So, here is a way of looking at the world where we have on the axis is the relatively likelihood of a disaster happening versus the relative impact that the disaster will have on the world. So, examples are electronic attacks, likelihood is very high but the impact might be relative. Well, climate should be also here. Pandemics, it's happening. It's happening all the time. You have the zika thing now everybody gets nervous about it. We have HIV, HIV-AIDS is a pandemic that is happening. 40 million people infected. So, we will talk about it next week as an example. And it's pretty scary. We have Ebola recently and it's pretty scary because it's happening more frequently and it's happening with different virus types. Not only mutations of viruses but totally new viruses come about that we never thought. If you look at that, which actually I shouldn't talk too much about it. But if you look at it so if you would plot the pandemics that are happening that are picking up every now and then. As you would plot to the death against the new viruses that are introduced into our society, it will actually grow exponential. So, we get an exponential growth not only of the number of outbreaks but also of the differences in the viruses. And that's actually the scary part. So, pandemics is an interesting example to look at, and let's do this for just a brief second just to again prime your mind a little bit. And now one of the reasons is of course that we are connected so much, that our networks, that our handshakes, we're just a six handshakes away. So, the question is are we reaching a tipping point? With all so many people around that if there's a very aggressive virus somewhere hidden in Africa, it can come here within a couple of hours of flights, right? And it can spread like that. So, one thing you can do then is actually study of what are the networks of spreading that describe that pandemic? We can actually quantify these things. We can measure those networks and we can measure how the virus propagates over the network, and we can predict what will happen next. And we can combine those networks so these are social networks with the phylogenetic network of the viruses and we can actually learn or teach us a lot about how those spreading things happen in the real world. And again, you see that network are a good way of doing that. The details I will show you, not now of course. And then you can play around with it. This is an important thing. If you want to understand complex networks, complex systems you want to play with it. You want to say like, "What would happen if I will change this? How would the dynamics change if I do that?" And that again, of course, the computer is our tool to do that. And so you can do things like you are introducing behavior changes or someone is connected to someone else and then they're changing behavior to connect to someone else. How will that affect without spreading of a virus for instance? And you can play those games and with that you can actually understand the dynamics of HIV where here the blue dots are the data points of HIV and the red ones actually the simulation. So, this is replay of the outbreak of HIV in United States actually. So, we can play with that things. You could do what if questions. So, you could say what would happen if we have better medicine or what would happen if we changed behavior? These kind of questions suddenly become of interest. And you can also look out this is now back to not to HIV but this is back to influenza. It is transmitted by being close to each other, right? So, you can look at places where you are close to each other, which are places where you share the same bus or you share the same bus station. And so what is the likelihood of transmission in a very dense city, think of Singapore? What is the likelihood of transmission of influenza? And by the way this is it not Singapore but this is an island in Saint Petersburg, Russia where we did this research where we mapped out the whole dynamics of people moving about the city and how the spreading of influenza would go there. And so we calculate not only the movement of the cars, the movement of the buses but also how long people are in interaction in those buses, and then we calculate the networks of spreading that come from that. And with that you can say okay you can try to understand what is the sensitive elements of that complex system to understand the spreading of influenza? So, a few words about crime. People have always confused the idea of people always thought that criminal networks are like terrorist networks and they are not. You can actually analyze that. Terrorist networks are very much up down, top down and the kind of kingpins there. And it's much different structured than from criminal networks. And so one of the things we did was looked by looking at criminal networks for cannabis production. That's the thing that's very much of interest in the Netherlands. We have a lot of that. About every week 30 cannabis production blends are being uncovered and discovered in the Netherlands. Netherlands is a small country 70 million people. So, there's a lot of this thing happening also at home actually. But it's also at home. And the whole thing is, so it has different components, right? So, there are you know there is a mastermind, there are land owners, there's electrician, there's some people who know about plants, and there are a lot of workers, and then there is the distribution of things. So, the interesting thing about this kind of particular criminal network is that it's actually a production network. It's actually what we call in Sociology a value chain. So, there's a chain of elements that need to be connected to each other. So, what we did is we got three types of data and that's if you feeling how you do this kind of research. We've got three types of data. We have arrest data of people that were arrested being criminal involved in cannabis trade, cannabis production. We had data from surveillance. So, from people from police guys going into the street saying, "Hey." Suddenly these guys start to talk to each other come back to their office making a note, digitize all that data. And we had the social media data so the Facebook data, Twitter data because criminals are also on social networks. And so you can actually mind it. So, we combined all that set, all those data, from that we were able to produce the network that actually recreate in the computer the network that produces those drugs. And taking into account all kinds of aspects like I said from the coordinator, the guy who does the cuttings, the toppings all those activities are then mapped into those three networks. One of the things you can do is play in the computer as if you are trying to stop the criminals from doing those criminal activities. So, and that's what we did. So, we were actually able to show that if the way the police is doing it now if they will look at the guys who are most visible or most connected the guys with the big hubs, the guys with the big gang numbers. There are a lot of connectivity. They're very much connected because they are in control of the whole process. And that's what the police are doing. They actually take out those guys that are in control of the whole process. And when you do that, you can actually show that the network is better. So, the production of cannabis increases if you do the way, if you try to solve it by the way the police is currently doing it. You should take out the ones that are mostly connected who make networks stronger. And that's one of the reasons why we still have so many cannabis networks, active cannabis networks in the Netherlands. And since we have published this, and they have changed their policy, and it's quite interesting and that's happened in Netherlands first, and then in the UK, and now it's a part of Europol strategy. So actually they are now, we know how to do it better. I cannot talk about it now, but we know how to do it better. But actually this is one of the strength, much I want to leave you with that to some extent. This is one of the strengths where you bring together all those different aspects, you map that into a dynamical network, you replay the situation in the computer, and you discover suddenly that actually you can understand things a bit better. So, the synopsis, this is the last, second to last slide, synopsis complexity arises from a systemic response from many non-linear interacting elements. The signatures are as non-equilibrium, phase transitions, and there's some kind of transitions anywhere between order and chaos, right? Again, this is hand waving, we can quantify bits and pieces of it but not all. Underneath most complex systems, we find the complex networks. I gave you a lot of examples of that. And we need a theory for complexity. It's not there. Like Feynman was saying, we're just emulating complexity, right? In the words of, well, he was talking about physics then, but still. So, we need a good theory, and not just an analogy of that. And the current methodology that we're using is not only those complex networks but also a thing we just called agent based systems. Individual based where we model individuals that are non-linear interactions and that actually gives us a good start. So, this is again, the bird flock I started with. There was a hawks somewhere around here. I don't know where he is, but he was somewhere around here. So, this is a bird flock, but now it's simulated. This is the computer simulation with three simple rules where each of the birds in there. Just three simple rules the one saying, "We have to get together." One says, "You have to go to the center." One says, "You have to avoid each other." These three simple rules, we actually are able to dynamically model that. And so, yes, we can model complex systems. Okay, thank you.
