← New Books Network
New Books Network
Xiao Huang et. al, "GeoAI and Human Geography: The Dawn of a New Spatial Intelligence Era" (Springer, 2025)
00:29:25
Loading summary
Transcript35 lines
- [00:00]
A
Hello everybody. This is Marshall Po. I'm the founder and editor of the New Books Network. And if you're listening to this, you know that the NBN is the largest academic podcast network in the world. We reach a worldwide audience of 2 million people. You may have a podcast or you may be thinking about starting a podcast. As you probably know, there are challenges basically of two kinds. One is technical. There are things you have to know in order to get your podcast produced and distributed. And the second is, and this is the biggest problem, you need to get an audience. Building an audience in podcasting is the hardest thing to do today. With this in mind, we at the NBM have started a service called NBN Productions. What we do is help you create a podcast, produce your podcast, distribute your podcast, and we host your podcast. Most importantly, what we do is we distribute your podcast to the NBN audience. We've done this many times with many academic podcasts and we would like to help you. If you would be interested in talking to us about how we can help you with your podcast, please contact us. Just go to the front page of the New Books Network and you will see a link to NBN Productions. Click that, fill out the form and we can talk. Welcome to the New Books Network.
- [01:07]
B
Hello everyone. Welcome to New Books Network. I'm your host, Devika Jain. In this episode we dive into the book GEOAI and Human the dawn of New Spatial Intelligence Era. This book explores how geographers and data scientists are fusing artificial intelligence with spatial thinking to address some of society's most pressing challenges. Featuring insights from leading contributors, the book examines how GOAI can help us map complex societal and environmental systems, why data justice matters in AI driven world and what the future of spatial intelligence might look like. This book has received 10,000 downloads in just one month on Springer Nature. I'm very delighted to be joined today by Dr. Xia Huang, one of the editors of this volume. Dr. Xia Huang is an assistant professor in the Department of Environmental Science at Emory University. His research spans human environment interaction, urban informatics, disaster mitigation and Geoai. Dr. Hoang has authored over 200 peer reviewed articles and more than 20 book chapters. He's listed among the world's top 2% of scientists by Stanford Elsevier Ranking. His research has received coverage in many media outlets and has attracted funding from several top organizations. Hello Dr. Huang, thank you so much for joining us today to discuss your new book. So first I would like to start with what inspired you to put together GEOAI and Human Geography? How did the idea of book come along. And what is your intended audience?
- [02:45]
C
Sure, sure. Thanks for the introduction, Devika, and thanks for having me here. Yeah, it's great to be here and it's great to see you. About two years ago, I assembled a team. So with me, Dr. Sicheng Wang from University of Southern California, being the leader of the project, we wrote a systematic review article how geoai has been adopted in various domains of human geography. And from that review we realized that GUI appears in every corner of geography. This is the coming trend. So then we decided to write or edit this book to reflect this great trend in human geography. And this book is meant for everyone who is interested in the geospatial world and wants to know more about GeoAI, the students, researchers and practitioners who care about how the intelligent systems can actually help us as Java first to understand and shape the human experience of space.
- [03:54]
B
That's very interesting and I like how you said that AI is in all parts of the work. But in the opening chapter I see how you discuss human geography in the era of big data and AI, and you talk about how AI is transforming geographic inquiry, in your view, how AI has changed the way geographers approach traditional questions of space, place and time.
- [04:20]
C
I think AI has reshaped how geographers think about old questions of space, place and scale. And right now, instead of just mapping the patterns, we are now modeling the relationships, predicting dynamics, and even generating synthetic scenarios using generative AI models. So I think with AI we are able to process massive spatial data sets that were impossible to handle before. Right. So think about satellite imagery or GPS traces, trajectories and or social media streams. Those are massive databases. And what's cool is that AI, you know, I don't think that AI will replace geographical thinking. I think it pushes it further. So we are still asking, you know, why here or why now? But just with the new tools and that can uncover this hidden patterns, hidden spatial patterns and the relationships that human intuition almost might need. So I think we're using that new tool, but still the questions will remain the same.
- [05:31]
B
That is very well put together. And I totally agree to what you said, that it is not going to replace what we do in geography, but sort of amplify it and push the boundaries in terms of scale and speed. Now, I like how you talk about in the chapter in the Rise of Geospatial Artificial Intelligence, how you trace geoai's development from early computational geographies to today's foundational models. What do you see as the key milestone in this rise and how they are shaping the future of geospatial research.
- [06:06]
C
Yeah, I think that's a really good question. If I had to pick the key milestones in geoai, I want to start with the early computational geography in the 1960s when geographers, we first tried to quantify spatial patterns. That was 1960s. And then came the machine learning phase in the 2000s, which opened the door for patent recognition, especially in the remote sensing field. And around 2015, deep learning, you know, came into play and revolutionized image analysis, for example. And suddenly we can map, you know, net cover urban features and then we can use, we can map that in a very automatic way. Right. And right now, stepping into the 2020, we're in the foundational model era with massive AI models trained on global geospatial data or, you know, contextual data. So I think each stage made spatial analysis more data rich and scalable. So each stage move us closer to the intelligent systems that can actually reason about geography, not just visualize it.
- [07:22]
B
So you would agree that AI is not completely new. It has been a gradual rise over the years?
- [07:28]
C
Yeah, yeah, I think so, yeah. It's always on the rising trend. So we have a series of key milestones that keep this trend forward.
- [07:36]
B
Totally agree. Yep. One of the very interesting chapter which I found was the pillars of GEOAI in human geography and how you identify data, algorithms and computational infrastructure as the three pillars of GeoAI. How do these elements interact and where do you see the biggest gap in current practices of how we are applying AI?
- [08:00]
C
Yeah, I think from my perspective, you've talked about data algorithms on computational infrastructure as the key pillars. I think those three are like the holy trinity of geoai. So I think they rely on each other completely. So you cannot run advanced models without good data. Right. So even great data won't help if you don't have algorithms or computational resources to. To process this massive amount of data. And right now I think that the biggest gap isn't just technical issue. I think is the access to it. So many geographers don't have the same computational release sources as the big tech companies or governments. And the data size remain unevenly distributed across the world. We talked about digital divide issue for many ages. I think right now we need more open, inclusive infrastructures and transparent algorithms that help smaller institutions or individual researchers to participate in this AI revolution. I think that's the biggest challenge.
- [09:13]
B
Yes, totally. I think the spatial equity in terms of infrastructure is a big challenge to solve as we think more about it and thank you so much for bringing this up. I think it's a very relevant point to be discussed as we advance AI in geography. One of the chapters focuses on integrating natural language processing in human geography. How do you think natural language processing is being used to address spatial problems? And how do textual geographies like social media, reshape our thinking of space?
- [09:45]
C
I think natural language processing either. NLP is quickly and quietly becoming one of the most exciting tools in geography. So I think that it lets us to extract spatial meanings from everyday text. As you mentioned before, tweets, news reports, or even policy documents. We can use NLP to handle that. So now we can map emotions or perceptions and narratives about the places that do not really show up in the traditional data sets. Right? So think about how people tweet during the disasters and how local media frames about projects. Right? So I think those patterns tell us a lot about the spatial inequity or belongings or how people feel about different events. So I think textual geographies are kind of reshaping how we understand the space as a leaf and a felt place, not just a measured place. So I think it's giving geographers a new way to blend the stories with data. So I think this is the most exciting part of it.
- [10:59]
B
One very important chapter which I found was explainable AI in spatial analysis. And it highlights the importance of transparency in spatial AI, which I think is very important because most of spatial AI is being used to solve some real world problems. So we need to be transparent for people to use our models and our solutions. So how do you think explainability can make spatial analysis more trustworthy and why it is so important to bring that in account?
- [11:30]
C
Yeah, yeah, I think that's a very good question. So if I have to give you an example, I think explainable AI is like turning the lights in a black box. So we always think that machine learning models or some deep learning models are black box models. Now we're trying to turn the lights on, right? So we can see why an AI model made a certain prediction in that case, say, why it labeled an area as a high risk area or a green space. So we can start trusting it more if we know why the main AI is making that decision. So I think this transparency helps communities, you know, policymakers and researchers understand how spatial decisions are made. And it's not just about technical issue. I think it's ethical issue. Right. So if we can explain AI's reasoning, we can also question it, you know, correct it, or even, you know, ensure it that it reflects the real world context. Right. So ultimately, I think at this explainability builds accountability, which is crucial when, you know, those models influence decisions that affect people's homes, health or livelihoods.
- [12:48]
B
I just love how you said explainability brings accountability. And that's so important for all of us to remember as we design these AI solutions. Later on in the book, there is a chapter on human centered computer vision for urban sensing where the authors explore how AI can interpret state view imagery to understand urban life. So in this kind of work, what are the risk of bias in how AI sees urban environment compared to the human perception of urban environment?
- [13:20]
C
Yeah, I think bias is a huge issue in AI's vision of the city. So you've talked about the street view, right? I think for street view imagery, it reflects, you know, who and you know, and what gets photographed and when and how. So I think AI models can inherit all those biases. Right? So for example, an algorithm might learn to associate, you know, certain building styles or street conditions with safety or wealthy, but in this case it can reinforce the stereotype. So I think this is a big challenge. Humans, we see context, but AI models, in many cases it only sees pixels. So this is the risk, Right? So the model can mislead diversity as a disorder, for example, or miss the social meanings behind those urban scenes, especially when we're using distribute data as the proxy. So I feel like we have to teach AI to see like a geographer, not just from a camera's angle, from the pixels perspective. So we need AI to be more context aware and to be more inclusive and to be more sensitive to how city actually move.
- [14:41]
B
Yes, that's a very good point. One of the things which I really loved about the book is how the book covers GEOAI across multiple subfields ranging from cultural to political geography to health and economic geography. And there's a wide area of things which you cover. One curiosity I had after reading this was which subfields do you think have been most transformed by geoai and which remain less influenced by AI as of now?
- [15:13]
C
Yeah, I think I mentioned about the review article that we wrote about two years ago. It's a systematic review that covers 1500 articles. Those articles merge GOAI with the subdomains of human geography. And from that review article we realized that health geography has probably been the most transformed by Geoai, especially during the COVID 19 phase. So when mobility data, remote sensing and AI models came together, we can use methods and big data to track outbreaks and understand the risk patterns. So again, GU also help link environmental Data, behavioral and health data at a scale that we've never seen before. Right. But on the other hand, I think cultural and political geography still, you know, left behind a little bit in geo AI adaptation. From my experience, I think that's partly because the subfields, as I mentioned, culture, political geographies, they deal with narratives and human meanings, you know, things that do not always fit into the data driven models we are using right now. So I think that's why they kind of lag behind. But I think that's starting to change as natural language processing and multimodal AI gets better in terms of capturing the textual data from that dimension. So I think all the subdomains of human geography will benefit from this AI revolution.
- [16:54]
B
Thank you for clarifying how that is about to change because that was going to be my next question to you. Do you see the change happening and other subdomains picking up on the AI trend? So that's very good to know that there is some kind of momentum change that is going to happen in future. Later on you introduce ethical spatial decision making and sustainable geoai, which I think are very relevant and important topic to cover. In your opinion, what do you think justice mean in terms of geo AI context and how can AI models be more accountable to the communities that they are going to affect and serve?
- [17:35]
C
Yeah, I think from my perspective justice in geo AI it's about who, you know, benefits and who gets left behind. And a couple of weeks ago my article Geo AI is widening the Digital wide got published out in Annals of American Association Geographers and we, I actually talk about this issue in that article. So it's about making sure that AI does not just serve those who with power, money or data access. Right. So accountability means evolving communities in how the models are built and how the data, what data they use and how outcomes are interpreted with the support from the community. So for example, if an AI system is mapping disaster risk and I think the local voice should be guiding what risk mean to them. Right. So I think justice also means transparency for example, with open data, interpretable models and fair representation about who participate in when we are building the model. So I think ultimately GEOAI should empower the communities to see and you know, to question and co create their spatial features. So not just as being the subject of analysis but being really participating, collecting the data and building the model.
- [19:06]
B
Yeah, that's a very important point. How do we educate the community on doing that? And I see that some of your work is also focusing on educating the community around this and make Them a part of this decision making. And from your answer, actually I got one thought because I see you are working globally, some of your research is global oriented. Do you see a global disparity in how Goai is being used? And are there ways we can address that global inequality in terms of AI access and use?
- [19:39]
C
Yeah, I think right now I have a project going on which is in Africa and it's educational project. So we're trying to use generative AI model to quickly generate teaching materials for the students in Africa. So I think right now Africa is facing severe educational inequity with outdated teaching materials and insufficiency of the teachers. So we hope that this generative AI teaching initiative will help address this issue. I think AI is kind of reshaping geography education in the best possible way because right now as geographers, the students who study geography can explore global data sets using automated methods to map risks. Or even right now with Genai, we have the chat box to simulate different policy debates. I think it's turning geography from a map reading subject into a more dynamic data driven discipline. But I think for teachers like us, we also need to evolve. We cannot just add AI tools to the teacher materials. We have to teach students how to think critically about them. So I think this is the most important thing. The students need to understand data ethics, for example, the model biases and how, you know, the social, how the decisions will have the social impacts, but not just the coding, but the social impacts. And I think that the ultimate goal is to make them both tech savvy, you know how to code and also social conscious. I think if the students have both of those two perspectives, it will make them better geographers who can use AI in the wise and creative way.
- [21:31]
B
Thank you so much. This covers my next question because the following chapter talks on generative AI and geography education and it explores how AI enhanced learning is reshaping geographic education and how the teaching needs to evolve in the era of AI. And covered some of those points. But I want to touch a little bit on what do you see in terms of workforce development? How do we get our students ready to get into the workforce in this time of AI?
- [22:04]
C
Yeah, I think right now we need to do more studies on the job market demands. Our team right now is doing a new article that use the job descriptions from different subdomains of geography and we will analyze what's the job demands, what's the market is demanding the students to have. What are the new skills that students need to have before they graduate? I think I can share more when the study has some preliminary efforts coming out, but I think we need to understand what the market needs so that we can develop better curriculums and better courses to teach our students.
- [22:43]
B
Yes, we look forward to this interesting study because I think that will be very informative in a lot of terms. Also, do you think that the educators need to be trained in how the education is given in this time of AI? So there is a need for not only the student training, but also we as educators. How do we change the way we teach?
- [23:06]
C
Yeah, yeah, I totally think so. Because right now, with the support of generative AI, I can design my teaching material in a much faster way. I would say so for the labs I'm designing with the support of AI, it will save me hours. It used to take me a day to design some lab materials, but right now it only took me probably half an hour or just one hour. So I think if we can use AI tools again, creative and in a wise way, I think it will definitely help geography teachers to improve the efficiency.
- [23:41]
B
Yep, thank you so much for that perspective. Finally, in the concluding chapter, Shaping Tomorrow the Future of GEOAI in Human Geography, you envision a socially equitable geoai. So what do you think a truly human centered future of spatial intelligence look like? And how can geographers contribute to that transformation?
- [24:04]
C
Yeah, yeah, thanks for that question. I think a truly human centered GEO AI future would combine computation with ASICS and with community voices. And as I mentioned before, those two things are very important. So it's not about building a smarter algorithm, it's about building a fair system. So in that vision, AI will help us tackle, for example, a lot of challenges like climate, injustice, resources, inequity and urban challenges through this inclusive participation and open science. Right. So I think geographers can lead that transformation because we understand that every data set has a place, a story and has a history, because we know the context of data sets. Those are geospatial data sets. Those are in our domain. So our job is to make sure that AI learns that too. Right. To see geography not just as a data points, but the living places in shape by human experience. So that's what geographers can contribute in this AI revolution.
- [25:17]
B
That's a very good closing remark for our conversation. I want to keep talking about more, but we are a little bit short of time. But before we wrap up, is there any upcoming project or initiative that you are excited about and would like to share with our listeners?
- [25:35]
C
Sure, sure. I have a lot of excited project going on. But I want to share a new project that we're launching. It's called Generative AI for Geography. And Deepika, I think you are also in that project as the co editor. So we are exploring how generative AI can help students and researchers co create learning materials, for example simulate spatial process and even build digital twins for the real world environments. So we're trying to see how generative AI is helping geography is to make geography better. So the goal of Gen AI I think isn't to replace human insights but through a augmentation process we try to augment it. So helping geographers us to think more creatively about the connections between people, place and data. So it's also making geography more accessible, especially for regions that lack resources. JN and I can be a huge support. So we are now back in this book projects, editing this book projects and we think that we hope that this book project will become a bridge between education, between research and also equitable AI revolutions in this geospatial world. So we plan to publish this book out earlier next year or at least by the mid of next year. So yeah, please stay tuned.
- [27:02]
B
Yeah, I'm very excited to be the part of that project and once we once it's out probably we can have another conversation focusing on that book.
- [27:11]
C
Sure.
- [27:12]
B
Thank you so much Dr. Wang for joining us and sharing insight on this very important and forward looking book. And thank you to all our listeners for tuning in. I look forward to seeing you next time on another episode where we'll explore another fascinating work shaping the future of geography. Thank you all.
- [27:30]
C
Thank you very much.