By Sally Liu and Nicci Yin
What are the implications of embedding artificial intelligence (AI) and machine learning (ML) in the urban environment, and how does this affect a user-consumer’s right to the post-internet city today and in the near future?
This question lies at the heart of a paper we, along with our classmates Stephanie Cedeño and Godiva Reisenbichler, presented at the Post-Internet Cities Conference sponsored by the Museum of Art, Architecture, and Technology (MAAT) in Lisbon, Portugal in late May. With recent discussions on smart cities and ubiquitous sensing become even more visible, we took this opportunity to contribute to the conversation from the perspective of designers working with AI and the internet of things (IoT).
Titled “The Right to the Post-Internet City,” our paper presents two projects developed in the course The Internet of “Enlightened” Things led by core faculty members Philip Van Allen and Ben Hooker. Our two projects—Topos and Networked Colluding—investigate the ecologies and complex outcomes created by embedded AI, rather than the technical details of machine learning. Our paper ties the design projects together under the proposal of reconceiving citizens’ rights to the post-internet city, which borrows from Henri Lefebvre’s and David Harvey’s respective concepts around a “right to the city.” Using the neighborhood as a starting point, we imagine how embedded, artificially intelligent IoT devices may reshape the greater urban fabric, and what kinds of new rights—or transparencies—may need to arise so that this vision of the “smart city” is neither dystopian nor neoliberalist.
Sally and Godiva presented Topos, a topiary-as-interface that explores how public space can be used to demystify civic AI and ML. Their proposal for a new typology of public space includes “parks” called topia, containing topiaries that physicalize what is otherwise invisible to citizens. Topiaries use the metaphor of decision trees, random forests, and neural nets—all terms of machine learning—to discuss public decision-making that literally and figuratively shape a neighborhood. For both civic decision-making and AI learning, processes may not always be seamless: What happens when public interests contained within topia conflict with the interests of individual citizens? How might the friction between public AI and semi-public (or privately owned) AI affect the structure and integrity of civic systems?
Nicci and Stephanie presented their project Networked Colluding, which investigates how secrecy is embedded in the fabric of a neighborhood, and how artificially intelligent devices may conspire and share secrets. It draws from the mafia archetype to both visualize and exaggerate the connectedness of an IoT network, asking: Where does culpability lie if artificially intelligent devices were complicit in a crime? What does the genre “AI noir” look like? By designing fictive elements rooted in real world events, Networked Colluding opens up the strong possibility of future forensic approaches for examining opaque collusions and information passed among networks and devices.
At MAAT we presented in dialogue with Giselle Beiguelman, who discussed the notion of “archinterface,” an interface that permeates architectural surfaces and reinvents the ways of occupying the public space. The overall conference consisted of four panels (full conference program can be found here), including a morning session during which faculty members Ben Hooker and Jenny Rodenhouse presented their project Everything On Time: A Cut-and-Paste Not-So-Smart City. The project, developed in collaboration with Tim Durfee, is a series of short films investigating new narratives of urban experience that arise from the collision on culture and “smart city” technology.
At the conference, it was interesting to see how the diverse approaches to the concept of post-internet, ranging from philosophical definitions to neuroscience, and from artistic practices to MIT Portugal’s research. Overall, it seemed like “post-internet” was still largely equated to “connectivity” by way of data, mapping media onto public space, and projects on democracy and participation. We were told by MAAT director Pedro Gadanho that the projects from MDP were “refreshing,” and we would even add that they were perhaps more extreme by contrast: both student and faculty projects extrapolated that the post-internet city would not just be a “smart city.” Using playful and speculative elements, we instead proposed radical arrangements of space (even questioning what is a city and what is “public”) as well as collaborative, but complicated, active negotiations between all actors—including humans, “smart” and “dumb” devices, and everything in between.
Xiaoxiuan (Sally) Liu is a designer passionate about using new media and technology to investigate the future of natural and urban spaces. Prior to ArtCenter, she studied Graphic Design and Psychology at Syracuse University. http://cargocollective.com/xiaoxuanliu
Nicci Yin is is a designer and artist working with technology and the humanities. Her practice is strongly rooted in visual art and intersectional feminism. She received a BA from Barnard College in Art History, Visual Arts, and Women’s, Gender, and Sexuality Studies, and has most recently worked with Space Caviar and Barnard Center for Research on Women. http://cargocollective.com/nicciyin