For many people, swimming in a pool on a hot July day with kids splashing and lifeguards hollering is the quintessential summer experience. Not only do pools provide a place for people to recreate and cool off in the summer heat, but access to pools is essential to making basic swimming and water safety instruction available to communities. Water should be a resource everyone has equal access to and can safely enjoy. However, a great number of people can’t easily access this resource. This is a major contributor to a significant public safety issue: 44 percent of the US population doesn’t know how to swim. As a volunteer swim instructor for children from underserved communities, I saw firsthand how many children had to ride a bus several miles to the pool where I taught to receive swim lessons. This observation led me to want to explore the relationship between wealth and access to pools. Hamilton Pool, 94949, Marin County For this project, I set out to map the availability of pools* in the San Francisco Bay Area, where I grew up, and compare that to population and median income data from the US Census by zip code. Using the considerable database of the US Census Website, I was able to collect individual data for each Zip Code Tabulation Area (ZTCA5) in the 10 Bay Area counties. * For this map, I counted "pools" as any publicly (i.e., public pools) or semi-publicly (i.e., HOAs and country clubs) available pools, but I did not count any backyard pools. Organizing pool and population data. Additionally, I collected my pool data through sites like Yelp that advertise pools to those looking for them. The resulting maps demonstrate the connection between median household income and the availability of pools, and show that community wealth is, indeed, a determinant of access to this vital resource. The pool availability map in process. These maps reinforce the need for local municipalities to prioritize funding for publicly accessible pools in lower-income communities to better provide for the health and wellness of all their constituents. Charles Scott is a Senior in high school. In his free time he swims, lifeguards, and teaches swim lessons. He has always loved maps, and wants to combine his love of maps and swimming to make a change in his community. A note from Guerrilla Cartography board member Zachary Loran: In late 2023, Darin Jensen (GC’s founder and president) connected me with Charlie, who had reached out to him for some guidance on this project. Like Charlie, I was first introduced to Darin and GC as a teenager, a parallel that made me a natural choice to be Charlie’s contact at GC. It was a pleasure watching Charlie create this project and assisting where I could, and I’m excited to see what this young cartographer does next!

Artificial intelligence is dramatically changing the practice of cartography, not just as a mechanism for extracting and manipulating data, but as a tool for creating, curating, and interacting with geospatial images. A simple web search for “AI in cartography” points to a vast number of articles, examples, and website resources—reflecting just how embedded in the practice of cartography this relatively recent technology has become. With so much information about AI in cartography already online, I thought it might be interesting to see how an AI-driven chatbot (specifically, ChatGPT 3.5) would summarize the state of the field. The prompt I submitted to ChatGPT requested that its output take the form of a blog post for the Guerrilla Cartography website. Even if that website wasn’t included in the chatbot’s training, it was surely familiar with the blog post as an internet genre. The ChatGPT-authored blog post on AI in cartography is reproduced here, unedited. Unsurprisingly, given how far chatbots have progressed in the past few years, the response is grammatical, well-organized, written in a breezy style, and seemingly comprehensive, but also somewhat generic. Examples and illustrations would certainly help, so I asked a few follow-up questions to flesh out the chatbot’s overview—with mixed results, as you can see below. You Write a 1,000 word blog post for the Guerrilla Cartography website on how AI is changing cartography. ChatGPT Title: The Cartographic Revolution: How AI is Transforming Maps In the ever-evolving landscape of cartography, a silent revolution is underway. It’s a revolution powered by artificial intelligence (AI), reshaping how we perceive, create, and utilize maps. Gone are the days of manual charting and static representations. Today, AI is at the forefront, driving unprecedented advancements in mapping technology. From automated feature extraction to real-time updates, personalized navigation, and environmental monitoring, the impact of AI on cartography is profound and far-reaching. The Rise of Automated Mapping At the heart of AI’s influence on cartography lies its ability to automate previously labor-intensive processes. Imagine a world where maps are no longer painstakingly hand-drawn but instead crafted with precision and efficiency through AI algorithms. Feature Extraction: AI algorithms can now scan satellite imagery with remarkable accuracy, identifying and extracting features such as roads, buildings, rivers, and more. This means that what used to take weeks of manual tracing can now be accomplished in a fraction of the time, allowing cartographers to focus on higher-level analysis and interpretation. Land Use and Cover Mapping: One of the most impactful applications of AI in cartography is in land use and cover mapping. By analyzing satellite data, AI can classify different types of land cover, from forests to urban areas, with impressive detail. This not only provides valuable insights into our changing landscapes but also helps in urban planning, agriculture, and environmental conservation efforts. Accuracy Redefined AI not only accelerates mapping processes but also enhances the accuracy and reliability of the resulting maps. Quality Control: With AI’s keen eye for detail, errors in maps can be quickly identified and rectified. Inconsistencies in data, such as mismatched road networks or misplaced landmarks, can be automatically flagged, ensuring that maps are as precise as possible. Data Fusion: By seamlessly integrating data from multiple sources—satellite imagery, GPS, aerial photos—AI produces maps that are more comprehensive and up-to-date. This fusion of data leads to richer and more detailed representations of our world. Real-Time Updates and Dynamic Maps The days of static, outdated maps are behind us. AI enables the creation of dynamic maps that reflect the ever-changing nature of our environments. Real-Time Monitoring: Whether it’s monitoring traffic conditions, tracking weather patterns, or responding to natural disasters, AI-powered maps provide real-time information crucial for decision-making. Emergency responders can rely on these dynamic maps to navigate evolving situations with precision and speed. Crowdsourced Data Integration: AI is also adept at processing and integrating crowdsourced data from platforms like OpenStreetMap. This means that community-contributed updates, such as new roads or points of interest, can quickly find their way onto official maps, ensuring that they are always current and relevant. Personalized Mapping Experiences AI doesn’t just stop at creating maps—it’s also revolutionizing how we interact with them on a personal level. User Behavior Analysis: By analyzing user preferences and behaviors, AI can tailor maps to individual needs. Whether it’s suggesting optimal routes based on past travel patterns or highlighting points of interest aligned with a user’s interests, maps become more than just navigation tools—they become personalized guides. Contextual Mapping: Maps can now adapt to the context in which they are used. Factors such as time of day, weather conditions, or nearby events can influence how a map is displayed. For example, a map might highlight nearby cafes in the morning and restaurants in the evening, providing a seamless and intuitive user experience. Visualizing the Future The visual appeal of maps is also undergoing a transformation, thanks to AI. Style Transfer: AI can apply different styles to maps, allowing users to choose between classic, watercolor, or even futuristic renditions. This not only adds a layer of aesthetic appeal but also serves practical purposes, such as optimizing maps for specific audiences or purposes. 3D Mapping: With AI, we are witnessing the rise of detailed 3D maps that offer a new perspective on our world. Whether it’s visualizing urban landscapes or exploring natural terrain, 3D maps provide a richer and more immersive experience. AI for Environmental Monitoring The power of AI extends beyond just human landscapes—it’s also making a significant impact on environmental mapping. Change Detection: By comparing satellite images over time, AI can detect changes in vegetation, water bodies, and urban areas. This capability is invaluable for environmental monitoring and conservation efforts, allowing us to track deforestation, monitor wildlife habitats, and respond to ecosystem changes. Habitat Mapping: AI’s ability to analyze satellite imagery is instrumental in mapping habitats for various species. Conservationists and researchers can use these maps to identify critical areas for protection and better understand the interconnectedness of ecosystems. Navigating the Future with AI The influence of AI on cartography is undeniable, but it also presents challenges and considerations. Data Privacy: As AI relies heavily on data, concerns around privacy and data security are paramount. Safeguarding personal location information and ensuring transparent data usage practices are essential. Algorithm Bias: AI algorithms are not immune to biases present in the data they are trained on. This can lead to inaccuracies or unfair representations in maps, highlighting the importance of diversity and inclusivity in data sources. Technical Challenges: The processing of vast amounts of geospatial data requires powerful computing resources and efficient algorithms. Continued advancements in AI technology will be crucial to overcoming these technical hurdles. Embracing the Cartographic Revolution In conclusion, we are witnessing a cartographic revolution driven by AI—a revolution that is reshaping how we map and navigate our world. From automated mapping processes and enhanced accuracy to real-time updates, personalized experiences, and environmental monitoring, the possibilities are endless. As we continue to harness the power of AI in cartography, it’s essential to remain vigilant about data privacy, algorithmic biases, and technical challenges. By addressing these concerns, we can fully embrace the potential of AI to create maps that are not just informative tools but also works of art that inspire, educate, and empower us to explore the world around us. The future of cartography is here, and it’s powered by AI. Let’s embark on this journey together, charting new territories and discovering the boundless possibilities that lie ahead. You Add three examples or images that illustrate the main points of this blog post ChatGPT Certainly! Here are three examples or images that illustrate the main points of the blog post on how AI is changing cartography: 1. Feature Extraction: Before and After AI Image Description: A comparison between manually identifying features in a satellite image (left) versus AI-automated feature extraction (right). The AI-automated process shows clearly identified roads, buildings, and water bodies with labels. Caption: AI-powered feature extraction drastically reduces the time and effort required to identify and map features such as roads, buildings, and water bodies. The image on the right demonstrates the efficiency and precision of AI in extracting these features from satellite imagery. 2. Real-Time Updates: Dynamic Traffic Monitoring Image Description: A dynamic map displaying real-time traffic conditions with color-coded lines indicating congestion levels on roads. Caption: Real-time updates powered by AI enable dynamic maps like this one, which shows live traffic conditions. Color-coded lines represent varying levels of congestion, providing users with up-to-the-minute information to plan their routes efficiently. 3. Personalized Mapping: User Behavior Analysis Image Description: A personalized map displaying suggested points of interest based on the user’s preferences and past travel patterns. Caption: AI analyzes user behavior and preferences to create personalized maps. In this example, the map suggests points of interest based on the user’s past travel patterns, enhancing the user’s experience by providing relevant and tailored information. These visual examples complement the main points of the blog post, showcasing how AI is transforming cartography through feature extraction, real-time updates, and personalized mapping experiences. The Chatbot’s Limitations—for now This small experiment demonstrates that ChatGPT can produce a reasonable overview of AI’s impact on cartography—and is likely just as fluent with any other subject included in its online training. For a blog post, though, there should really be more specific examples, which is why I submitted a follow-up request. The examples were interesting, but lacked images because ChatGPT’s current public version is only text-based. Since the chatbot’s examples might have been based on actual online images, I requested their URLs. ChatGPT was happy to comply, but the links led nowhere. Thinking that the links might have expired, I looked for the images on the Internet Archive’s Wayback Machine, but the images weren’t found there either. Most likely, the images were chatbot “hallucinations.” Google’s chatbot, Gemini (formerly Bard) is a ‘multimodal’ chatbot that is trained on images as well as text. So I tried a similar exercise there, asking for specific examples and images. Gemini produced a similar text and some map images, but they weren’t very helpful. They didn’t illustrate how AI might have been involved in creating the map, and they didn’t relate well to the text—almost as if they were selected at random from a collection of maps. But chatbots are improving rapidly. More capable versions are being introduced every few months, so it won’t be long before multimodal chatbots with advanced image analysis capability can go beyond the limitations we see today. Whether the results are interesting and useful is something that humans will still have to decide. Charles Drucker has been a Guerrilla Cartography board member since 2019.

Guerrilla Cartography aims to promote the understanding that there are many views of the world, that how we understand space and place can vary, and that we should think critically about the maps and the graphics we make and consume. Our organization has designed processes for creating our atlases that support these goals. 1) By Featuring Diverse Narrative Viewpoints A different person or group of people produced nearly every map in our atlases, each contributing their indi­vidual aesthetic and experiences to the broad theme. We can never tell all stories about food or water, but crowd­sourcing content allows us to glimpse a few things that we might not otherwise encounter, giving the atlases a variety of perspectives. An example of an unusual narrative viewpoint is the map “Holy and Unholy Spirits Along the Ganga: A Map of Polluters and Prayers,” which juxtaposes spiritual sites with polluting industries along the Ganga (Ganges) River in India. This map was a collaboration between social ecologist Bidisha Banerjee and cartographer Luc Guillemot for Water: An Atlas. Holy and Unholy Spirits Along the Ganga: A Map of Polluters and Prayers by Bidisha Banerjee and Luc Guillemot in Water: An Atlas (2017). The atlases also exhibit a variety of literal viewpoints, including different projections and scales. One example is Garrett Bradford’s map of “Global Almond Trade and California,” featured in Food: An Atlas. Global Almond Trade and California by Garrett Bradford in Food: An Atlas (2013). The central map employs the unusual Peirce quincuncial projection in order to place California at the bottom, so as to better showcase the almond trade that emanates from it. The cartographic choices highlight the narrative of California as the center of global almond production. 2) By Placing Maps in Relationship to Other Maps and Information The atlases are meant to be both informational and entertaining, readily accessible to anyone with an interest in maps, in the theme, or both. After the call for maps is sent out, the maps that the crowd chooses to make end up creating a narrative for the atlas. We do not prescribe the narrative in advance; it grows organically throughout the process of building the atlas. We are then able to group the maps into narrower themes within the broad theme, giving a structure that makes the atlas more than simply a collection. This is the part of the atlas that is editorial, and that makes the board a part of each map and its story. While each map by itself has a story, that story also becomes contextualized in the order of the maps. We experiment with multiple linear “stories” as we work to pull the narrative thread through the atlas. There are other stories that can be told with these maps and other ways they could be grouped, but due to the constraints of producing a physical atlas, the editorial board decides the final grouping and order for the maps. We hope our arrangement of the maps provokes a response in readers, prompting them to agree with or question the narrative that is being told, and to translate that to their viewing of other maps and atlases. 3) By Building Legitimacy for Marginal or Atypical Cartographic Voices Among other things, Guerrilla Cartography is concerned with authority — the authority of who gets to produce and distribute maps and why. The narrative construction of the atlas and the editing process itself both help give legitimacy to voices that may not have access to traditional atlas or map publishing venues. Our process of pairing people who have map ideas with our volunteer cartographers is another way that we make space for many different viewpoints, such as the collaboration mentioned above for “Holy and Unholy Spirits Along the Ganga.” Banerjee had been immersed in a project about the Ganga since 2009; after submitting her idea to the Water: An Atlas call for maps, we connected her with Luc Guillemot, a Guerrilla Cartography volunteer. They worked together over email to bring Banerjee's vision to life. Seeing the ways that a variety of people represent geographic data is informative and instruction­al. While anyone can publish maps that they have created on the Internet, distribution may be limited. Without a digital home, these maps may also disappear. Guerrilla Cartography’s atlas publishing model enables a broader audience and a more concrete presence-both digital and physical. The crowd­sourced nature of Guerrilla Cartography’s atlases also helps viewers critically examine their assumptions of who has the authority to produce maps. 4) By Promoting Critical Evaluation of Content, Authorship, and Authority The myriad styles, narratives, and scales of the maps contained within Guerrilla Cartography’s atlases invite readers to question their assumptions about how a map is constructed, by whom, and for what purposes. The atlases also encourage readers to think critically about the very data that the map representations are constructed upon. For instance, in Water: An Atlas, we open with a chapter titled “Imagination.” Here we are mapping imaginary data, or in some cases actual data on imaginary or legendary phenomena. The map “North American Water Tensions in the Year 2028,” for example, depicts a dystopian vision of water scarcity-caused conflicts in the not-too-distant future. North American Water Tensions in the Year 2028 by Bryce Touchstone and Melissa Brooks, in Water: An Atlas (2017). For someone to read these maps, they must begin to understand that, while the map portrays a certain authority, the mapped data may not exist in the real, tangible world. Unreal data are being mapped. What does that mean for all the maps we see? Does it make us wonder about the “real” data that are being mapped elsewhere in the atlas? Alicia Cowart and Susan Powell are members of Guerrilla Cartography’s Board of Directors. An expanded version of this article was first published in Cartographic Perspectives, Number 92, 2019.

Interculturality as Shelter is a map made of many maps. It examines the ways in which young migrants experience interculturality in the neighborhood where their school is located. It shows their ways of feeling, walking, and making the city. A neighborhood where nobody cares Ciudadela is a town located in the southern part of the Buenos Aires province district of Tres de Febrero, Greater Buenos Aires, Argentina. It is a place of passage, circulation, trade, and exchange. Located halfway between one of the most important train stations and bus terminals in Buenos Aires City, and the train station that bears the name of the neighborhood (located in Greater Buenos Aires), Ciudadela unfolds as a diverse, heterogeneous, unequal, and insecure space. A territory that is home to many others: Ciudadela Norte, Ciudadela Sur, Villa de los Paraguayos, Fuerte Apache, Villa Herminia, Villa de los Rusos, and Villa Matienzo, among others. Places that bear ethnic, class, race, and national marks in their names that refer us to historical formations of otherness. As young people themselves point out, they “come from many trips.” Mobility takes place in the context of family migration, involving multiple movements between countries (in general, Bolivia and Argentina, but also Paraguay and Peru), provinces, cities, and/or neighborhoods. The multiple mobility experiences of these young people are, in turn, experiences of diverse territories and multiple border crossings, an extremely relevant aspect when reconstructing and analyzing their experience of the city. “Why is our neighborhood dangerous? Why can’t we walk without feeling fear? How did the neighborhood come to be what it is?” are some of the questions that young people framed in the context of an ethnographic and collaborative work experience carried out in a secondary school in Greater Buenos Aires (Argentina) between 2018 and 2019. In this sense, displacement cuts across this map in various ways. First, as a way of inhabiting, traveling, and “making” the city of the young people with whom we worked. Second, as a broader repertoire in their experiences of multiple territories linked to the migrant condition of most of them. Finally, as an activity or operation subject to control or “government” by the State, and specifically by security forces in urban space. Interculturality as Shelter shows the backside of this experience of the city as danger. Interculturality is a complex term that young people were debating at school in the context of a broader project. When I asked them what they meant by it, words such as “respect,” “tolerance,” and “cultures” emerged. When I told them about the idea of mapping interculturality, new aspects came to light: political and affective dimensions that were not visible before. It was the first time they carried out such a mapping experience. Taking an official map as a departure point was a limit but also a good opportunity to point out the tensions between their lived experience of the city and the one cartographers took into account. It was hard for young people to place themselves into the maps and find those places they linked to interculturality. Most of them worked in small groups but others preferred to do it individually. Part of my challenge in doing this map was to turn the eight maps I received into only one. Three of the eight maps that were combined into the final map. Mapping is always political Far from notions such as tolerance and respect, when mapping, young people link interculturality to places where they feel safe among others, such as friends and relatives. Places where they feel safe in their differentness. The production of other cartographies and maps enabled the creation of spaces for dialogue and the production of collective knowledge about the experience of the city. Reflection from a visual device allowed the articulation of ways of experiencing the city and narratives that dispute, and challenge hegemonic urban perspectives. The Interculturality as Shelter map appeared in Shelter: An Atlas, published by Guerrilla Cartography in March, 2023.