number of scientific papers. So the power of sorts in that sense. But in fact, that often depends on where papers get classified. Apparently in the old Soviet Union, basically, all the papers that were written by the physicists were spread right across Siberian places, were all collected together and catalogued in Moscow, basically. So you've got all this sort of scientific innovative stuff, which was classified, you know, in a place for where the headquarters was really in Moscow, the same in terms of retailing in Britain. So a lot of the supermarkets, for example, report the data to head offices. And if we want to look at the hierarchy of supermarkets by the amount of income they have and so on, we have to disaggregate that data. We can't take the data from head office because it's absorbed. So in other words, the space which constitutes the city is really all important and different definitions change the nature of what we're actually studying in that sense. And then arguably, we may not, cities at the end of this century will be very different from what they were at the end of, let's say, the 17th or 18th century before the Industrial Revolution. And in some senses, we may even, we probably will continue to use the word city, I think, because it is used very generically. It's not very precise, in that sense. And what we're talking about here is trying to make it precise in different ways and we get different articulations, we get different results, different ideas about a city we classify in different ways. So what do you think this future city will look like? Do you mean in terms of functioning, in terms of form, in terms of role as, you know, economic power or? Increasing differential between what people generate things and where they associate them with. In other words, there's going to be an increased differential between home and work, for example. And that, so in other words, breaking the link between traditionally, in the simplest way, the city centre and the suburbs, I suppose, in some respects, that that really relates to the nature of work and where work can take place in some sense. And that's something that, of course, the pandemic has exaggerated in many ways, although it's very much on the cards before the pandemic. There were people writing about the divorce of work from home. 150 years ago, 100 years ago, E. M. Forster wrote a little story called The Machine Stops. And Alvin Toffler, who wrote the book The Third Wave in 1969, I think, and then Future Shock, about 10 or 20 years later, he was very much talking about the divorce of work from home and people working at home in their electronic cottages. And there have been a lot of studies about this break between work and home in the 90s and so on. And it never really happened. If you look at the commentary on it, it says, OK, it's now easy to work at home, but most people don't. It's been like 3% to 5% of the workforce working from home for about 50 years, something of that sort. But of course, recently, with the development probably of information technologies again, I mean, the very meeting we're having here on this Zoom call is a classic example of something that is a non-place urban realm type thing. It's a non-place, and it's the divorce of work from home, basically, in that sense. It's very interesting. So the city will change. And again, this is another quite interesting issue that when we look at cities, we don't really, we don't necessarily see what the deeper picture is, that the physical fabric tends to disguise what's happening underneath, really, in a sense. We've got to look under the hood much more. And that's hard, especially in a global world of information technology. Actually tracking all this stuff is almost impossible, really, in some sense. So some big question marks about how we define the city and how we actually study it with respect to these new patterns of work and interaction and so on. Yeah, you mentioned that there is, I mean, the city went from something being very spatial to today, something non-spatial or a spatial, so indeed, yeah. So because you say we need new ways of looking at the city, understanding the city and all of that, you had an entire book called The New Science of Cities. I'd like just to spend some more time on this, and could you perhaps share some aspects or key features of this type of science needed? Yeah, yeah. One of the things in that book, which has never been picked up by really anybody, I don't think, and maybe that's due to me not pushing it enough, and it's also due to it not being a very current idea. The idea is that when we look at relationships between objects, so we might look at relationships between places within the city, we tend to look at the immediate effects. If you look at divide the city up into zones and you draw a graph of transportation links, draw a network of transportation links, etc. that you then begin to study the network. What we don't tend to do is basically predict the network, basically. We assume that it exists, the social networks, we might be interested in trying to explain who talks to who, but we don't really predict the actual network in this sense. A lot of our models to assume that the network exists, they don't predict it. In the book, I'm much more interested in predicting networks, and in the first half of the book or first third of the book, I actually introduced some ideas about how we can predict a network. By taking, for example, two sets of objects, we might take workplaces and we might take home places, and if we form a matrix of who works versus where people live at home, basically in that sense. That gives us an asymmetric relationship in the sense of where people work and where they live are two very different sets of places. From that data, we might be able to predict the network itself, rather than assuming the network, actually predicting it in a sense. That really is thinking of the problem as what's called a bipartite graph, as opposed to a simple graph, etc. Throughout the book, a number of models that I talk about, location models and so on, really dwell on this idea of the bipartite graph, the bipartite model, where we don't necessarily know the network, but we're able to predict it or get some handle on it by putting two sets of objects together, which you're unlike in this particular context, whereas most networks in cities are put together by looking at how many people move from A to B, where A and B are treated in the same way, in some sense. That's one of the key things in the book. It's a little bit abstract in some sense, and you'd have to dig into it and think about the graphs and so on in the book. In some senses, there's not much in that book, if anything, on scaling. Now, the book was written in, well, probably about eight or nine years ago now, there's not much on scaling. We were all very aware of scaling at the time. I mean, the Lewis Battencourt and Jeff West were really popularized, the whole new ideas about agglomeration in their PNAS paper in 2007 or 2008, I think. But there's a little bit maybe about scaling, but not much. There's a little bit about power laws, but not in the same sense. Lewis Battencourt's new book, Introduction to Urban Science, is very much about scaling. The whole book's about scaling, in fact, in this context. So it really relates to some earlier traditions. There's a fair bit in the book on space syntax in my book, that is, where space syntax is a good example of this notion of changing the focus. One of the kind of key features of the space syntax is that it does deal with bipartite graphs. It treats the nodes in the streets or the nodes in the segments as different sets of objects, and how you put them together really relates to the street network or the space syntax network. And so it relates a little bit to that. Also, there's a fair bit of stuff on space interaction. And then the last third of the book really deals with something quite different. It deals with these same tools, but looking at almost like social networks, relationships between people who are involved in a problem of design, basically, in that sense. And how you get some degree of convergence in terms of the mechanisms that were defined on these graphs. So there's a little bit about the pros and cons. There's nothing in the book either about, there's not much at all about diffusion. If I was writing a book today on a new science of cities, if you like, I'm not sure I would call it that, but because in some senses I call it the new science of cities, but really it might be relatively new, but it's not the only science of cities. In fact, I think in the first few lines of the preface, I say something like, there are many different sciences of the city. To third one, it was completely wrong. And anybody opening the book saying, yeah, this is a book about one point of view, it should be surely disabused by the fact that, although I say it a number of times, whether people take it to heart, I don't know, but it's one of many ways. And all it's saying is this is a new way of looking at it, pulling things together and looking at it. It's not the only way by any means in that sense. So there are a number of themes of that sort. So it's a rather eclectic book in some senses. As I say, if you're writing one today about the new science of cities, I think you take a much more traditional point of view. Some people have done, Mark Bartholomew in Paris has done some stuff on that. He, I think, has taken a more classical view of this. He has a Cambridge University press book on on some of this sort of stuff. The Louis Bettencourt book is much more about scaling, I think, of different contexts in a way. So I mean, that would be my reaction to your question about what goes on in the book, The Science of Cities. Probably a more classic book about the complexity stuff is the previous one I wrote on on City of Complexity, because that really is very much about what complexity is. It's about bottom up. It's about path dependence. It's about city size and a number of other things. And more importantly, the Cities of Complexity book definitely has very specific models of cities of complex systems. First of all, there are cellular automata models, which are classically a dynamic in contrast to those models I talked about in an earlier age, the Land Use Trust. They're classically dynamic cellular automata models. They're highly formalized and so on. They're very easy to apply and to articulate. And then also that merges into agent-based models. So there's quite a bit in the book on agent-based models, really. Agent-based models being agents who sit on cells, basically, and move around in a slightly more elaborate version of the cellular automata model. Yeah, I was I was asking this because as you just said, there's a number of people talking about this urban science or this new urban science, and everybody has a different angle. And I was, you know, it's always important, I think, to come to this discussion by saying, okay, this is my point of view. This is what I bring to the discussion and all of that. And I was wondering if you could, so you also said that cities, you know, are defined also through time because of different either value systems or contextual problems. And perhaps one of the biggest problems today might be the environmental one. I'm wondering, you know, we all have these statistics in mind that cities are the most are hosting most of the people, but also are responsible for most of the environmental effects. So how do we reconcile these two in your perhaps bipartite graphs? Or, you know, is there a new urban science in your head that can bring these two together? I think one of the great difficulties is that we have a massive, we have some massive disjunctions, massive disconnections between the kind of science that I'm talking about here between cities in terms of building science, right? Okay, building science, which is quite different from what I'm talking about. When there are puzzles, we're all talking about relationships, equations and so on. But cities is building science. And then cities in terms of other systems that are clearly part of the city, but treated in separate silos, such as the environment. And you can see why this is so, because basically the environment is about environmental issues that the connections through to people and the way they're making impact on the environment is an indirect one, in some sense, as opposed to models of how people work in cities, which is much more direct. People literally working and producing things and moving around and going to work and shop and all that sort of thing, very different from the impact that people have on the environment, which is more indirect in the sense of the modeling in that sense. And so consequently, and when it comes to ecosystems basically, there are a number of books recently looking at cities as ecosystems, but none of them have ever been able to grasp this central notion that we're in the business of trying to relate people who live in cities to all of these different features of cities, which are highly problematic, such as degradation of the environment, basically degradation of ecosystems and so on. Climate change, climate change is a little bit easier, because we do, you can see where we generate in terms of the way cities develop, where our carbon footprint is in this sense. So that's one of the big issues, that in some senses, it's a different language and a different conceptual focus that we put on with all of these different things that makes it very difficult to reconcile it in some sense. And so it's not that we don't understand that there are big issues related to people and the environment, it's just that I don't think we know quite how to think about our future as people working in cities and having an impact on the environment, basically. I mean, if you take something like climate change, if you take something like flooding, which has an immediate impact, you can see a strong physical representation of what goes on in cities in terms of land, land use and so on. And you can see where we might be able to make predictions under what if type scenarios about where places would flood, but actually doing something about it is a much bigger question that seeps into many other things, you know, which really relate to the whole big picture. It's as though we invent all these different things and they scale up and they produce these massive effects, which we have as climate change that then come down again in very detailed focus, basically. And to do something about them, we need to somehow do something about the top level, basically, which is everything, really. So there are some real dilemmas involved in integrating ecosystems with ecosystems with environmental systems. If you take metabolism, circular metabolism, and that kind of thing, I mean, waste, basically, or processes in cities, we can identify them in some senses and we can think of some very simple ways of actually solving some of these problems related to waste disposal and so on, basically. The issue, I think, is not necessarily knowing how to solve them. We kind of in an ideal world know how to solve all these things. But when we work it through into a plan, it becomes extremely problematic because it's highly ramified. It's highly political. It very much depends on ideologies and so on in this particular context. Whereas some of these other things in the scientific cities are a little bit less so than that. I'm not saying that they're not ideological either. I think that we need to think about these things a lot more in those terms. Not a very satisfactory answer, I don't think, to the question you posed, really, but it is in the nature of things that we need to think very hard about how we put together things that we've talked about in a different language. I can talk about environment and I can talk about a new science of cities and you wouldn't think that we were on the same type of tour quite differently about lots of things about the environment which we'd all agree about and lots of things about the science of cities. But when we ask the question, how do we put all this stuff together, basically? How do we use these different insights? And that's a really big issue because it sort of presupposes that there may not be an integrated theory of everything out there, basically. Many people would believe that. I mean, I think we used to think that there could be integrated theories about everything but I'm not convinced at all that that's the case. And of course, we tend not to talk very much about that so we don't know what people think. In my colleagues, for example, if I went round the table with our 12 lecturers, professors or whatever and we had a debate about what they believed in in terms of putting things together, I think we get some very different perceptions dependent on their own history, their background and a whole range of things really in that sense. So it makes it very hard to know how to proceed into this. Yeah, as you say, I think I have this romantic idea of this integrative approach somehow but when you start collaborating with cities as well, you also see how ill-equipped or ill-informed they are for dealing with these complex issues, which are local and global at the same time, which are of different nature and all of that. When they ask a number of questions, you feel also ill-equipped as a researcher to provide a satisfactory answer because you know already that your answer just covers just the first layer as we said. And so I think models still help to at least help them understand, either through cursors or whatever, so that they visualize this entity that lives and that transforms and that if you do this, then there is a range of things that change at the same time, although we don't know where and how and how much. But yeah, I feel that there is also a lack of tools of communication between the practice and the research somehow. Sure, definitely. I think it's important to think about the implications of all of this, definitely. And we should do more writing about it really in that sense. I mean, this is the issue about comparing different points of view and basically taking a much more pluralistic view about how these things can be used in this particular context used and implemented in that sense. I think we have to apply things that are speculative and get things wrong as much as we get them right. And so that's always been the case really in a way. And that's the way we might make some sort of progress. Yeah, before we close the episode, I generally ask two questions. If you have any work or research that you want to spend some more time in 2022 that you would like to share and also do you have any books or articles or films that you would like to recommend to to go a bit deeper on this discussion of series and complexities? Well, there is a there is a book that we produced about let's see about the middle of last year, I think, called Urban Informatics. It's an edited book by Springer and it's open access, right? Urban Informatics. It's a Springer book. It's open access. You can download it all and it deals with a very wide number of techniques much wider than what we've talked about here. But there are some sections on what we've talked about here. It's mainly from the Geomatic Engineering GIS point of view of this book. Although there's a lot of articles in it that people looking at this podcast might be interested in. There's a very interesting article by Sidney Horrible from Illinois dealing with industrial ecology really in that sense. So that would be worth doing. So just go to Google or any search engine and type in Urban Informatics. Great. She S I H basically and it'll come up basically and you can download the book basically etc. So it's it's open access in that sense. So that might be quite useful to some of the participants in the podcast basically in that sense. Also you could have a look at our journal Environment and Planning B, which is getting a bit bigger at the present time. And it's fairly vibrant in a sense and there's lots of new ideas and new thinking I think that we're trying to pick up in this area in that sense. But these are two things that you can do straight away basically really just by googling on the net in that sense. Great. And is there something that you want to focus on for the rest of the year in terms of research or a new project or a new book or something like that? Well yeah what we're doing and what we have been doing in the last year really is we've had a task force looking at what is loosely called the digital transformation in planning basically. It's a digital task force related to traditional planning, well in Britain we call town and country planning in the sense, but it's urban planning basically. And we've been looking at how the planning process and practice in Britain might be informed by new information technologies, by new uses of computers. And this task force you can download the report it's open access and you can download the report from a website called digitalforplanning.com. Digital, D-I-G-I-T-A-L-4 which is the number four planning all one word digital for planning all one word.com. And that really deals with what we think there's a group of 10 of us, a panel of 10 members, what we think should happen in the system in Britain, which is very very behind in many senses in terms of the application of information technologies. Like I think in Switzerland and many other places you have a planning system of sorts which is probably similar to ours where you have to apply for permissions for development and matched against a plan. And that's the way the process works. The data that's contained in the planning permissions, there's something like well over half a million planning applications a year in Britain, there's 60 million people, well over half a million planning applications a year. And that's an enormous stack of data about what is actually being developed or not being developed in British cities. And it's data that is never used. And so some of our some of our proposals are that this data be mobilized in some sense it's it's national data it's in the public domain all this stuff but there's not been the effort or money really or resources required to actually do this. So our task force is really about how we begin to informate the urban planning process in a way that is much bigger and better than has ever been before. One of the things that we're very worried about I think in terms of planning profession is that a lot of a lot of thinking about planning is lost in terms of the use of information technology information technologies models that we've been talking about here are not used really at all in a planning practice basically in that sense. And so in some senses the whole task force is really to make recommendations how this can be mobilized really in a sense in that sense. I will definitely have a look that looks very promising we've been thinking about hopefully having this type of data set so I'm very curious about how this looks. Well thanks so much Michael for your time thanks as well for everyone to listening watching until the end and once again don't hesitate to share this with fellow planners computer scientists or with colleagues and tell us what you've learned today in the comments. Thanks again. Thanks very much.