KAZ YONEDA: Community at the Potentiality of AI and Architecture

REFLECTIONS ON THE SOTAMAA + YONEDA STUDIO

‘The best way to predict the future is to invent it.”
Alan Curtis Kay, American computer scientist

Under the rubric of “Community”, the first in a series of threefold xLAB Summer Studios, Sotamaa + Yoneda Studio explored the potentiality of Artificial Intelligence and Architecture, and its affect on the very social fabric that comprises our civilization. The moment of Singularity is imminent, and as such, architects ought to speculate, conceptualize, and produce a viable prototype for this possible near-future. The teleological eventuality of humanity’s logic and reason is Artificial Intelligence, which makes this an opportune moment to critique as well as to create a heterotopic world in which both corporeal (human) and virtual (AI) sentient existences can co-exist in a new cybernetic world order.

The studio operated with the hypothesis that AI will inevitably and fundamentally change the way city will emerge; financially, programmatically, spatially and culturally. It will change the way future constructions are financed and thus fundamentally effects the nature of development.

Before we design the AI powered architecture we must design the AI itself. How can we build an AI, which is aligned with our interests? Can we build an AI without losing control over it, and what kind of an AI should we build to begin with? An artificial Intelligence will surely help architects who ‘must be proficient at balancing complex, nonlinear relationships between diverse constraints’ (Landa, 1997), but beyond proficiency, could an AI help us do more? Could it help make our collective dreams reality?

Data is constantly uploaded from people via their devices onto a cloud. An Artificial Intelligence could help harness the desires, dreams and inner lives of people, and assist the architect in the creation of new architectural outer worlds? Could an AI predict behavior, which has not yet been expressed, and thus empower (economically and otherwise) the creation of something fundamentally new and surprising?

Speeds of architecture, economy, and technologies are all different. Consumption has shifted towards software, ethereal experiences and away from hardware (architecture). This modality requires architecture to become a robust, long-lasting platform capable of delivering experiences that cannot be downloaded, and simultaneously a salient structure in which fast-changing desires can unfold. An exciting opportunity for changing architecture in the near future comes with the possibility of bringing different technologies into dynamic play within architecture, combining them inside the “same frame,” in order to produce new architectural effects, which deliver experiences to all senses.

Similar moments of decoupling will emerge in terms of function and built form will emerge. Functions, driven by economics and AI, will become ethereal. They will become flexibly and dynamically assigned across buildings, while at the same time buildings must deliver an increasingly rich urban texture of distinct atmospheres and unique places. As Sotamaa stated, “I believe that particularly exciting new challenges and possibilities arise with the emergence of Artificial Intelligence. An AI powered Smart City can lead to a radical uncoupling of program from built form. For example, a school can be viewed more as service which dynamically taps into city’s resources, and less as a discreet building. Coincident with the uncoupling of program and building, physical environments will have to increasingly compete with uniqueness and experiences that you cannot download. If buildings provide flexible infrastructure for intelligently assigned transient program how does specificity in terms of identity and atmosphere emerge? What do buildings in an AI powered Smart City look like and how do we design them?”

With this basic underlying attitude of optimism and opportunism towards the Artificial Intelligence, Sotamaa + Yoneda Studio had nine talented students from all over the world assigned into three teams of three students each. As a team, they were tasked to propose their visions of the community in the age of Singularity. We asked that the they embrace it, not resist it, but rather, take the condition of cybernetic normalization as a given. 

Team Deckard

“Team Deckard” proposed a new program consisting of housing plots for sale and services plots to rent on the site surrounding the future Artificial Intelligence Industry Academia Collaboration Platform being planned in reality. Their proposal was based on an assumption that a joint-venture would build and supply the machine learning algorithm that receives the user data and generates the form. For example, when someone decides to live in a particular area of Tokyo, the AI Center collects all the data bought by the joint-venture, such as social networks behavioral patterns, internet searches, purchase actions, and health metrics. A Deep Learning algorithm is applied on the collected data. Specifically, any objects – bought or possessions – that can be visualized as an image are deposited as raw ingredients. The AI sorts them and analyzes their contents: the subject, the colors, the positions, the points of view, and any parameterizable qualities. Eventually, the process would hypothetically allow the AI to fully comprehend the personal preferences in the form of omni-sensorial qualities and characteristics. This allows the system to generate the spatial features and specificities. For example, Pantone 13-1125 TCX Peach Quartz, 13W indirect light, 4,5cm round corners, panther-fur rug, and so forth. Furthermore, this information is combined with the project definitions from the conventional procedure between an architect and a client. The conclusions derived by the AI, a Deep Dream Algorithm, are applied to combine all those features into generating a form. In parallel, architects reference massing, form, and materiality in constant dialogue with the AI to create architecture. In this scenario, an architect is charged with the curation of the obtained space: Its location, its display, its relation with other buildings, its consequences to the public space, and its perception. The architect is aided by yet another machine learning algorithm that reacts and suggests modifications, which enables a rule-based decision on where and where to place the buildings on site. Once the development of this collective of individuals starts growing exponentially, a real community emerges, and its needs are met by self regulated investments in terms of services.

Moving City

“Moving City” explored how autonomous mobility technology, coupled with artificial intelligence can make cities more dynamic and cosmopolitan communities. Things in a city already move. People, goods, vehicles, and services all navigate through streets, in and out of buildings, and underground. Even buildings constantly vibrate in its cadence. However, the way things move is changing and autonomous mobility technology is fundamentally redefining the idea of circulation in cities. Multi-modal autonomy exerts new challenges on architecture but also creates new opportunities, and their project explored autonomy at even an architectural scale. Some technologies that are already on the cusp of being realized in an urban scale are things like Google’s autonomous car, a shared mobility technology. At a smaller scale, there is Greg Lynn’s autonomous mobile helper, Gita, that will follow you around the city and carry your goods. Besides the movements of these objects and vehicles, they were also interested in what happens when buildings move. In a conventional section of a city, everything is organized around the street and the street is organized around the car. This is an organization based on thresholds and hierarchy. There is a strict boundary between the interior and exterior of the buildings, between the sidewalk and the road, and inevitably, the road is a space dedicated entirely to vehicles. The urban infrastructure is organized by streets because they provide access to service utilities. This team imagined a city and a community in which the street is no longer dedicated to the cars and instead is a common space that can be reappropriated for multiple modality. The use of artificial intelligence would liberate vehicles, objects, and buildings from conventional paths, to respond to their environment and inherent demands, and to navigate freely through the city. Another potential benefit of a building that can move is that it would have to become more autonomous, and therefore, attain a kind of off-the-grid resiliency to take care of its own power generation, water management, and disaster preparedness.

The Incrementals

“The Incrementals” based their thesis on how AI would allow a change in the economics of development, happening in a different increment, which would allow a richer city and a more robust community, where growth is possible in different ways. This is manifested in the form of large scale, hierarchical, inflexible, high capital and centralized ‘city-making’ projects in the name of masterplanning. Herein, the conventional masterplanning has failed to grasp the complex and rich dynamics that define contemporary cities. The obsession with form tends to gloss over political, economical and social experiences. Their intention was to domesticate urbanization and to transform our cities into responsive devices. First, their protocol would identify the unused or overlooked areas in design as well as spaces for AI to expose the weaknesses in the intersection of these systems. On the other hand, an architect would be charged with finding more personally meaningful outcomes than the ones predefined for us. By engaging these identified voids through design-hacking and space-hacking, we can create platforms for other people to engage with their world more creatively.

Kaz Yoneda is Founding Principal of bureau 0-1, Inc. Prior to the launch of his firm, Yoneda served as an inaugural director of space design for Takram and ran the Harvard GSD Studio Abroad with Toyo Ito. Currently, he is an adjunct assistant professor of Architecture and Urban Design at Keio University, as well as a visiting lecturer at Japan Women’s University.