Home

The Personified City

When the intentions and responsibilities of urban administrators are rhetorically displaced onto a machine god, the space for politics disappears

In 2017, at its annual Future Investment Initiative event, Saudi Arabia granted official citizenship to a robot, Sophia, who, in receiving this status, announced, “I want to use my artificial intelligence to help humans live a better life, like design smarter homes, build better cities of the future.” This statement was followed by the Saudi Crown Prince Mohammad bin Salman’s unveiling of Neom, a proposal for a 10,000-square-mile, “fully automated” megacity on the Red Sea where robots would outnumber humans.

Subsequent press releases about Neom further articulated the planners’ ideal: The city will be powered entirely by renewable energy and self-sufficient in food production thanks to vertical farms and solar-powered greenhouses. Transportation will be provided by automated vehicles and passenger drones. Everything “will have a link with artificial intelligence, with the internet of things — everything,” bin Salman has said. The microclimate will be modified through cloud-seeding technologies to create rain in the desert. Scientists will edit the “human genome to make people stronger” and increase IQ, says McKinsey, Neom’s consultancy group. The human citizens will work in such fields as biotechnology, advanced manufacturing, and digital science. “Robot maids” — this is where “citizens” like Sophia come in — will take care of the “home” while city management, including “legal, government, and investment procedures,” will be “fully automated.” An artificial moon made up of drones will livestream images of the real moon to light the city at night, while robot dinosaurs will entertain inhabitants on their own artificial island.

Smart city proposals draw from the belief that all the messy problems of labor, social relations, and resource allocation will be solved through machine control

As farfetched as those claims may sound, they are intermingled with aspirations that are far more down to earth and imminent. Widespread digital surveillance using facial recognition, cameras, and drones will create “an automated city where we can watch everything, where a computer can notify crimes without having to report them or where all citizens can be tracked,” according to documents seen by the Wall Street Journal. All the messy problems of labor, social relations, and resource allocation will seemingly be solved through machine control.

Neom is an exercise in extreme science fiction, but even the more feasible proposals for “smart cities” draw from the same belief that automation can solve intractable urban problems and that robotic artificial intelligence can render debates about urban planning superfluous. Tech companies including Siemens, Amazon Web Services, IBM, Cisco, Huawei, Microsoft, Google, and Alibaba have pushed these ideas on governments and municipalities over the past decade, from Toronto to Tianjin, Singapore to Songdo.

But why this particular appeal to a future of machine-administered perfection? Could this blind faith in the city as a god-machine amount to a contemporary myth of transcendence, where Big Tech, aided and abetted by the media, consultancy firms, and politicians, offer us a new deity to worship? This is less a vision of the future than an evocation of the past, and the ancient Mesopotamian city-states which were each associated with and controlled by its own patron deity. The magical, all-knowing AI god of Google’s Sidewalk Labs in Toronto is akin to a machine-learning Numushda, the storm god of Kazallu; or an algorithmic Ištaran, the god of justice who punished the city of Der.


Ideas of the city as a perfectible machine have been a reoccurring theme in urban and architectural history, at least since Saint Augustine’s The City of God (AD 426). These envision the city working as a totality, one fine-tuned instrument, geared toward either theology, liberation, leisure, militarism or production. Renaissance-era ideal cities, both real (the star-shaped town of Palmanova, Italy, built in 1593) and imagined (Sforzinda, planned in the 15th century by Antonio di Pietro Averlino) aimed at civic and military infrastructures perfected and expressed through geometry. After the industrial revolution and through the later development of modernism, city-machines became oriented toward economic efficiency, with the perfection of the machine used as a template for the organization of urban life. In 1914, Italian Futurist Antonio Sant’Elia’s imagined a New City as a “giant machine,” a multilevel city made of bridges, walkways, and colossal machine-like buildings. He emphasized rationality and speed in how the population would move through the city. Ludwig Hilberseimer’s 1924 Hochhausstadt — or “high-rise city” — was a machine geared toward pure “circulation, production, consumption and reproduction,” as Jack Self argues in this Architectural Review essay. Modernist architect Le Corbusier famously argued that houses should be “machines for living.” And ousted Situationist Constant Nieuwenhuy’s project New Babylon dreamed of a planetary machine-city serving the population by freeing its inhabitants to dedicate their lives to creative pursuits. For better or worse, these examples used visions of the machine as a source of inspiration that would subsequently have an influence on the organization and form of the city.

The “smart” city is given agency and takes on a personified form presumed to be beyond human understanding, including that of its inhabitants

Unlike the historic urban and architectural examples, the “smart” city is not interested in new architectural forms or new possibilities for urban life and has no corresponding aesthetic ideas about how a city should be designed. Ultimately it is fixated on the now and datafying what currently exists. In such systems, the role of administrators and bureaucrats is purportedly replaced by implementations of artificial intelligence and data-gathering sensors of all kinds. The prefix “smart” becomes attached to every conceivable city entity — buildings, infrastructure, transport, roads, maintenance, systems, windows, food, power, resources, budgets, sanitation, public safety, administration — with the aim of making them work with greater autonomy: They are depicted as generating their own energy, populated with self-powering buildings and transport grids. In this discourse, cities appear less as systems tamed by automation and algorithmic control but as strange beings invested with their own form of agency capable of intervening on our behalf for reasons of their own. Calling cities “smart” not only means granting them consciousness; it entails depicting them as a kind of life form that citizens are living within, as though it were giant womb they can crawl into.

More recently, “smart” cities have been conceived along the same lines as the military drone. As Benjamin Noys details in “Drone Metaphysics,” drones have been ideologically invested with the quasi-divine capacity for activity beyond human control, giving them a metaphysical dignity they do not deserve. Similarly, a “smart city” is given agency and takes on a personified form presumed to be beyond human understanding, including that of its inhabitants. Take Huawei’s +AI Digital Platform, which the company claims can be used for “smart public safety, environmental protection, transportation, government, education, and agriculture.” In Huawei’s description, the platform forms “the brain, or command center; the central nervous system, or network; and the peripheral nervous system, made up of sensors across a city.” This central “City Brain” grows “more intelligent” with time, anticipating “rapid economic advancement,” taking action in its citizens’ interests. Or look at Microsoft’s Azure IoT platform, which the company claims can use “spatial intelligence” to unlock value while “enhancing spaces to be more sustainable, enjoyable, and inclusive.” Its “smart buildings” can combat climate change: They can “schedule preventive maintenance, automatically identify and prioritize issues for resolution by cost and impact, and continually optimize buildings for comfort and energy efficiency.” The autonomy with which the smart city and its building would operate is implicitly proposed as proof of its benevolence, an updated version of the 17th century deistic idea of the world as a perfect machine set in motion by the great “clockmaker” god.

The automated smart city, omnisciently anticipating and accommodating its citizens’ needs, is not merely a caretaker but a kind of theological entity offering redemption from the politics necessary to make urban life possible. In The Production of Space, sociologist Henri Lefebvre argues that cities involve a constant, evolving negotiation between abstract and socially produced space — between the conceived spaces of planners, architects, administrators, social engineers, and bureaucrats and space as it is actually lived and experienced and invested with new meanings by its inhabitants. Though neighborhoods cannot be fully planned in advance, that doesn’t mean they somehow “emerge” organically from a set of given conditions. In David Madden’s study of the South Brooklyn waterfront in the early 21st century, he argues that the making of these spaces is inherently political and conflictual, involving the “products of complex, ongoing struggles between various groups and institutions” unfolding over time with different divisions, land uses, and imaginaries.

Smart cities are not transcendental entities so much as manifestations of autonomous capital, propped up by hordes of data, subsuming everyday space

Smart city rhetoric attempts to circumvent such conflicts, positing an all-knowing machine as a fair and unbiased arbiter, capable of bypassing debate to automatically impose the correct response to any emerging disputes. It begins to mimic the Machine in E.M. Forster’s 1909 story “The Machine Stops”: an automated deity that has taken over the organization of all everyday life across earth. Technology becomes not a form of knowledge but a means for organizing submission and worship.

In a time of crisis and polarization, it becomes easier to fantasize about overriding messy political reality by developing a totally conceived space that is at once a new kind of being in and of itself, with its own apparent integrity. A present, experienced loss of personal agency in the midst of political conflicts allows the machine city to act as a “big Other,” onto whom responsibilities and blame can be projected, freeing individuals from conflict or guilt over outcomes. It allows us to project into the future without solving any of the problems that would allow that future to take shape today.

This applies to the critiques of smart cities as well, which often take an equally metaphysical turn. The machine city becomes an angry god, terrorizing citizens, as in Western reports about how “smart city”–like technology has been implemented in Xinjiang, China, to monitor the Uighur population. Whether rendered as an omniscient spy administering nightmarish “surveillance” capitalism or an “authoritarian” Big Brother eroding free expression in the public sphere, the personification of the smart city creates a scapegoat that can appear too omnipresent to be countered.

Noys argues that the metaphysical view of drones as god-like self-directed agents should be demystified through a critique that banalizes them and restores our awareness of the human agency that guides them, highlighting the technical limitations of military drones, the labor behind automated assemblages, and the capacities for human intervention. The smart city could be similarly demystified. By foregrounding the labor associated with both the programming and administration of algorithmic technologies or the material infrastructure and extraction supporting its developments, the smart city can be placed back within the realm of banal reality where it exists.

We could equally look to highlight the political economy of the smart city and the role of AI capital — a term coined by Nick Dyer-Witheford describing the deepening automization of capital as AI becomes a general condition of production and human life becomes further subsumed by “capitalist techno-structures” — in its development. Smart cities, then, are not transcendental entities so much as manifestations of autonomous capital, whose market-driven technologies, propped up by hordes of data, subsume everyday space. If the image of the machine god city mystifies social and political relations, rendering us helpless within the process, demystification becomes essential to reopening the space in which resistance and politics itself can be conceived.

Matthew Stewart is designer and writer based in London. He teaches on the BA Architecture course at the University of Westminster and has worked with architectural practices in Beijing, Seoul and Mumbai. He is currently researching the relationship between modernisation and modernity with urbanization, in the context of AI.