 Thank you for staying for the session and there's probably a few things to say first before I get into the meat and bones of the presentation. Firstly, it leads on from what Marco was saying earlier really. I've been asking myself this question about why we see cycling manifest in certain ways in cities around the world and in particular those cities that had marginalised cycling. So in the last kind of 10, 20 years we're seeing it come back into cities. But my question and my argument in some ways is that we're only seeing it manifest in certain ways. We're seeing it manifest as a solution to health problems. We're seeing it manifest as a solution to economic productivity, keeping cities moving. Part of the argument I think being is that actually our cities should be more than just about that. They should be places for humans to live and flourish. They should be places of play, places where we can be our irrational, inconsistent selves and do stupid stuff sometimes. So it is more, that's the kind of question driving me. It's not necessarily saying it's bad that we're seeing cycling manifest in these ways but just asking that question. Which versions of cycling are we seeing kind of coming back into our cities and is that all cycling is about? So this presentation is based on a chapter of a book which I am currently writing very slowly. I expect it to be out roughly when the UK leaves the EU. And that will be by my estimation roughly 2030. So that's when to expect it. Don't hold your breath. So this is basically a quick gallop through one of the chapters there. Other things to say. I'm not anti-bike share. Once you see the end of this presentation you might think wow he's a bit negative. That is not the case. I'm quite pro-bike share but maybe there are ways of doing it that are better and ways of doing that are worse. Methodologically this is still a work in progress in some ways. It's based on a series of interviews with people working at Mo Bike and OFO and in the Shanghai Government and advocacy groups. It's based on news reports. It's based on looking through investment records and it will be based on looking through the patents of companies like Mo Bike and OFO as well and what they can tell us about what they are hoping to do with what they've learnt from users. So with those caveats kind of out of the way and I'm sure there's some more that I've missed. I'm going to try and run through. So has anyone read a lovely book called Surveillance Capitalism by Shoshana Zuboff? Cool. Excellent. If you haven't read it, read it. It is brilliant. It will probably take you a little while. It's quite big. But when I, you know, when myself and Dr Lin we had this data and we were working through it it really was one of those light bulb moments. You're trying to make sense of your data and here's a book that provides a framework that really helped us fit what we were doing kind of into that. So I'm going to give you the world's quickest gallop through Surveillance Capitalism kind of 101 in a way because it's about 600 pages. I'm going to be challenging cat here for a number of slides. How quick can you get through a presentation? So then I'm going to go on. My main argument being that companies like Mo Bike and OFO from the Dockless Public Bike Share fit into this mould of Surveillance Capitalism. Whether they intended to to begin with is another question. But what I'm arguing is that they are currently on a trajectory that fits them into this model of surveillance capitalism. That they, in terms of their origins, their expansion who's invested in them, they fit the likes of Google and Facebook in some ways. In terms of how they're operating, in terms of their architectures of extraction, extracting user data. They work on the same ideas about scale, scope and action which I'll talk about. And their products are the same. They are predictive products. They are there to make, you know, so that they can put some data together. They can sell it as a product which will help someone in the finance industry or the advertising industry make a better bet effectively. So they're in the business. Ultimately of selling predictive products is what I'm arguing. They may not end up realising this. They may just fail as companies. They may find another way of operating. But at the moment my argument is that they are on this trajectory. So here we go. Surveillance Capitalism 101. So Zuboff is effectively arguing what this is. A new economic order that claims human experience as free raw material for hidden commercial practices of extraction, prediction and sales. So effectively the idea that there is a shift from extracting surplus value from labour relations to social relations. So our everyday social reproductive activity, whether it's web search, whether it's wandering around our house and talking to Alexa, whether it's using MoBike, whatever it is, our everyday social activity becomes a source of surplus value in the form of our data. The idea that these extraction architectures find out something about us. They find out something about our preferences, our likes, our dislikes, our income, how we move, where we go, all this stuff. So that is kind of the linchpin of surveillance capitalism. Then companies will analyse it and they will try and bundle it up and turn it into a product that they can then sell to someone else. Whether it be a finance company or an advertising company, because it tells them something about their users, which means they can target advertising more effectively, they can say whether they should get a loan or not, or whether they should have a higher insurance premium. So it's these kinds of predictive products that potentially this data can be used for. So as Zuboff says, the key data source that can be monetised is our behaviour. Zuboff uses very much. She doesn't talk about public bike share. She talks about the likes of Google and Facebook and the idea that in the early days they used web search and Brinham Page just wanted to create the best search. They wanted you to be able to find the page that you wanted, the thing that you wanted as quickly as possible. But out of that, at some point, which I won't go into, they also realised that there was a valuable resource from that as well. That they started to understand something about users which they could use for other ends. So beyond making search optimised, they could do something else with that data. They could monetise it effectively. This becomes what Zuboff calls behavioural surplus. How do you use that? It was interesting as far as Zuboff is concerned that we get an empowerment and disempower from this. That we get better search. That we get a free or very cheap mobile bike ride or whatever it happens to be. We are empowered to some extent by this process but perhaps we are disempowered as well. Because we are not sure who's making the decisions now to some extent. We don't know what data is being taken from us. We don't know where it's going necessarily. We don't know how it's going to be used. So there's an issue potentially about disempowerment. As I've already said, central to this extraction is this idea that it's in the realm of social reproduction. What's interesting particularly about things like public bike share is that you can gain a certain amount of knowledge by just letting people cycle around. Perhaps you could give them some kind of app and you can know about that. If you really want more users and you want more data then why don't you induce that action? Why don't you provide them with an architecture which enables them to do the thing that you want them to do? So something like public bike share is quite a good example of that. So it starts to kind of, as Creavy says here, provide a terrain which guides us in certain ways. And as I've said, the ultimate goal of this is to create what are called predictive products. So we feed that behavioural surplus into particular processes fabricated into prediction products to anticipate what we might do now, soon or in the future. And that's valuable to lots of industries to know about us a bit more. If they're going to spend a big advertising budget, why not at least know a bit better that you're going to respond to those adverts, for example. Okay, so enough about that. The research that we've done took place in Shanghai. When we started on this, we didn't even think about data. We were just, we'd been doing public bike share research in Taipei and London and then we were in Shanghai. So we started looking at it over there and it just happened to be at the time when the explosion of the dockless operators took off 2016, really. So Shanghai, 25 million residents. A very dense city in some places, up to a thousand people per hectare, average density 200. Relatively low ownership of motorized vehicles compared to some context. 20% of journeys in central Shanghai are made by car. A falling ratio of powered two-wheelers and self-powered two-wheelers. And that was really because the government, late 90s, early 2000s, really had a policy promoting the car and wanted to marginalize cycling. It's only really in the last 10 years that cycling has been pushed from a policy perspective again. As part of that, the city had its own docked public bike share system. For a number of reasons I don't have time to go into. There's a paper going to be published on it soon in Urban Studies. Look out for it. Is that it wasn't massively successful. It was only instituted in a certain number of districts. It was quite disjointed. It never got huge kind of ridership. And into that kind of scenario came the dockless public bike sharing companies. At one point in the city you had around 30 operators operating upwards of 1.5 million bikes at one point, now back down to just a few. So if we look at the origins, how are we doing for time, Kat? 8 minutes, right. Better get going. If we look at where the companies have come from, if you look at Mo Bike in particular, came out of the ashes of Uber and DD to some extent. Operations that were regulated out but had a certain knowledge base in terms of ICT and mobility and was then applied to a different kind of model. So coming out of tech companies, but they've been funded by venture capital and internet giants. Very much mirrors the investment that we saw in Google and Facebook in the late 90s and the early 2000s. And you can kind of see that if you look at their investment. Mo Bike, early investment, small venture capitalists, later rounds, people like Tencent, big Chinese internet giants investing a lot of money. By June 2017, one billion US dollars invented in Mo Bike. Mo Bike was acquired by Mytarn Dian Ping for 2.7 US dollars fairly recently. We look at OFO, quite similar, early venture capital, but then later rounds dominated by Asian internet giant Alibaba as well. So certain similarities between those two companies we could argue. Now it's not to say that they were always going to be surveillance capitalists, these companies, or that they even will be. Maybe they don't monetise their data. And they were certainly started if you look at a lot of the press releases and talking to people in the company. It's undoubted that they wanted to make a difference to transportation in cities. They had quite pro-social goals to begin with and again that mirrors the likes of Google. They wanted to make search better. Facebook maybe wanted to connect people together. So it's not to say they didn't start with very pro-social goals. But, okay, do they actually make any money out of doing that? Mo Bike has a massive monthly operating deficit. Just the example of Hello Bike. It's operating costs of four times. It's revenue currently. It's another provider. When Mytarn Dian Ping purchased Mo Bike, it assumed 700 million in debt. Evidence that even after a few years of operation, no money is being made here. High rebalancing costs, low high fees, high rates of vandalism and attrition. All things that have contributed to the fact that these companies are making any money. But the question I'm asking is about, okay, how do we get to this pro-profit situation? And it comes from Zuboff. So she asked the question if Google is a search company, why is it investing in home smart home devices, wearables, self-driving cars? If Facebook is a social network, why is it developing drones and augmented reality? So we can ask the same question of PBSS and Mo Bike and the companies invested in them. If Tencent and Alibaba are search companies, why are they interested in bicycle rental? Why are they investing in these companies? So Zuboff says activities that appear to be varied and even scattershot are actually guided by the same aim. Behavioral surplus capture. They want to know a bit more about users so they can bundle that up into predictive products. Okay, generation and extraction. So, as far as Zuboff is concerned, if you want to create a good predictive product, you have to have economies of scale. You know, you have to have as many users as possible. You have to have scope, i.e. you need to infiltrate as many areas of people's lives as possible in order that pie chart that's being assembled up in the cloud of the user. You want to populate as many chunks of that as you can. So you need scope and you need action. Why wait for something to happen when you can induce that action to happen? Okay, three minutes. So scale. The closer the N is to the entire population, the more accurate the predictions, the more valuable the product. And if we look at Mo Bike, for example, if you look at Shanghai in general, the growth of bikes, fairly meteoric even within the space of six, seven months in one year from a quarter of a million bikes to 1.5 million bikes and the peak of operations in cities around the world, 200 cities for Mo Bike, 250 for the peak for OFO. Scope. It doesn't just need to be vast, it needs to be varied. We need to know things about different aspects of people's life. We don't just want to know about their income or if they've got a cat, we want to know a bunch of other stuff perhaps as well. So scope relies in order to get that on the disappearance and the diversification of the extraction architecture into mundane objects. And that includes bicycles. So moving out from these fixed virtual architectures into the real world, public bike share in this form is a manifestation of that, I would argue. And then action. So we don't just want to wait for this stuff to happen. Why just wait for people to ride bikes when actually if we give them a bike, we're going to get that much more information about their daily movement. So we can create, we've got more scale and certainly at that point. OK, and then in terms of predictive products, how long have I got left cat? OK, that's not great, but let's see what we can do. OK, predictive products. So arguably the future success of PBSS will lie in its ability to do something with this data, that it will be valuable to someone. And there's a couple of kind of more obvious ways. So some of the operators have kind of a point system in the app, a credit scoring. What sort of person are you? Do you damage the bike? Do you bring it back on time? Where do you leave the bike? It all says something about you as a person, which is potentially valuable to someone. And some of the kind of products we see potentially being made. And we're already seeing transport models come out of this commercial and financial products as well. So from people we talked to, for example in MoBike, the big data is very valuable. We know it's a gold mine. They weren't quite sure what they were going to be doing with it at this point. They just thought it was valuable. They also knew that they needed to combine the data about that pie chart on its own. Maybe that MoBike data about movement wasn't worth that much, but if we combine that with other data sets that the internet giants might have, it becomes more valuable. So they've been experimenting with transport models, working with the municipality to some extent in testbeds. When we were there a couple of years ago, that process was just starting, but certainly as far as the new area construction and transportation commission was quite slow going. But there is perhaps the potential to create transport models and say something to the city that is valuable to it about how its citizens move around. I'm not saying there's a problem with that, but perhaps in the ownership of the data there is a problem. We also see that the citizen there is empowered as a sensing node, but we're not empowering them to interpret their data or act upon their data in other ways. We can also use root data to target advertising. We see precedence in this in things like Pokemon Go, where a particular store can pay to have a poker stop near it. We've seen this already because you are literally going to make people go to that place. It's inducing action. In the same way you know from some of this MoBike data, you know where people are going to go, and that means advertisers, shops, retail, whatever, could target things a bit more closely. Then financial products is an interesting one. The credit rating function of things like the MOBI Act have the capacity to render user data on things like trustworthiness. We already see precedence for that in the finance industry, low cost loan industry, the insurance industry, taking data about trustworthiness that tells them about something about you. What's interesting there is that we might think it's a very benign transaction. We might get a free bike ride and some data goes somewhere, but that data might start acting back on us at some point in the future that we are not aware of. When we can't get that loan, is it because something about our mobility, because we happen to make a particular journey, lived in a particular area of the city, that it was a proxy for actually this person will be a bad credit risk. That data-fied style starts acting back on you in ways that you didn't know at the time. It's going to come to pass in public bike sharing, but it's a potential that happens already in other industries. So very quick round up, because I know I'm over time, issues about sustainability. If these operators cannot monetise and they go bust and they were the gap, they were filling the gap in the transport system, what happens for that kind of city administration? We've seen that in places like Manchester and the UK that were reliant on MOBI for a while. They've withdrawn. There's a gap. There's a transparency for users. What exchange has taken place? For what purpose? Do we know? Do we care? Weak to non-existent citizen participation here in terms of our data being used, taken. We are just a sensor now in terms of our journey. It represents a shift in governance in the city. It positions the city administration as a customer for this data on transport modelling as opposed to the owner of this data. So arguably there are some shifts there in terms of city governance. Thank you. Do you want to use this one? No. Thank you, Justin, for your presentation. You were a few minutes over, but I cannot be a judge. Does anyone have any questions for this very different kind of presentation? Can you go back to the GEC ready guide and tries to go out of the door and says to him you have to buy flowers for your... Oh, you're right, yes. So you have to change it. It's not Siri who says to him. It's his bike who says to him. You have now to go to the shop and I'm going to bring you to the shop. You're not going to work. I'm bringing you to the shop because you have to make up for your wife and you have to buy flowers. Okay, okay. I don't know if this is positive or negative, but yes. Okay, good. I will reach all this. Sorry. So you were totally correct in your prediction that it seems like you're quite negative. At least to me. So I'm trying to think... I mean, some of us might be more of an advocate when we think about cycling. So what could be the positive spin? To me, it's not... Again, it's not to say that this is a bad thing. I think the key issues for me are that issue of governance. If this data is privatised and we have seen instances, Manchester was one of those cases where they wanted to work with Mo Bike. But one of the things they wanted was access to the data and they wanted access to the data and not have to pay for it. Mo Bike are no longer in Manchester. So draw your own conclusion. There could be a number of reasons that they pulled out of Manchester. Perhaps one of them is that they don't want to give that data away for free. Now we could argue, okay, perhaps it's just the way we're going. The municipalities are now going to have to pay for data to know about their citizens and the movements of their citizens. Shouldn't we keep that in-house? It's just a question. I'm not saying one way or the other. Perhaps this is the brave new world. So I think that's an issue. And it's not confined to cycling is about transparency in terms of us as users. What data is being taken? Where's it going and what's being done with it? There needs to be more transparency and again it's not just about bike share. I think that goes across the board. We seem not to care. I think cities need to be more self-confident and sort of dictating the conditions under which private people can operate there because without the space they couldn't also let the bikes travel out. But I think there's a lot to learn from this kind of data. If you look at how people move on bikes through the cities, if you also think about the presentation from a Mexican person, where are missing links within the urban bicycle network and you would be able to see where people break harshly, where they accelerate, where they have a good flow, where they need to take a lot of sharp turns and so on and so on. You can see all the stuff from the GPS data in the bikes and there's no rocket science. So any city could basically, or all the cities in the country are durable, whatever, they could be like, let's make all-way feats that has this sort of anonymised even GPS tracker and they're just to learn about the infrastructure of our cities. Yeah, no, no, absolutely. I'm not saying that there's not some valuable data here. I guess what I'm saying is that the city now has to buy that data. Even though it's supplying its space for free to the public bike share companies and it gets that service, but should it also have the data for free? But that's just that they don't, though. When they make agreements to allow these companies to operate within their cities, the data, giving the data back to them simply has to be in those agreements. That's a very easy thing for a city to execute. Well, you say it's easy. It's easy, but in cities that maybe are still suffering the effects of financial crash in cities in the UK, administrations have got no money. A lot of them have welcomed this. They've got, oh, thank God, we don't have to now pay to have a public bike share scheme. So they're quite willing to have that trade-off, but that trade-off, they are selling their citizens' data to get that bike share system. But there are several cities in the US that have been very smart about this. Yeah, no, it's not to say you can't be smart about that. Yeah, yeah. There are still more hands up, and there is a debate going on, but unfortunately this is a conference and it is my turn to say, turn this over into the drug session at the end.