 Beijing or in New York or Canada or anywhere in this world. First, I'd like to welcome to you all and thank everyone for attending. I'm Deng Qingmao, the President of Urban China Network at Columbia GSAP. And for today, let me briefly introduce Urban China Network and Urban China Forum this year. The Urban China Network is a GSAP urban planning-based organization formed by students in Columbia. We aim to create a platform of communication and exchange for Chinese urbanism across multiple disciplines, through hosting events and bringing together planners, entrepreneurs, government officials, scholars, and students. Urban China Forum is an event which we host annually and each time with a different thing related to planning in contemporary China. If you're interested, please check out our official WeChat platform for the digests of speakers' presentations from previous years. Thank you so much. And this is the eighth year of Urban China Forum with the theme Smart Cities and Info Universe. I would like to thank our sponsors first, of course Columbia GSAP and the Weatherhead East Asian Institute and Minty Mentor for making this forum possible this year. Weatherhead East Asian Institute is a hub for the study of modern and contemporary East, Southeast, and Inner Asia at Columbia University. And Minty Mentor is a professional career consulting organization that gets international students to learn their dream job. It provides networking opportunities and training courses, et cetera, on the related service to those job applications. And we are proud to host a two-day international forum here. And this is virtual meeting room with all of you, of course. All right. Urbanization in China is accompanied by a profilation of urban data. Contemporary metro politics produce countries, countless information and data, which are then captured and utilized to inform the urban policies. In recent years, local authorities have planned for the development of digital industrial parks and have promoted politics in favor of the growth of tech sector, envisioning cities to be served by the data they generate, innovative interpreters emerge to provide data-driven urban governance and planning solutions. In this forum titled Smart City and Info Universe, we aim to analyze the impact of urban big data in global organization and to shed light on the future development of smart cities in China. We bring us together and we are very glad to be joined by five leading participants, the scholars, especially this day, this year. And for today, we are glad to have three of them in two groups. I think I believe everyone see them in the video box. And Professor Xin Yue Ye will present on AI-driven framework for urban flood resilience to your design. And Professor Alan Smart and Professor Dean Karin will present on data-driven governance, smart optimism, and inequalities in China, security, surveillance, and informality. Yeah, so, prior to the start of their presentation, I would like to clarify the logistics. So, we will mute our audience and turn off the videos during the speakers' presentation. However, please feel free to type in all your questions throughout the speech. Our UCM members, Xin, will collect their questions and read out to the speakers during the Q&A section. You can also raise your hand during the Q&A section, as well as to ask questions directly with the moderator, Kaosong, and Amuse you. You could find a schedule in the chat box after each speaker section. And please remand it that the forum will be reported and the schedule has been posted. Yeah, and I would like to introduce Professor Xin Yue Ye, our first speaker of today's forum. We are very glad to have you here. Professor Xin Yue Ye is a Harold Adams-induced professor on interdisciplinary built environment sense research. He is an associate professor of stellar faculty progress target for urban computing at Texas A&M University. Also, he is a director of the urban data science lab. His research focused on geospatial artificial intelligence, big data, smart cities, and urban planning. He is one of the top 10 young scientists named by the World Geospatial Developers Conference 2021. We are very glad to have you here and please let's welcome Professor Ye. And you can share your screen and I will stop sharing here. Thank you. Let me share my screen. Yeah, we can see that very well. Okay. So how many minutes do I have? 30, 40? Yeah, yeah, 30, 40. Okay. Thank you. Thank you for the opportunity for me to come here to give the talk. One year ago, I was still in the Metropolitan New York because I was a faculty in the College of Computing in New Jersey Institute of Technology. So I have a very diverse background. Before I came to U.S., I was a professional urban planner. I got a PhD here in UC Santa Barbara in Geographical Information Science. Then I moved to geography. Later, I even become a computer science faculty. Only one year ago, I relocated to Texas A&M University to join the Department of Landscape Architecture and Urban Planning as well as the Department of Geography. So my research interest is to integrate urban planning, geography, and computer science for urban sustainability. So my work these years, I mean, especially ongoing projects, mainly supported by Microsoft and National Science Foundation, if you look at these projects, it's very much, it's about spatial decision support and smart cities, especially after I came to Texas A&M because Texas has lots of challenges, two challenges. One is definitely a good thing, the keep growing population. And so it will bring the burden to the building environment, but definitely also further expand the many opportunities. But on the other side, we have many kind of hazards. One very unexpected hazard is February. This February is no stop. So we are working, it clearly gives us a picture is we need to work on some kind of digital systems which is highly interdependent because on one side we have population and urban dynamics. On the other side, the urban extremes, sort of extreme weather and also the different governance mechanisms across locations. So it let me think of like over almost a similar time of how the when the geographical information system was introduced in the United States around the 1960s, as a Mark Hager, when he developed the School of Landscape Architecture in University of Pennsylvania, he promoted the idea of convergence. So he not only hired landscape architects but also introduced soil science, hydraulic geography, all kinds of different disciplines because he strongly believed we needed interdependence among disciplines to achieve kind of the optimal outcome for landscape design. So this kind of idea influenced the geographical information system technology development. And nowadays, if you look at like it's a recent two years ago, the report from Princeton is looking at the Houston urban sprawl will increase rainfall. So in other words, we do need as a multi-discipline to understand in the coastal building realm and how different disciplines from the kind of the climate science to urban science together will influence kind of the sustainability and resilience of our communities. If you look at the national science petitions, kind of they have a webpage called the communities in the 21st century. For that is putting together many national foundations, relevant projects for which is to fund the urban relevant research. I checked, I just pulled down their five major questions. Very interesting, very much relevant to the smart cities and big data technology, because first they want to ask how we can make the prediction, right? Then second is ask what kind of theories? The third is how to serve the people. The fourth is if you have successful innovation, how can it be transferred to others? This question is actually is a very artificial intelligence question. That means how you can do the transfer learning, because we spend money mainly in the large metropolitan areas. We accumulate so many experience, how we can move the similar experience to other communities which lack technical support, lack financial support. The last but not least is how to engage communities. To engage communities to me is more like how we can do the knowledge co-production, how we can do the co-learning, learn with other people. The interesting thing is when we talk about smart cities, many times we are working on well-developed or well-established research questions. It's very much developed by scholars, but if you go to the field, you will have many open questions. You have many local data you have never expected. So how we can deal with that? So in today's presentation, I will show some of my experience on that. And one more thing is I read this report. It's a free for download from National Academies of Science, Engineering, and Medicine published in 2019 talking about building and measuring community resilience. Actually, it's a mainly action plan for the Gulf research program, but it applied to most of the coastal communities. This report asks for three action plans. One is we need to adjust the dynamic state of communities and their changes in risk and resilience over time. So emphasize its dynamic change. Second is we need to link the data information together. So this is I will also emphasize it later. We have many scattered data information, separated efforts. We need to link them together. Third is decision making. No matter how you do research, we need to really benefit, adjust the challenges from the community. So we need to think about how we can better our decision making procedure. NSF, three years ago, NSF established a large amount of money for what are called convergence accessor data. That's another effort is to turn research into action to benefit the society. Very luckily, I was one of the first group of researchers to receive this funding. And for that, we established something called collaborative dual decisions. Currently, I'm the vice president for spatial decision support consortium. And for that is because we realized for planning from the top down, move from top down to more collaborative decision making is you do need to get in many stakeholders, experts, local residents together to make a decision. However, we have many different terms of understanding the world. We have different interests. We use different terms because we are from different disciplines. We even though we speak English, but we speak different terms now how we can make each of us understand and make the decision transparent. So we are here, we utilize definitely the data is geographic information system based. But we also implement new generations of knowledge graph for the collaborative workspace. So you will see the examples later. And also last year, I published one of my books called spatial synthesis, computational social science and humanities. We invited the member of academy of science, Dr. Goodchild to write the front page. He argued is nowadays a problem faced by humanities are more challenging than they have ever been, especially if I use coastal sustainability and resilience as example, because we have more people move towards coastal areas, but we we are not very much ready for so many extreme weather and population pressure and definitely social equity. For that we by nature is collaborative. However, for discipline is by nature is a central petal because everyone mainly focus on their own research question. We really need to reward the broader perspective. So in other words, again, is how we link the connected dots together. So if based on what we have reviewed so far, I can summarize we have four big challenges now. What first is we have fragmented research, resilience actually is mentioned in many, many disciplines, but we really need to add and not only in different disciplines, the discipline look at the things at a different scale, a climate science is very macro level, but landscape architecture or architecture is a very building and a neighborhood scale. So how we can link them together is called a cross scale holistic investigation. Second is knowledge divides as I mentioned before scholars with different specialties, they tend to speak a different domain language, and they have different criteria. And about we need really need something to be commutable. So how we commute, how we communicate these kind of different ideas. Then we come into decision making challenges for decision makers. Most of them will not really have time or knowledge to read all the models, equations, diagrams, everything, lots of overwhelming information, how to give them very much clear message what to do. So that's a third challenge. The fourth, as I said, fourth is not the least is community engagement, how you can get as a public getting involved and how to raise their awareness and their awareness, how we can turn their awareness into data coming back to the decision making procedure. So for that, there is some initiative solution we have been doing develop a large scale knowledge graph to link the knowledge pieces together and also develop the information hub for collaborative and collective knowledge sharing. And it's a call for that is for artificial intelligence driven system to assist and facilitate the decision making. For me, artificial intelligence is not a come to replace people. Instead, artificial intelligence is a come to facilitate the human computer partnership. So artificial intelligence will really promote the collective decision making or collective wisdom. So eventually, we will get the other general public involved and help grow the platform together. So when we talk about how the knowledge now is distributed across skills, I utilize that there is an example, as example, by two Japanese scholars, they talk about resilient urban form, a conceptual framework, very much in detail to list all possible attributes from macro scale to find very like a and let's say the fine skill from macro skill, they are looking to the here is at least I will not read through is from the kind of the regional connectivity urban sprawl urban development type and then method level is land use makes the access to parks or open green space all the way to the building and proper level. For each, to be honest, for each average attribute, it's worthy as sub discipline to explore there. And I don't believe there is one person who understand every attribute here or can handle the kind of the mechanism or theories for each of that. However, that's the reality is this is a can maybe still incomplete the list of the knowledge we need to understand as a resilience. So I will come up a kind of something like called a resilience knowledge graph. You see here, can we link everything together from the from the there's a macro level like architecture all the way up to the urban planning from micro to macro. And for that, we, for example, we can at a building level, for example, I just use some things that given the size of a diagram, I just pick pick up some representative indicators. For example, in the building, we need to know the foundation structure, materials, information and the neighborhood level, we grab as a green infrastructure, these kind of soil, your condition, vegetation. And when we go to the more macro level, we are kind of temperatures of rather plain information, vulnerable, vulnerability scores that together we form a kind of a bigger knowledge network allow us to decision to at the same time, we involve the expert and the resident. It looks like he is a beautiful conceptual framework, how we can really implement that. Definitely, we have some belief is as a point out by Barabash, he said, actually, the dynamics for many social technology economic phenomena are very much driven by individual human decision or actions. So now we really need to put the human into the center to develop the system. So that's, it's just due to the size of a diagram, we show the partial on project for to understand the flooding. For example, based on the flooding literature, we develop oncology to see how, for example, if you give a house, what kind of the features of the house will either be relevant to flooding. So for example, a house, you will see the types of house. The house is what's the relationships of a house, where does the house attach to what does the house next to and the house itself also have certain its own components from foundation to story. So for all of that, we can link to the spatial information or text information and also image information. When they are linking them, we are developing something called a scene graph. The scene graph is some recent growth in computer vision. So previously in computer vision, it will help you to identify the items in the picture, but a scene graph not only help you to identify the item, they also help you to identify the relationship. So in other words, though this is a picture showing the house, car, but through the scene graph, we not only identify which is a car, which is a house, we also identify their relationship. So on the both on the right side, you will see a building the relationship, which is in front of which, which is on the side by side of which. Since we have many such street view images and pictures, and we also know each house information and also we know which is the link relevant to which, then we can build a huge connection from bottom up all the way to the original level, because when first we know the house is next to a park, as a park is located in as for the census block group, and this block group is in census check, census check is next to a highway. The highway will link multiple census checks together. And this highway, for example, will also extend to the coast, and the coast will link to other infrastructure for like a flooding protection. So in other words, through these kinds of the image, the text, the relationship, we can link all the elements together. And not only that, I remember we also mentioned as architecture design, I worked with architects. For example, if you have a building here, can you list all the possible criteria telling us is does this building has any design go against the flooding resilience or any design is a good feature to promote the resilience. We have we also changed thousands of these pictures to allowing us to translate on one side, translate the picture into text. If we can translate the picture into the text, it means we can link the picture or link the design scenario with many, many of the laws of design. So if it functions, it means even the two pictures looks like so different. But however, the design philosophy might be same or similar, we can link these pictures together. So it allows us to further strengthen the connections in the knowledge graph. So the good example is we developed an annotation system to allow ask the designers and the students and the local residents to upload the pictures. And for in each picture, they are talking about how they feel about the design of this building, how good it is for preventing from being flooded. And they just use a natural language is they just use your common language to type your comments. The good thing for that is I think everyone has experience to search in Google where Google is not only when you type in the Google, Google will not give you the same or similar words. Instead, Google return your answer, right? Because Google why people like to use Google because Google is a semantic based semantic space search. It means they try to understand what you want and reply the answer instead of send back the same or similar words. Why because Google is a knowledge graph based. So we are thinking about is can we utilize that to build as something similar as a Google, but we use the wisdom of the coastal flooding resilience. So when you throw in the pictures, I can match the similar picture for you. For example, in the left, if I load a picture on the left, the left the most picture. And my system immediately returns the right four pictures for you. If you just look at the right four pictures, I mean, at least in the first two pictures are not so, the right two pictures are not so similar as the quality image on the left. However, they have many, their similarities because the front door, the lowest floor has a similar design feature. It's immediately being detected and it returns a similar image to you. So in other words, it's a kind of image intelligence. By checking the same, they will return the similar image to you. Why we have a similar image? Because I change the image that you know whether the design of the building is consistent with certain criteria. What does a computer do? If they are going to look for the criteria, the text criteria, when you have, for example, 90% or 80% similarity, I will return, certainly I can design it more complicated. I will say based on relevance, I can even tell you what's the percentage of similarity. It's almost think about now you search literature. The literature is also can be sorted by relevance. I can, I mean, image I can also sort it by relevance. I can make it even more transparent telling you is what's the percentage of similarity. So for that, that's another important thing about artificial intelligence. Nowadays, we have a big concern of AI. It's a black box. It especially cannot be explained, but it's linking the image to text. And we can clearly tell you why we think it's the same. Then it's a highly explained, it can be explained. And not only that, because when we talk about digital team, another big thing is we do not have the lowest floor elevation information for most of the house. I know there's two ways you can get the lowest floor elevation for flooding resilience. One is a fly jones and a fly jones. And a fly jones, I mean, definitely if you are a wealthy community, you might already be fried quite a few times to help you to monitor the building environment. And all you have some building to help building certificate, they tell you exactly what's the first floor elevation. But we have many long down communities. We have many low income communities, or we have places which will not allow you to fly jones. So we have been thinking about some very inexpensive way to deal with that way. Recently, we just published is using the Google Studio and the image processing approach to automatically retrieve the lowest floor elevation. And in case someone want to curious why we need to know lowest floor elevation, because when there is a flooding, if the water come beyond your lowest floor elevation, it means your house is flooded. If your house is flooded, then your property value will be immediately damaged. And given, I mean, definitely if it flooding, we can have more equations telling you when the water going by inch, how much damage it will bring to the house. And if we go to certain level, people cannot get out from the house, then there will be life loss. And if one several inch further, the whole house might be brought down because of the pressure from the water. So this is certainly will allow us to understand the long term impacts of flooding, long term impacts of urban climate upon the real estate market. So this is the fundamental information. So we develop this technology. And this is because we just how we further understand to detect the lowest as a door in the building is we calculate, understand how Google Street View car pass by the street, how they take the photos, then use algorithm to re calculate, then your computation to detect the doors in the building. And certainly we have some detect some doors is accurate. Some is wrong. But we that's that's normal for computer vision task. We keep updating our algorithm training and using the correct images as a reference to keep enhance our algorithm. And and for that, during that procedure, we also realized some more difficult thing, because we call the data challenging community, because for the street view images, for I will notice that in quite a few some communities, they haven't updated street view images for many years, or in some community, they even know street views. So there is a data challenging communities that also is a big research topic in small cities, especially in the lack of for the flooding resilience, because if you do not have a holistic picture of the community, you only pay attention to the wealthy community. It's a big problem because sometimes the wealthy communities are side by side with low income communities or actually is a city as a whole is a you you cannot even know any other any building cannot be left. So we need to get hold of all this information. So for that, we are developing, we are using as founded by National Science Foundation, develop as a GPS enable the video detection. So in other words, we drive into these communities to take the videos linked to with our narratives and other information. And look at this is in the in the middle of the map. We use this is an example I used before for ACRON. In ACRON, people drove through the streets and they talked about their perception of the environment. Then we can link it with the images because these street views or the images taken around the trajectory can be linked together. For that, we can build for that use is at that time for the for the looking to drug issue. But it's a similar thing we can apply to the resilience elements. As I said, if people say, hey, there is a problem of the resilient flooding issue, they can always speak up. However, people from different disciplines, they might have different perspective, local people, experts. And you can consider on the top two maps is people's opinions or their attention points are mapped by different color means different group of people. Then you will see how if they are highly consistent on certain location, yes, that's a place we need to action. But if they are highly like a divergent in certain locations, it's also worthy of notice. So we use that we develop something called a geo-visual system is allowing people coming to a landscape to talk. When you talk, it will turn into transcript and even later, you'll definitely you'll have opportunity to revise your text. So that means our understanding of the building element will be combined by your movement, the pictures, your perception, right, the mental space or in other words, we have absolute space, mental space, perceived space together. So as I said, is we need to understand what other people also think. So if you ask different group of people to pass by the same landscape, based on what they describe the difference among them, so you will see how you can better coordinate everyone's kind of interest, because when you bring people into the field, see what's there and see their response, sometimes even based on interesting information, how long they stay in certain location, whether they speak or not. So what we did is say there's a yellowish thing, the yellowish means people talk, they have some discussion here. So if they have discussion, they mean they are interested in here. So even these whether they talk or do not talk, how long they stay in certain location will also reveal people's attention or their understanding of the landscape. And that's all definitely for someone might think is whether that is kind of the privacy issue. What we also consider that is developing the polygon and the more larger polygon to hide information inside. So in other words, if you are in each of these polygon, your information will be the same. So we will not as information inside will not have any difference, it will be summarized. So we also using, as I said, there's many ways that we can use natural language processing to link these images together. So we also developed some keyword traits to link the urban image. So remember when you travel into different locations of urban building environment, you will always have a feeling as if you go back to your hometown. Or you remember that you saw it before, but you cannot remember. Or you went to a park or you walk to a neighborhood. You suddenly you feel you are inside is similar as your house, right? You will find some many similarities. So we allow us to link pieces in the building environment together. So as Michael Betty in I think three years ago, he published a book called Inventing Future Cities, because he said it seems for many years we're talking about how we predicted the city. But given the many inevitable technologies, like as Kevin Kelly mentioned, we have many technological forces that will shape our future. Can we, instead of model our city, can we invent our future cities? And for that is I just started my new National Suns Foundation projects on digital trains, the Texas A&M homepage also reported my project. What we do is we develop digital trains for the Texas coastal communities to promote the resilience. And for that is what we do is, because as early in my presentation, I mentioned in early February, Texas has unexpected snowstorm. And after snowstorm, you need to report your loss. However, after several months, quite a few communities still have not reported their loss due to the lack of technical support. So we are thinking about things Texas has experienced so many all kinds of disasters. Can we build a huge disaster level using the integrators of big social data and the weather simulation to have a quick estimate of damage. So it will allow the resource to be quickly deployed. And for that, we receive the support from Microsoft. So, so as I said, when we have all these new advancement of technology, we really need to link it to the location and it need to be input as a knowledge graph to connect things together. However, in the center of that is human being. And for human being, I'm not only talk about individuals of perception, I talk about many different people's perceptions. So for each of us, we have different knowledge. I appreciate everyone's knowledge. So our purpose is building a platform allow everyone's knowledge can be merged, can become a much larger knowledge graph. So I mean, in China, we know that all the things, the three common process knowledge will be much stronger than one wise man. So the so the knowledge learning is through human computer co evolution. It will make our system much smarter and sustainable. So for that is this is a big actually my big dream of in the future, how the smart cities should function is on one side, we need a link that make our system more can be explained. So it means we need to have a link many diverse data unclear links, putting together but also facilitate the communication amongst different stakeholders and since many information from different disciplines, we need a knowledge graph on behind to make it closing the gap. And we are developing more data fusion modeling skills to linking these elements together and engage and people because I fully believe it through that is if we grow our wisdom, grow the collaborative technology ability, the system will also grow. So the software will become smarter because of human become more collaborative. And since today, I do not have too much time to talk about the how we integrate the crime data with local data because it's relevant to downskill downskilling technology to link data across skills. For that, I will say we will need we will integrate as a social and a physical convergence for digital twin. And if it functions, it will allow us to have many fantastic applications in the for coastal resilience. So I think that's all for my talk. Thank you. And any any information any comments and any questions are very much welcome. So participant, you can raise your hand and I will ask you to unmute. You're unmuting me. Thanks. Hi, I'm Thad Pulaski. I'm from the Center for Resilience Cities and Landscapes at Columbia University. And I really enjoyed that presentation that was really enlightening. And the image capture software for the doors of the buildings is something that could be really useful in New York. We don't have a good idea of, first of all, like how many people live below the flood elevation? Like we know where their house when their houses are in the flood zone, but we don't know necessarily if their first floor is above that base flood elevation. We also don't know if people are living in basements. So this this image recognition software would be a great addition to this platform that the Center for New York City neighborhoods has already created called flood help New York, which intended to do this, but it didn't have the machine learning piece. It does like bring together the other flood risk data with like our municipal data about property. So just don't know where that first floor is. So that's great. But I did want to ask a question about the the the perception like community perception of a place and how you how how we can use that data to better understand community resilience. Because there's been a lot of debate, at least in my my working world, which is like some of the urban resilience people in the Rockefeller Foundation and other otherwise about measuring community resilience and how how it can be achieved because it's a lot of the a lot of the data points are very subjective. And then this perception of community health or community resilience or just safety in some cases, that sometimes that data is really hard to trust because perception makes up so much. I mean, bias often informs our perception, I guess, would be a simple way to say what I'm asking and how do you like correct for bias and how do you sort of distinguish perception, knowledge what we often need for like community conversations about resilience is like sort of a collective wisdom that isn't necessarily based on like, you know, my individual perception, but something that like, as a community, as a collective, we agree that like we need to get the drugs out or whatever, you know, so I'm going to stop there and let you respond. But thanks again so much for this wonderful presentation. Appreciate it. Thank you. Thank you for the wonderful questions and comments. And the first is about the first floor elevation estimation. So for that is, I will be very glad to share my other thoughts later is we are we are now largely implementing this kind of idea in Gaviston, Ireland, if you ever know Gaviston, Ireland is is a very frequently studied in the national literature for hurricane because it's Ireland is just near the Houston. So we use Gaviston Island as a test bed nowadays is using my technology to verify with the building information and also they also fly Jones also get information we are verifying these three different data sources together to see whether we can come up with some really inexpensive way to identify the to identify the lowest floor elevation. So I'm more than happy later can discuss with you is how we can use similar ideas in New York. I have some other projects also going on with New York is using New York's open data to understand how the business resilience of business to every disaster. For example, no matter is hurricane, but also like a public health crisis, how it will impact the business operation in local business operation in New York City. Okay, so that's one thing. A second thing is about the participation. I totally agree is if you ask people for their opinions, everyone will speak up of their interests. And then now some time sometime it's hard for us differentiate what is the interest of the community, what is the interest of individuals. So I think certainly when we develop the decision making tools, it's not help people to say it's not making decision for people. Instead is you can consider why we need a whole thing together is for let's say to help the decision makers or other experts to better understand what why there is so diverse thoughts about allocation. Previously, many of these information are stored in the meeting notes. Or you bring people in, they just talk and after talk, everyone forgot what other people is opinion is. So we for that is we want to link everyone's opinions to location, to the trajectory. And there's one other benefits is when you're putting people talking together. Nowadays, a natural language processing skills can also find out how much you repeat your opinions. How to what extent you go against yourself in the statement. Because previously in decision making, we use some using some very hard numbers, after you do some selection kind of quiz selection to find out how inconsistent a person is. If you are too inconsistent, then I need to disregard the person's opinion input. But for that is sometimes too arbitrary is because you only use several quiz or try to understand whether people can say no, no, you need to let people talk. So I so you know, so what we do is just let people to speak up as much as possible using the natural language processing to detect the inconsistency, use the natural language processing to to extreme their like a topic, their like a main arguments, then we go to overlay with other people. We strongly believe is if we can select a representative samples from the community, we and I believe we will be able to if they cannot find out their common concern, if they really represent something is commonly represented by the discussion, it will, it's really valuable. Right. So it's really will be very valuable. Thank you. Did I answer your question? Sorry. Yes, absolutely. That was a great answer. It's so interesting. I feel like we are in this world of convergence already. And so like the perceptions that are put into Google Maps or to like, you know, that are fed through social media are already informing and influencing our opinion. So in a way like how this translates in a positive way to the urban sphere is something, you know, I feel like it's already happening. But like maybe if we could design it so that it didn't live at the hands of the hands of like Google, but was like more or Amazon or whatever big tech, but if it could be like more of, you know, a citizen led initiative, I'm not sure what that means exactly, but because I do worry about that. Like I feel like maybe, you know, I like walk score or other attempts to sort of evaluate an environment based on community based on perception have resulted in like a kind of a reinforcement of some racist and otherwise like, you know, like historically informed biases or like, you know, like, um, and instead of, you know, the translation that often happens in the community engagement sphere, which is not taking people's perceptions, but oftentimes people's like aspirations. So I don't know. I think following what your question I today I today in the presentation, I forgot to mention another point, why this kind of well, why bring these things convergently convergence in location has another greater benefit is sometimes we, I as an urban planner, I visit the local community, we find out that some people are so knowledgeable about their local issues. I was so wonder is if the other community also have such a person, it will be great. But since we most people only live in their own community, in their lifetime, they stay in their local communities. So how wonderful this knowledge can be transferred to other similar communities. So I mean the similarity is when you have all this kind of data, you how we know similarities not only big picture or language is based on the essential knowledge graph behind that. If we find out that the two communities have 90% or 80% of the similar and one committee, for example, has experienced deal with snowstorm, the other community has never experienced that but one day suddenly because of extreme weather now have snowstorm, where the experience I can borrow from. So if I have a digital twin though it's not 100% the same, let's say it's a 90% the same in somewhere else. And let me quickly to borrow what they deal with snowstorm for me to make a I mean borrow the experience for this issue making. Thanks so much again. It's a great presentation and great responses. I appreciate very much. Thank you. Any other questions, comments? Well, if no one have a question, I will go with the question. So how do you see like the challenges from like scaling the like a project like a Galveston Island pilot project to a larger city like Houston or New York? Okay, so at least that's certainly a great question. We this year, as a senior personnel will receive the five million dollars of national transportation to build a high performance computing platform for emerging sciences. So because I wrote part of the need for do the coastal communities and resins modeling that's certainly is a priority one for the priority for this funded project for the deal with extremely large scale computing. I know many people will ask is when you deal with Ireland, then you have expanded to the metropolitan area. Even I will say how I will expand it to the whole gulf coast or the whole coastal area of the United States. It's suddenly is the purpose of high performance computing at the same time is the data structure thing. For today I showed as an example of our trajectory mapping by trajectory, our geo-visual research for that it has been funded by my three previous national science foundation projects. And for that we already developed some database structure and there's a media compression technology to deal with huge amounts of data because today I haven't mentioned one thing is when you send people to the field, then you ask them to talk, speak, and upload the video. The video will be size will be very large if you went to a place without too much internet connection, how you can easily upload these data. So we already solved this problem. Yes, so we have the technology. Thank you. Any other questions? Any participant have any question? You can just raise your hand and I will ask you to unmute yourself. One thing I want to add is the reason we use the Gaviston or Ireland as a case study is we want to use that island to test the convergence ideas because we want to use that to get the input from local people, officials, and various amounts of data together to build a very comprehensive digital copy of Gaviston and when it functions then the other cities can follow a similar way. But we do not value the computing capability too much because we already developed the data structure or functions very well. Yes, the only reason we use Gaviston as an example is we want to use that to test comprehensive digital twin. We make sure it functions then we use a much larger city and a much larger area to run the model. Yes, any other questions? If there's no more question then we will have like five minutes or around five minutes break and then we'll come back with the second groups of speakers. So now it's 10.05 and we'll come back at 10.10. If you have more questions please like text in the chat box or you can save it for the panel discussion that involves all three speakers. Thank you. Thank you very much for your great presentation. Thank you. All right, let's come back together and let me check if Waitroom has more to share this screen. Thank you so much, Professor Yi, for the presentation Q&A section. And thank you, thank all the questions we have. And now let me introduce our second and last group of speakers of today's forum. Professor Alan Smart is a professor in Mentors in the Department of Anthropology and Archeology at the University of Calgary, Canada. PhD in Social Anthropology from University of Toronto 1986. Research interests include political economy, housing, urban anthropology, Mercedes, etc. He has conducted field research in Hong Kong, the line China and Canada. He's the author of Making Drones, Scattered Clearance in Hong Kong, and also other publications. Professor Jim Coran is an associate professor of sociology at University of Calgary. He has previous degrees in economics, the philosophy of social science, and the PhD in sociology. His research areas include race, economic sociology, social theory, and inequalities. He also has many incredible publications in the British Journal of Sociology, Economy, and Society, etc. Professor Alan and Dean would present the topic data-driven governance, smart urbanism, and inequality in China, security, surveillance, and informatically. Let's welcome our speakers and I will stop sharing my screen now and yeah. Okay, can you see it? Yeah, yeah. Perfect. Okay. Okay, so as you thank you very much for the introduction. As you've mentioned, the title of our talk is data-driven governance, smart urbanism, and inequalities in China, security, surveillance, and informality. All right. So broadly what we're going to do in this paper, firstly, we're going to briefly discuss the rise of smart urbanism and data-driven governance in China. Secondly, we're going to propose risk class as a prism to evaluate these changes. Then I'm going to hand it over to my colleague Alan Smart, and we're going to talk specifically about some of the dynamics of risk class and digital governance, including impacts on security, surveillance, and informality. All right, so what we're seeing here over the last decade in particular is the rise of a new form of governance, which has been called data-driven governance, and this has emerged alongside with smart urbanism. And the current form of data-driven governance, I mean, of course, you study people like Scott seeing like the state and Ian Hacking, Foucault, there's a kind of guidance as well, this long idea of information as being fundamental to state governance and state control. But what we're seeing here is a kind of new form of this, specifically in terms of the use of big data and artificial intelligence. And this is particularly the case in China, lots of studies have shown, at least there's lots of research coming out, you know, some of it, a lot of it in newspapers, things like that, to show how much data is being produced in China, collected, and used to train algorithms. So just one example of this is that China uses 50 times more mobile money than the US does. And we can see with the FT graph right there, just the massive difference between the two. And mobile pay is a particularly significant source of big data information. It provides not only information on what people are purchasing, but it's also geolocated, which can in turn facilitate governance projects such as the training of cutting edge AI systems, such as those used in Alibaba as a city brain, which have become more and more popular, more and more used, not only in China, but in other places as well. So not only are we seeing data produced and collected in China, we're seeing also a subtle shift towards China in terms of maybe subtle is not the perfect actually significant shift in terms of the development of these algorithms in China by Chinese companies. So while much of the technology is produced by Western corporations, such as IBM and others, Google's also involves the usual suspects. Increasingly cutting edge technology is always also being developed by Chinese companies such as Alibaba, Tencent, E2, and financial, and I should mention Huawei as well. So the centrality of big data to machine learning forms of AI is giving China a significant advantage because of the larger size of the population, much greater adoption of mobile phone shopping, and at least until very recently, less restrictions on privacy and use of data by government and corporations. And we know what the new privacy regulations coming out in China, and we'll have to see how these work in practice. But nevertheless, it's been a massive advantage for the amount of data that's been able to be collected in China in terms of pushing forward on data during governance and smart urbanism. So one of the things we're proposing here, and I definitely, of course, all of this is co-authored between Alan and I, but I think in particular, Alan's been really important here in thinking about this, is the idea of provincializing Western smart cities. There's a long running tradition, which you can be traced back to Mark's, among others, that the West is basically the most technologically developed, which in the 19th century in that context, you can see how Mark's would be seeing that in terms of the development of the capitalist mode of production. And the idea here is that what we still see as the kind of lingering effect of this is a lot of focus on Western smart cities, and in particular, a lot of focus on Western smart cities as at the cutting edge. And I mean, there are a couple kind of very famous cases from Asia that are used, but what we're suggesting here is really given how much of the cutting edge China is getting in terms of the employment of smart urban technologies and big data and AI that we might want to see it the other way around, that there might be a kind of use in understanding what kind of future for the entire world are Chinese cities generating. Okay, so in terms of smart city projects, again, this is very high level, very brief summary. Already by 2015, there were 311 Chinese cities implementing smart city plans with claims that already 158 of them have been built. Admittedly, and this is important in all the things we're saying, I think it's important to identify the new, acknowledge the new and the novel, but often those things that are new or novel are overlaid on top of existing, they never made kind of ex nihilo out of nothing, they're generally overlaid on top of existing structures. So admittedly, many of these described smart cities were kind of token or rebranding of other initiatives such as eco cities, digital cities, creative cities. Nevertheless, though, if we go beyond the formal plans and consider data driven governance more generally, which is not only instantiated in the city, but more generally in any kind of state governance, China has probably moved faster than anywhere else in the world. And again, this ties us back to the issue of big data. So of course, we've, you know, heard this oft said analogy in the last few years that big data is the kind of fossil fuel of the 21st century. In particular, what we've seen since 2016 is Alibaba has been slurping up video feeds, social media data, traffic information, and other data for example from Hanjo for its city brain project. And not only have they been developing this project within cities, but they're starting to export this technology, not only to Macau, but to Kuala Lumpur and potentially some other cities as well. And so what we're seeing here is a kind of scaling up as an exportable urban governance package with a kind of homogeneity of the basic framework, but then value added customization. And here is one example of the kind of view of the city that can be provided by these types of technologies, building on a massive amount of data to simplify to provide a few key categories. And this is again, city brain, which has become extremely, extremely popular. I mean, one recent article said that, you know, almost every city official is trying to get a city brain within their city in China. So as I've mentioned already before, privacy has so far been a huge competitive advantage. In terms of weaker constraints on government and corporate use of personal information, the new kind of consumer privacy rules, it's still very early to tell, but again, these don't put constraint on government use of data. So generally, it's still a kind of significant advantage over the West, especially when you compare it to projects like Sidewalk Labs in Toronto. I mean, ultimately, the whole project just ground to a halt because of concerns over privacy and rights of revolving around privacy. Okay, so that's a kind of, you know, brief overview. One of the things that we're trying to propose here is how to think about the new forms of inequality that are emerging from this, how to provide a kind of categorization or kind of how to make them more intelligible and to be able to, and how to make them commensurable as well. I mean, there's so many different processes going on, so many different effects. How do you try to make them, you know, how to try to make these understandable? So one of the things we're proposing here is this concept of risk class. So the concept came out of a debate that I had with Ulrich Beck in 2013. And so it basically builds on Beck's idea that risks, like wealth, are the object of distributions and both constitute positions, risk positions and class positions respectively. So this is from Beck's, not his first book, but definitely his kind of his classic work, risk society, which was written, published in Germany in 1986, just so the manuscript was submitted before Chernobyl, but came out just after Chernobyl emerged. And for many people, the kind of Chernobyl accident crisis was a real wake-up call that, you know, capitalist modernity, technological development wasn't only producing goods, but really producing the potential for significant risks as well. So in thinking about this idea of a risk position, we need to think theoretically and empirically about how do we generate this idea of a risk position? And so what it involves is needing to integrate a multitude of socially produced risks into a systemic social position of disadvantage or advantage. And the key here is that despite lacking the kind of commensurability that economic goods and in particular, exchange value generally generates. So, you know, when it comes to distribution of goods, the primary reason why we can make them commensurable and kind of give a general class position is because they're tied to exchange, they're tied to markets and the more valuable something is, the more we can generally consider that value. I mean, of course, the distinction between exchange value and use value is significant, but nevertheless, there's some kind of mechanism of commensurability. As we know, there are extremely imperfect markets on, if there are any markets at all, on externalities and on risks. So is it kind of both a theoretical challenge and an empirical challenge? But as Amartya Sen has said previously, you know, we need to identify the core processes going on here first rather than just looking to build frameworks on what the existing data provides, right? And part of what we're trying to do here with bringing class position and risk position together is to address the systematic nature of knowledge, the production, but the distribution of risk in modern societies. So you almost have like this double economy, the economy that produces goods. And at the same time, an economy that produces and distributes risks. So to be clear here, while the framework built on one of Beck's claims, Beck actually kind of undermined the potential insight of this work by arguing that that basically what would happen in the risk society is that the risks would be significant enough that inequalities would decline. So basically, what we have here, you know, one of Beck's famous quotes is some poverty, poverty is hierarchical risks are democratic. And the argument that I made in the debate with Beck is that while his model or basic framework definitely can illuminate inequalities, he employed an overly catastrophic account of risks. And this overly catastrophic account of risks just neglected the way in which even if things are getting worse for everyone, they often get worse for everyone in a highly gradational way, a highly unequal way. So the concept of risk class is meant to explore the interactions between class position and risk position. And to be clear, this is not a kind of, I'm not trying to reduce class to risk or risk to class, but rather to explore the interaction between the two. And so in my work, I've developed some kind of toolbox of concepts for this. A lot of the work's been done in finance and environment, risk arbitrage, organized irresponsibility. And one of the key concepts here, I think, that I constantly keep in mind is this idea of private escape routes for risk. And now I was trying to think about a kind of very concrete and spatial way. I mean, this is an urban forum to make sense of this. I think the picture, this is, of course, the picture of my book, I'm sorry, it's a picture of New Orleans, which I put on the front of my book. But I was reading the Financial Times yesterday. And I think one of the articles by Simon Cooper, I think gave a great example of a kind of urban, the spatial inequalities that you see was private escape routes. So I'm just going to read you briefly this. This is about Miami. So he says here the phrase climate, I just saw this last night. So I didn't put it in the slides yet, but forgive me the brief interlude. The phrase climate gentrification was popularized by a Harvard study of Miami house prices in 2018. The researchers found that price rises in higher lying areas, such as little Haiti, had outpaced those and lower ones like Miami Beach since about 2000. So again, this fits with many kind of stories, which is the salience of risk has continued to increase since the 80s. And you kind of see these tipping points since 2000. Back to Cooper. This is climate change hastening the gentrification of poor neighborhoods. Miami real estate agents used to ignore rising sea levels, then pretended the problem had been fixed by what looked like waste high garden walls. But lately, and this is the key element here, if you're thinking for analogies to think about spatial qualities, but lately Miami market thinking is shifting from, quote, location, location, location to, quote, elevation, elevation, elevation. He also says I also liked sell low by high, which is a little cheeky, but we'll forgive him there. And this kind of idea that elevation, having elevated areas of land in high risk areas, I mean, this can be traced back to a whizner al's well known book at risk. And I talked a little bit about this in, in Andrew per dash. What we're seeing here, this is becoming a more general process that people are starting to be the second those with economic power are aware of these risks, and they become salient to them. They can use their superior wealth and purchasing power to outbid others for these more secure social positions. And so the analogy you can almost think of as a, as a kind of auction, where if you have a limited number of secure positions, these generally can be bid for those with superior economic power. So, and of course, we, we are well aware of the kind of inequalities that come with climate change in terms of those who produce climate change and those who bear the costs. A lot of the research has focused on global inequalities, which is reasonable and significant, but the inequalities within countries, which is highly correlated with class inequalities are in fact, just as significant or in many cases much more significant in terms of the ratios as, as international inequalities. And so the kind of economic elite are not only an economic leap, they're also a risk class and that they can produce risks, and then avoid them. So three key goals here in thinking about risk class analysis, first of all, to explain existing identity, identified equalities in new ways. Secondly, to extend the scope of systemic inequalities in new domains and to provide novel basis for normative critique of these inequalities. And, and Alan Smartnove, I have already, I mean, started to try to do this. I mean, admittedly, when you developing new frameworks, often the, the data just isn't there yet, and you have to work with what you have. But in our recent urban studies paper, we tried to work on through some of these themes on relating to social credit in China. I've published a recent paper on oil and gas as well. It looks more specifically at climate change. And so, you know, this is still a kind of emerging research program. Okay, so bringing it, you know, bringing it back to the very specific topic we're looking today, which is risk digital governance. Of course, this is Zuboff's very famous book on surveillance capitalism. What we're looking here is how things like surveillance and security that are involved in new forms of data driven governance, and the shift towards formality, how those are increasing the advantages of those who already have security and social legitimacy. And almost you could say kind of dumping risk on others. And as I've mentioned before, again, this is really well known book by Virginia Eubanks, Automating and Equality. So I mean, there is existing research, there's definitely a lot of existing research on digital risk and inequalities, but bringing the kind of risk class framework is still emerging. Often these inequalities are layered on top of the existing configurations of private and public. And as always, the state is always mediating them. So, and that's another reason to have the kind of specificity of analysis when we're dealing with China and not just bring kind of existing Western approaches to it because the specificity of the type of Chinese state that not only has emerged since the 1970s, but that is continually being remade, especially in the last decade. And how we think about these things in particular when it comes to the role and impact of algorithmic governance. Okay, I'm going to switch over to Alan now. Thank you. Okay, thank you, Dean. And thank you for the organizers for inviting us to participate in this forum. So just to follow on from what Dean's been saying, I think the process of kind of facial recognition that's going on really illustrates the kind of things that Dean was talking about so that China has built already by 2016 a really massive data facial recognition database of 1.8 billion individuals at that time. You might wonder why the number is more than the total population of China. And that's because anyone that passes through a Chinese airport gets added to the database. So you're so that's accounts for that. And so that you have the and the AI that's been developed makes it possible really to search this almost unimaginably big database in just a matter of seconds and recognize things there. Of course, you might ask what's going on with COVID and everybody wearing masks. So there's a lot of work being done on that, but also moving beyond faces to what's called gate analysis, the way you walk and the way you move, which apparently it's very surprisingly difficult to disguise your gate. And so there's all kinds of other ways that are going on. So increasingly Chinese police are equipped with database connected cameras allow them to scan faces and identify suspects. And there's been a number of notable cases where people going to a football game or a concert or whatever have been arrested as they go through the scan scanning system. What's surprising is that those cases aren't in the much larger numbers. So there's again, as Dean said, it'd be great to have more information. So we're kind of looking in from the outside of what's going on. So the under implementation of that kind of tool of policing is something that's worth recognizing. And the other thing to kind of qualify this is that we shouldn't get carried away with just focusing on the technology because the surveillance system in China builds on paper file technologies that took form in the 1950s as the household registration system, the Hukou. Just as you banks points out that the automating inequalities in the US builds on all kinds of other analog technologies of surveilling families through child and youth protective services and things like that. So these are not new, but they're really adding on. And as you banks points out for the US, the digitizing and automating of these kinds of systems kind of makes it much harder for people to avoid the risk. So the risk gets very much amplified on those who are already disadvantaged. Next slide, Dean. So one of the things that as Dean points out that a lot of this is really under recognized in the mainstream smart cities literature. And things are moving very, very quickly. And you have to kind of pay attention. So if you just kind of search smart cities, you might not come across what's been going on, particularly with Huawei's version of database governance, governance, so that they're promoting what they call safe cities. And so safe cities and smart cities are starting to merge in some significant ways as Iona data has documented for India. And the smart cities are particularly being promoted by Huawei and some other companies along the digital Silk Road, which is extremely significant because of the kind of global geopolitical conflict between the America and China and creating kind of divided digital systems. And increasingly the Belt and Road Initiative is creating an alternative digital governance system for those cities heavily involved with the BRI. Next slide, please. So in Pakistan is a good example of that. So I got involved in this smart city research not directly, but because through informality. And so I want to explain a couple of things, why we're putting informality because it seems like what do these two have to do with each other. And in Professor Yaz, really interesting talk, he talked about data challenging communities. And I think informality is basically the core element of those data challenging communities. So I came to the question, well, what is what is smart about a smart city? And there's a huge debate about the definitions. And largely it has come down to an emphasis on the density of information and AI technology. So it's basically a high tech assumption about that being smart. So smart becomes the technology. But it really begs the question of what makes the city smart. Making traffic flow smoothly in a sprawling auto dependent urban region is a pretty limited conceptualization of smartness. And talking about auto, you know, self-driving cars, the worry is that, you know, there's really smart technology, but a lot of people live even farther out so they can nap or do work in their car while the car does the automation. So the point is that it's really easy to do very stupid things with very smart technology. And the best example of that is the 2008 financial crisis, which was based on some really highly paid, very smart people using cutting edge technologies of algorithms for or, uh, securitizing, uh, mortgages and things like that and ended up really collapsing in large part, the global political economy. So just assuming that because you have smart people using great technology, uh, is going to make a smart city really neglects what is smart. Next, please. So that's where informality comes in. So the assumption really behind those who intervene in urban governments is that the informal is backward, it's problematic, it should be either reformed or usually bulldozed and replaced with something modern or in the terminology often used in China, civilized. And so one of my first work within smart cities was really to ask whether formalizing informal practices like urban villages in China or illegal street vending and things like that makes the city smarter or not. And it's a very big question, which is almost completely neglected by mainstream urban studies researchers because globally, at least 50% of all work in the world is done informally. And if we count non paid work like domestic labor within houses, it's probably more like 70%. In places like Sub-Saharan Africa and South Asia, even paid work is about 80% informal rather than formal. So the question rises for me is if smart city projects displace or erase informality, they'll have huge impacts. But will this in fact increase urban intelligence? If not, those displacements may have extremely negative outcomes. Next, please. And so a lot of studies show that informal practices are better than formal institutions at meeting the real needs of citizens, particularly but not exclusively in the global south. So if you have a kind of a risk polarized system that mostly benefits the middle class and up, informal practices are often crucial in allowing those at the bottom end of the pyramid to be able to cope and survive. If that's the case, and I'm convinced it is, then strategies that undermine informal economies make the livelihoods of the majority of the city dwellers worse rather than better. But as I said, most proponents of smart cities strategies either neglect informality or see it as an obstacle to an efficient and sustainable city. Next. So there's a big divergence between kind of work on smart cities in the global north and the global south. In the north, they mostly ignore informality altogether. In the global south, it's often paid some attention, but usually as examples of the things that have to be removed to make a smarter city. India particularly has taken this kind of approach. So you see more references to informality, but they're generally very negative. Okay. And so in China, you get a little bit more mention of it, but still it's probably more in the explicit smart city strategies like the global north cities where it tends to be assumed neglected rather than talked about. But if you look at not just what we call explicit smart city strategies, but the implicit smart city strategies that are part of the broader big data based governance, what we see is that China is very much caught up in the idea that informality has to be replaced and that things like urban villages are backward, they're problematic, and they need to be replaced with something shiny and futuristic and modern. Next. So that implicit in China's projected smart urban futures is a lens that really sees certain places and people as parts of the past to be excluded, perhaps reformed, but largely removed and replaced. So that in much of the world in what's called elsewhere, squatter areas, but it's kind of maps on somewhat loosely to urban villages, there are lots of attempts at what we call it in situ upgrading to try and make it improved often kind of more user friendly than the public housing projects that are built on the fringes of the city. And I point out here that if you google smart slums, you find almost no references. And yet there's no reason why we shouldn't be using the same technologies to optimize resilience and other kinds of processes of mitigating risk in low income areas instead of just assuming that they are part of the past eventually going to be replaced. So my assumption and argument in our work is really that what's going on is that hundreds of millions of urban Chinese are still heavily reliant on informality and smart urbanism biased against these informal practices could be and largely is being immensely destructive. It's good that the resettlement projects from urban villages tend to be quite inclusive of at least owners of property within urban villages often discriminating against renters who are usually migrants, but you know so there's good housing being produced usually on a more fringe area and a lot of the residents are happy about that, but it is really ending a lot of the kind of urban vitality of the informal institutions within the city that old parts of the city or old parts of the fringe that have been incorporated in the city and so on. Next please. So I just want to conclude and Dean may have a couple of comments afterwards is that coming back to this idea of provincializing smart city research is that in the context of mainstream smart city research a lot of what's going on in China is really arguably kind of catching up if not equal to the sophistication of the technology in the West, but that it's also being implemented at a much faster rate kind of at Shenzhen speed of urban development and that with the kind of trade wars and sanction wars between Washington and Beijing what's happening is the world is kind of big data-based governance systems are bifurcating they're diverging in different ways and that the Belt and Road Initiative and the Digital Silk Road is actually presenting a kind of a new model for digital governance and that largely it's so certainly I'm not saying that there isn't publications on Chinese smart cities in the good journals but they tend to be specialist articles they're just about China and then when you look at the broader arguments that are kind of unmarked they're talking about smart cities. China is at best a footnote and often not mentioned at all so there's a real need to point out how provincial western smart city research is and they're neglecting some really big things going on and I think the other this is a kind of an older image I use but the most recent numbers I've got is that Huawei has over 200 safe city projects which involve basically at a core of it is cameras linked to databases linked to the police forces so smart safe city has become essentially a secure and surveilled city presumably largely for the benefit of those who are already privileged so the consequences of the types of smart urbanism that are being developed in the context of the Belt and Road Initiative I think are really of kind of world-shaping consequences and we really need to pay a lot more attention to what's going on so thank you I'm done excellent yeah I'm good thanks Ellen very fascinating thank you so we're going to start the Q&A section so if anyone wants to ask a question or participant please raise your hand I will I will ask to any of you yeah thank you that's a fantastic speech and I really learned a lot but I was kind of confusing because I feel like sometimes equity and say like safety are like contrary to each other even that's to say in the topic of smart city even in our like normal urban planning we it's really difficult to get some of some of groups like involved into some kinds of planning especially for smart cities some kinds of like advanced technologies may be applied like say we have some groups of elderly people it's really difficult to get them to learn how to actively use those advanced technologies to involve in either in like community planning process or in decision making but we are trying to still use those smart city the idea of smart city to apply to like all of our residents like equally and like no bias on any kind of choice um that makes me kind of thinking is that possible or when the idea of smart city was raised is it possible to like we got some group of people so first get involved then we can use the technology upgrades or like after like 10 or 12 years we got like either mode to like the face recognition to them like let those residents to passively get involved into those smart city ideas instead of like ask them to actively enjoy the technology okay why don't I jump jump in then so yeah it's a really good question and it's definitely a challenge as you can tell I'm pretty gray and I'm definitely not that the cutting edge of use of technologies tend not to be on social media very much because my email does everything I need so but so the question is how you end up with what the Europeans kind of emphasizing is smart citizenship and it is extremely difficult to get people to participate but I think that is often the way in which the urban planning participation system operates that even in North America there's a kind of a cynical attitude that this is all a token exercise that yes you can express your opinions but is not really going to change what the urban planners are doing so that if you're busy and you have to take the time to read documents download PDFs and find a way to translate your way of thinking in your concerns into language that fits with what the planners have in mind then you're probably not going to bother to do it particularly if you don't think that anyone's actually really going to listen what you say so I think it's necessary to find ways to make it easy to participate and in fact that's what all AI is theoretically about it's about big data it's about being able to include the concerns and issues of all of the citizens not just a handful of elites or at least it should be feasible because you're getting the technology you're getting the processing capacity in order to use information about what people are doing as opposed to what they're willing to say in formal participation activities organized by governmental agencies so I think we need to find a way to implement it but one thing that is universal is if your community is going to be demolished people have things to say so there are certain decision points at which it would be a very good idea to use the technologies that people are familiar with like we chat in China and so on to try and tap into that and urban planners come from a particular background and they're probably not necessarily going to automatically be aware of the concerns and this is there's a long history of this before smart urbanism of resettlement projects that neglect what the real priorities of the urban poor are and there's great work done way back from the early 70s by the architect John Turner that shows that things like amenities that urban planners tend to emphasize and kind of modern looking buildings are much less important than central proximity to where the jobs are and things like that so there are there are there are ways of incorporating the ideas and concerns and interests of more marginalized communities but we're not very good at taking advantage of them probably because we're not that interested in hearing things that don't fit our our projects yeah i i think it's an excellent answer i think and part of the issue here so first of all any new intervention is is going to have differential impacts i mean the only way it wouldn't is if the people were the same that's it like any new intervention is going to have differential impacts and the hope here is to reduce some of the inequalities right i mean part of in a publicist paper in antipathy in 2018 and one of the things i talked about was what is the relationship between injustice and inequality so is is every form of inequality unjust and it's actually a really really sticky matter like conceptually right emotionally we might feel it but when you start looking at it it's very hard we know there's a connection between these two but what is the connection so my proposal was let's look at relational inequalities because we know those are problematic and what a relational inequality is when the advantages of one it's not like everyone's gaining but some are gaining more than others it's where the advantage of one is to the cost of right and that that to me kind of cuts through the the gordian knot because we know those are problematic we know that we know and especially when we're adding new processes that are making the already advantaged more advantage at the cost of the least advantage right and i i think one of the i've been slowly reading Piketty's new book he doesn't write small books i don't know if anyone's noticed this i think he needs to hire someone to abridge his books but i think one of the things he's really pushed on and i think it's quite an important reflex of critique for academics and some recent french scholars have started to point out how the kind of merchant elite and the kind of cultural academic left um actually share many common like we we feud but we also in many ways share a common project of of excluding others from the basically from the governance of the world and and one of the issues here i think smart citizens enabling people to have an impact and a say in how the society works is absolutely fundamental and i think allen's work on informality that's a huge part of it right but sometimes when i see conceptions of smart citizenship developed they're very much developed by people who um are so immersed in the gown that they they don't really they're not really interacting with different cultures whether class cultures you know ethnic communities and and i think it's a real challenge to think that we can find one framework to integrate all of these different ways of communicating and i think that's one of the issue with any kind of platformatization of of of communication of information is we kind of either has to reduce it to this you know to speak to the algorithm but a lot of people just don't work i mean every time there's a poll online it's people like myself you know it's people highly educated people in universities who are vastly disproportionately participating right and i i think even what we conceive of as smart citizenship has to be rethought in a way i think they can that can take much more seriously the informality and the time scarcity of people you know single parents um you know people who are juggling i mean i feel busy but i i'm not even close to being really busy if you know what i mean and so i think it's a great question and i think the most important thing is to keep this puzzle in mind and not to pretend we've solved it in a simple way you know what i mean to can and and not to think somehow because when we haven't totally solved it that we can't make progress on it right so thank you thank you thank you for a comprehensive answer i really enjoy the discussion of info as an informal building environment or informal urban activities versus discussion of smart cities it reminded me many other discussions on artificial intelligence side is a talk about the the how to say the social how to say justice issues and bias in artificial intelligence because we in AI or computer vision we have quite a few libraries to label the urban images as safe or not safe or happy happy or not happy or they have many criteria i would question who label these images have those people lived in these something they call it informal human building environment really i mean live there so so that the problem is for example i have visited quite a few the cities along the border of us and mexico border many of these cities are very poor but their crime rate is low the reason is the generation living in the same neighborhood they have a family owner so they were i mean actually the crime rate is low they're pretty safe however if you just based on the looking of the neighborhood so you you might think it is a high crime rate neighborhood but actually it's not so the issue will make us think is do we need a smart city technology to change all these neighborhoods do you do you have the job opportunities to afford those people to remodel all these buildings according to the need of technology apparently it's not so i totally agree is definitely through the we need a more collective intelligence from artificial intelligence as well as human centered urban planning theories help us to make our cities the more as a living structure instead of it's a pile compiled by technology so what is the most needed is by local people the interesting things for many times we design a city we are not the people who live in that city or community so that's the reason we do need some platform a computing platform allows us to collect more thoughts from people no matter is a very so that's the reason we need to use natural language using all kinds of the signals from them how they walk around when they walk how they talk how they perceive what they say all kinds of things so we will see the living is a we treat the city them as a living structure instead of a mechanical machine so yeah so i that's my reflection although there's a wonderful presentation thank you well thank you yeah i totally i totally agree and it's good to hear examples like the us-mexico border where there's been some really interesting research about you know the as you said low crime but also very poorly serviced communities that are often very much stigmatized and misunderstood from the outside and that's certainly something that is pretty common throughout the system my my doctoral research was on squatter areas in in hong kong so we were living there and one of the things that was clear was that kind of the middle class assumed that they were dangerous places and a survey that was done by NGO found that not only was objectively the crime rate quite low but that the perception of the people living there was that it was i think one third thought it was safer than the city in general and about half thought it was about as as as safe as the rest of the city and of course hong kong is a low crime rate city in the first place so uh it was uh you know there were lots of problems but those problems were really ones that the community couldn't really do with uh deal with by themselves like sewage and so on yeah um i raised my head but i will speak to it i have a question for um generally both of our speakers is that uh it's really interesting and really fun to see that the uh the visualization part for professor years point is because we um we think that we have seen this place before and we want to search for this place and it seems like for me the user the potential users can be um for example the young people and a more broader user group but on the other hand i think the technology productions are really expensive to afford so i kind of wondering like what's the mainly users now so that's do they facing the affordability of the product issues or um what's the uh like the future roads if the technology is getting more and more easy for like larger publics to use very well uh okay for thank you for your question and due to the time limit today i didn't show that we developed this data collection device uh through cell phone app we already make it developed the free app called joe app which now is if you search either in iphone or android uh there is free for download and for that is for any people who have a smartphone you can use that to go around in your neighborhood to talk and if your the your voice will be turned into transcribed link with your uh trajectories and we also have our free to use software allow you to upload the uh data uh upload the data to the uh the open or open tools now for that is to facilitate the people's data input you don't you can just speak or and we also developed a new tool is allow people just use a finger touch finger to access our visualization uh platform so so as if is if you can drive and say you don't need use a computer the keyboards you just say as so if you can play with a digital device you will be able to do that but as you said as another concern is if for people who do not use a digital device at all um at this time our we work with uh some we begin to work with some local communities in texas we do help people who do not use any digital device then we have our student our uh volunteer social workers to work with them because that's another of my long term uh plan is i really want to bridge our international students with american communities because for many of them after they came to us what they realized is for example in texas enm we are in a city or college station it's a typical college town for many of them after four years or five years of phd study what their understanding of us society is a college town or they definitely sometime will travel to the big and fancy cities for tour but they never know the other small cities or rural town in the us how they function so i'm through these ways how we can match the students with the local communities and so they can learn from each other the life experience of from local people from local decision makers through this digital device can be a commuting medium between them so i i'm thinking in small cities or any kind of academic research is very much is to train people to be a citizen to be a collaborative collaborative citizen just to follow up on that so that just give an example of the way in which ordinary people without much education have been collaborating in dealing with some of the issues that professor yeah is talking about particularly in uh africa there's been a lot of effort put in by ngos to fill in the invisible spots of african cities because very large parts of african cities are basically absent from google maps because there's no official streets there are just pathways and so that there's uh efforts to incorporate people not just to put you know you could you the technology will allow you basically to follow a path and describe the path and add it into a uh collaborative mapping mapping software and it's fairly simple with it as long as you have a smartphone apparently but then you as uh professor yes you can also add in annotations and describe the place and what are the uh key features and the kind of all the kinds of things that uh are often usually available on uh google maps but are completely absent in parts of the cities that don't have mailing addresses so there's also some interesting work being done on providing universal addresses for places so that you can start to get deliveries which would be useful during the pandemic and things like that so there's there is work being done but it's still it's the funding of it is marginal compared to what's going on for uh obviously uh profitable corporate activities one thing i want to add is map is not a territory even we use a for example we have google map we print a google map it still is not sufficient for a community to use so i work with my students saying yes in that for example it's a neighborhood you can print a google map if you can print a print of google map by the way prefer with open street map we can do more more programming on that can we add the places where the elderly old adults might easily fall down focus on some location because google map or open street map will not put these marks for you but it will be very useful for seniors in the community to walk around and for that is definitely some computer vision all the people's experience can oh yeah so pay attention there is a place it's very easily you will fall down right so we mark these places out or you can mark places for like in oh i think in many states we have safe to school program or seeing how we can i mean kids can go to school safely then you also need to avoid there is certain like traffic issues or some sidewalk broken sidewalk for that no no google maps or open street map will give you such information it's very much from the citizen cloud source the information so this is also how to this is a really good way to make our how make our community much safer much warmer much welcome for local residents yeah and i just wanted to build on one of the things alan said i mean we had the slide about surveillance capitalism and we haven't really talked a lot about capitalism in this in our talk we talked a bit about the state but it's any of these developments are always mediated not only by the state but also by the type of economic organization you have and globally surveillance capitalism is dominant and so if you look at the amount of resources that are devoted to things that are expected to i would say yield profit or in fact i mean often the goal isn't necessarily profit in the short term it's a net present value so companies will often incur losses in the short time to engage in acid inflation right and but things that are not anticipated to provide profit in the mezzo term or at least economic i mean amazon still doesn't make a lot of profits because their focus is on building their market share and and even things that start out sweet i mean just the whole digital economy is just filled with the bait and switch the second you think i mean what's up is the perfect example of this right the way facebook and i'm not saying what's up is a perfect tool but it was shifted from being something in private between individuals to being linked to the facebook instagram ecology right so i do think there there are important things going on there but they're often they don't have the weight of resources behind them that they definitely could right and if you think about the amount of wealth and that's being created from this data very little of it is being devoted to these kind of issues right thank you so much yeah thank you for answer this questions um i received some like private chat messages about with more questions about with regards to um urban data and it's like particularly urban governance so um so one of the participants is addressing the like the part you mentioned that all of tech companies are in china are applying um smart cities um governance technologies um in chinese cities so this participant mentioned that there's another online service company called DD which is the online service like real cofa hire service company that went public in new york and after its IPO the company app was forced out the mobile stores so the question is how the city travel data sort of become kind of a sensitive or something threatening to the administration um that's your thoughts on that now and you've done more i mean this is part of the battle between the two uh digital giants right yeah um so i think DD is one of the examples of of something that we're writing about right now and that's the way in which uh you know i've mentioned the kind of the fight between washington and beijing but uh the tech giants in china are increasingly under pressure not just from washington's sanctions and uh blacklisting and so on but also on the tech lash the technology backlash from beijing itself and they crack down on uh the free ranging activities of ant financial and alibaba and tens and and and so on so that i think this is the kind of the divergence that we're talking about that essentially the world is seems to be on a path to having two very different digital governance ecosystems one based in china one based in uh the the us and that uh the uh increasingly the technology giants in china are having to kind of uh rely much more on the favor and support of beijing so it's kind of supporting a much stronger state capitalist rather than free market uh approach to digital governance and so that the companies that aren't playing along like dd are being punished so that the the models are increasingly coming more directly from beijing and i think that the new privacy law data privacy law is a big part of that and as dean said earlier how that's going to play out in practice will be fascinating to watch but i think there's even compared to when we wrote the urban studies paper you know things are uh completely changed in in just two years in terms of what's going on in the data ecosystems in in china so i'm not sure if i have an answer to that question but i think it's a really good question thank you um is there any other comments um if not i have another like also another private message um both questions so um in still like it's all also about like the informality and the like urban villages we talked about in the um in the presentation so you seems to discuss a lot of policies from the chinese government is that to to replace those informal areas with more shiny quote-unquote shiny urban areas and you mentioned that people will be left out of this process and it could be better if citizens like informal in the informal sectors could be could apply those smart technology and could they sort of utilize those smart technology um the public this person is curious if you could talk about this like the inclusion process um in this like transformation or the like the transformation of technology from the more formal sector to the informal sector okay yeah so uh i think when we're talking about informal settlements where you have a fixed space and it's already being developed informally it's it's it's a challenge to incorporate uh uh infrastructure but in some ways the smart infrastructure is much easier to implement implement than say things like sewer systems improving the uh street pattern and so there's lots of uh ways in which one could take uh you know basically all you really need is uh wireless towers and the right kind of technology to for example give everybody their uh own address to give access to uh uh things but so far most of that that's some of that is happening but it's being used often used in a way which is exclusionary rather than uh collaborative so for example a lot of the surveillance technology is being done to try and ensure that uh people have to have a hook go to live in a hotel or to stay in a hotel or to rent an apartment and so that the african population in guangzhou for example has shrunk quite dramatically because it's very difficult for them even if they get a visa to find a place to stay because they're excluded from large parts of the accommodation possibilities so as long as you have a system that is selectively inclusive for example benefiting the original villagers in an urban village as opposed to the usually the uh large majority of the population are migrants with only temporary hook go uh usually they are not taken into consideration very much in resettlement plans in improvement plans so the original residents are the ones who are kind of seen as the those to be included there is there is some uh moves i've just been read a manuscript for a book manuscript on urban villagers and with the new emphasis on a community was a community of shared prosperity the new mantra from xijing ping there is more effort to equalize access to public goods for both migrants and original villagers but as long as you have a kind of a strong exclusionary attitude that these are temporary people and temporary places i don't see an awful lot being done to kind of smart implement smart technology but you know one will see but the trajectory seems to in china seems to be mostly replacement and resettlement and i would say one of the biggest areas of inequality that needs to be addressed is dealing with tenants rather than property owners within the urban villages thank you um thank you for answering my questions um any other more questions all right uh so i guess i have a hand back to danching yeah i will share the screen today's topic are so fun and thank you all for coming here and also thank you for our audience here it's been so great to have all our speakers talking about the sort of for me the very insightful and the connection between the technology with equity data the climate resilience except for all the heated topics uh nowadays and um yeah thank you so much and tomorrow we will have two more speakers uh Mr. Sun Tao and Mr. Yang Sun Tu and Mr. Yang Tao and they will present us with the topic smart city business and technology in internet companies and also the topic of the practice of city information information modeling platform in china yeah and thank you so much for today it's been a long forum here and thank you for staying here and hope to see you um the audience tomorrow and remember that there's a time changing because our winter time so yeah we will also make announcements yeah thank you so much thank you so much thank you for the organization thank you thank you to the whole sand professor thank you thank you thank you yeah bye bye we'll see you right now yeah bye