What happens if we let tech companies design our cities? Photo: © SIDEWALK LABS, HUFTON + CROW

Would you let Google run your city?

Google promised to create the neighbourhood of the future in Toronto. Then residents rebelled
January 28, 2019

On 17th October 2017, Eric Schmidt joined Justin Trudeau, the Canadian prime minister, and John Tory, the mayor of Toronto, on stage in Corus Quay on the city’s waterfront. Schmidt, then executive chairman of Alphabet (Google’s parent company), was there to announce a new partnership between Toronto, the Canadian government and Google’s sister company Sidewalk Labs to redevelop a 12-acre plot called Quayside, with a view to extending it by a further 800 acres.

“We cannot wait to get started,” said Schmidt, on building what his company called the “neighbourhood of the future.” Everyone on stage seemed to assume that Toronto residents would welcome the project—and the Google money that went with it—with open arms. They were soon proved wrong.

Sidewalk Labs was aiming to create the city of tomorrow, “with connectivity designed into its very foundation.” Its 196-page proposal envisioned self-driving shuttles, modular housing, rubbish-collecting robots, data-driven public services and what it described as a “programmable public realm.” Technology would be seamlessly integrated into the physical environment: ceaseless streams of data would make the city efficient, adaptable, responsive and ecologically sustainable. The result would be the first “thinking” city, fed with information which would then be crunched by algorithms.

From the outset, Sidewalk saw Quayside as a “global testbed” for urban innovation, a “replicable model for the world.” “What happens in Quayside,” the project vision said, “will not stay in Quayside.”

Yet conspicuously missing from Sidewalk’s proposal was any mention of politics. Developing a new city from the ground up inevitably raises a cat’s cradle of political choices. And the deployment of all that data does not reduce those political questions, but only multiplies them. How would all those streams of data be used to make decisions about public services? Who would control the platform, own the algorithms and potentially profit from the knowledge gleaned? And who would decide what could get built, whether bricks and mortar or digital infrastructure? On these and other political questions the proposal was either silent or else reverted to the passive voice, sometimes wrapped in mystifying jargon.

For example, in relation to something called Quayside’s “platform standards layer”—which, the Sidewalk proposal explains, would define “the rules for residents, administrators, and developers using and building atop the platform”—there was little indication of who would set these rules. The spectre hovering behind the project was Sidewalk’s parent company, Alphabet. What did one of the most powerful companies in the world stand to gain from its development of the “city of tomorrow”? Was its aim, as one resident suggested, to become the city’s digital overlord?

By late 2018, a year on from the grand launch, Sidewalk’s plans were mired in disagreement. Four Toronto women had started an advocacy group, Tech Reset Canada, to campaign for guidelines around the private use of public space and data. Marc de Pape, a tech entrepreneur who had been interviewed for a job at Sidewalk, abruptly withdrew citing its overweening ambitions. “I was talking to a company,” de Pape wrote, “that aspired to export a complete platform for cities, and civics, using Toronto as an incubator.” In May 2018, at a public roundtable organised by Sidewalk at Ada Slaight Hall, a member of the audience suggested its plans might be steering Canada towards totalitarianism. The press picked up on the concern, and from then on the adjective “controversial” was used in coverage of the project.

Sidewalk was naive to be surprised by the level of disquiet provoked by its plans for Quayside. People rarely like to be used as lab rats for corporate experiments—especially ones run by corporations as omniscient as Google. Matters were not helped by the global wave of anxiety about the power of the tech giants and misuse of their platforms that has followed the recent scandals over Cambridge Analytica, Russian trolls, alt-right activists and Macedonian teenagers pumping out fake news.

Download your city

But is there a danger that the backlash has gone too far? Here was an opportunity for perhaps the most data-rich company in the world to tap its unrivalled knowledge and re-engineer a city to work as smartly as possible for its residents. And, rather than deal with the pre-existing jumble of buildings and roads, it could start from a blank slate. All the hidden wiring, physical and virtual, could be the most ingenious ever designed.

What would it be like to live in a Google-smart city? Based on the picture painted by the visionaries, it would not just be different but perpetually changing—dynamic and transient. It will be filled with Lego-like buildings that can be put up, taken down or shifted around overnight. Even buildings that stay in the same place will be multi-purpose. One day you will walk past a grocery on the corner, the next day the same space will be occupied by a health centre. Public spaces will be adaptable so they can host makeshift theatres, transitory music venues, pop-up markets or mobile food trucks. The paving slabs beneath your feet will be hexagonal-shaped so they can be easily removed. The new smart city will be less like a fixed space and more like a smartphone screen, in which the buildings are apps that you can download, move around or uninstall at any time.

The people who reside there will be as transient as the city. Everyone will become, in Theresa May’s phrase, a “citizen of nowhere.” They won’t lay down roots, but instead carry much of their lives around with them like digital tortoises. But this is a weightless world, and so it doesn’t involve anyone lugging objects around on their backs, but rather our virtual lives being carried on our phones and in the cloud. This disembodied information will allow for personalised public services (if that is not an oxymoron), and relieve the need for things like GP surgeries, since we can be treated anywhere via our digital devices. When we travel it will be on foot, by bike, or by self-driving taxi or bus. If you drive your own car—which will be discouraged—you will be nudged, guided, assisted, speed-controlled and, if you err, penalised. It will be a city that goes with the grain of the communications revolution—as frictionless, flexible, convenient and disruptable as possible—where work, home, social life and family life blur and blend into one another.

Data will be the animating force of this city of the future—its lifeblood. Data captured by sensors built into the physical infrastructure. Data from our homes, our vehicles, our phones, and even our bodies. Along with the city’s waste, its water and its energy, this data will be pumped through a vast network of underground pipes: the 21st-century digital descendants of Joseph Bazalgette’s sewers. Of course for this city to function it will need a guiding intelligence. An organisation capable of interpreting and integrating this data in real time. An organisation like, for example, Google.

So irresistible is the dream of building a utopian tech-driven city that another tech titan got in on the act a month after Quayside’s launch, and on an even bigger scale. In November 2017, a subsidiary of Bill Gates’s firm Cascade Investment bought 25,000 acres of scrubland in Arizona, planning to build a new smart city, to be called Belmont.

But Bill Gates may soon discover, as Alphabet and Sidewalk already have, that one person’s utopia is another’s dystopia.

The behavioural psychologist BF Skinner is best known for the “Skinner box,” in which lab rats are taught to do something—like press a lever—by being rewarded (say, with food) when they get it right, and being punished (perhaps with an electric shock) when they get it wrong. In 1948, he wrote a novel, Walden Two that he considered to be a utopian vision but which later readers have bracketed with Aldous Huxley’s Brave New World and George Orwell’s 1984.

As with Sidewalk’s Quayside, Skinner’s Walden Two is a constantly evolving society based on experiment and evidence. Everything has been socially or culturally engineered to make it more harmonious and efficient. From living quarters to dining arrangements, from the way in which people carry food (on elliptical, transparent trays with specially shaped indentations), to the way they can gradually build up their tolerance for bad experiences to achieve self-control. As a result, residents would “get the satisfaction of pleasant and profitable social relations on a scale almost undreamed of in the world at large,” and “they get immeasurably increased efficiency.” Compare this to Sidewalk’s plans for Quayside where “fine-grained information about the public realm and the objects in it enables greater efficiency and quality of life.”

In Walden Two the cost of this satisfaction and efficiency is depoliticisation. Members of the community do not get to vote, or influence policy. Day-to-day, society is run by “the Managers,” specialists who take charge of different areas of life—like food, health, play, arts, education. The only government is a “Board of Planners” who serve a fixed term and then choose their own replacements from names put forward by the Managers. While the 1940s rhetoric of “planning” is alien to the cheery anarcho-capitalism of the 21st-century west coast, the underlying idea—gathering as much information as possible to refine and then optimise every imaginable social arrangement—is the same.

And like any self-respecting Silicon Valley entrepreneur, the founder of Walden Two, TE Frazier, has little time for government. For him, people are happy despite government, not because of it. “Our government won’t accept the responsibility of building the sort of behaviour needed for a happy state,” Frazier says. “In Walden Two we have merely created an agency to get these things done.”

Get political

But back in the real world, politics is not so easily brushed out of the way. Sidewalk’s Toronto timetable has been upended, with publication of its final revised plans pushed back to some time in 2019. Quayside is still likely to go ahead, though its ambitions may eventually be diminished by a thousand compromises. The Sidewalk Labs approach may have proved to be hubristic.

Nonetheless, Toronto and all other cities will inevitably get smarter and more data-driven in the end. From Canada to the Philippines, from Belfast to Bangalore, national governments and civic authorities are convinced that big data can unblock their traffic jams, rationalise their energy use and fix their housing problems, not to mention their benefits systems and even their healthcare.

In all of these cities you will also find transnational tech platforms, fighting for the chance to make the city smarter. Few of them are doing this from a top-down blueprint, like in Toronto, but rather incrementally, from the ground up.

It happens in stages. First, they create an app (in the Uber mould) which makes a public service more convenient and efficient, and offer it at low or no cost. Next, they frantically build up a user-base until they can benefit from the “network effect,” which kicks in when sufficient people are using it to give it the comprehensive coverage that makes it valuable for others to join. Once that base is big enough it becomes not only economically sustainable, but is also protected from official intervention by its popularity.

At a certain point the data starts to give them phenomenal insights into people’s behaviour, and opportunities to influence it. After that, city authorities either try to rein them in or partner with them. Waze, a Google GPS app that has been downloaded more than 100m times, knows people’s traffic movements better than most governments.

It has partnered with over 70 US city authorities, and more than two dozen outside the US, to share real-time traffic and accident data. Uber and Lyft are battling to provide shared transport and public transit solutions in cities across America. Other tech platforms are making similar moves into healthcare, education, energy and housing.

“The most disquieting aspect of the data-driven city is the opportunities it provides for social engineering”

Just a few years ago, city authorities were remarkably relaxed about this colonisation of their civic services—particularly the collection of personal data. This was partly because it had taken them by surprise, and partly because they felt impotent to stop it. But although cities have recently pushed back, as London has done by threatening the operating licence of Uber, they still see huge potential in the integration of tech platforms and civic services: less congestion, more strategic use of resources, better planning. On top of which, the tech platforms have something that most civic authorities do not—deep pockets.

And so many cities are left looking to work with platforms in public-private partnerships, in which—they hope—the municipal authorities can still retain ultimate sovereignty and data control. That may sound like a common-sense solution, but when these partnerships get too cosy, there can be significant risks for the citizen.

The all-seeing eye

To see those risks consider Hangzhou, a city of almost 10m people just southwest of Shanghai. It is home to China’s equivalent of Amazon—Alibaba. In 2016, Alibaba launched a new initiative with Hangzhou authorities to help make the city “smarter.” The commercial platform then started ingesting all the data it could access—traffic movements, public transit information, video feeds, social media chatter. It then connected data sets together to identify links and find patterns. Working towards an eventual product called City Brain, the aim was to develop a neural network that could help Hangzhou function more efficiently. The partnership appears to have been a success—at least from the city’s perspective. Traffic is moving faster and incidents are being spotted more quickly.

The city’s residents, however, are now constantly watched and tracked, and their behaviour directed, rewarded and penalised based on their data trails. Sharpening the political implications is Alibaba’s affiliation with Ant Financial, which is piloting a credit scheme with links to the Chinese government but separate from the state’s social credit system. This scheme has been linked to an episode of Charlie Brooker’s dark satire on the tech age, Black Mirror, because it constantly assesses citizens’ trustworthiness and gives them a Sesame credit score, based on everything from their online behaviour to their friendship networks. A low score can make it harder to travel, prevent you from getting a bank loan, or stop you finding a job. Transgress rules or social norms in Hangzhou and it will cost you much more than just a fine.

Of course we are lucky enough not to live under an authoritarian dictatorship. But the starting point here is not about what the politics allows, but what the technology makes possible. Of course we enjoy certain protections that are not enjoyed by the Chinese, but many of these—for example, those in the UK Human Rights Act—are only protections against the government and the public authorities. Because it was until recently impossible to imagine a private company strangling the possibility of private life, traditional civil liberties are not framed with this danger in mind.

More of the dangers are on show in Singapore, which—on paper—looks like a more directly comparable case, seeing as it has a multi-party system and regular elections. In 2014, the Prime Minister Lee Hsien Loong set out his vision for the city state as a Smart Nation, “where people live meaningful and fulfilled lives, enabled seamlessly by technology.” In practice, this has meant putting sensors in roads, buildings, vehicles, even lamp posts. Its eventual aim is to connect “E3A”—everything, everyone, everywhere all the time. This way, Lee believes, it can ensure its citizens’ safety, convenience and prosperity. Yet, as in Hangzhou, the cost to the citizen is the freedom to go about their lives without being tracked, monitored and managed by the authorities. Dissent, already difficult in Singapore, is becoming virtually impossible. The prime minister’s own brother and sister recently left the island complaining that “we feel Big Brother omnipresent.”

Nor should we assume the Hangzhou smart city model does not travel to other countries. City Brain is already being exported to Kuala Lumpur in Malaysia, and Alibaba has ambitions to take it much further.

In the US, Amazon already uses its hometown of Seattle as a laboratory to test futuristic civic experiments, like the checkout-free shop. And very often, whatever suspicion the tech giants might attract, cities still end up approaching them as supplicants. After a ferociously competitive civic battle to accommodate Amazon’s expansion beyond Seattle, the online retail behemoth has settled on two locations to develop its new headquarters. These two, Crystal City in Virginia and Queens in New York, are unlikely, having won the privilege of hosting Amazon, to deny it the freedom to help make them smarter.

Boxed in

Toronto journalist Brian Barth makes a telling parallel with Amazon’s own Alexa, a virtual assistant device you can buy to control your lights, heating and burglar alarm from outside your home. Writing about Quayside, he observes: “It’s one thing to willingly install Alexa in your home. It’s another when publicly owned infrastructure—streets, bridges, parks and plazas—is Alexa, so to speak.”

Jane Jacobs, the great 20th-century critic of technocratic urban planning, wrote that cities pose a “problem in handling organised complexity.” The question is, who handles that complexity? And if it is all organised by one—or a handful—of huge, data-hungry technology platforms, what will that mean for civic freedom?

For many people the most disquieting aspect of the data-driven city is the opportunities it provides for social engineering. Initially this may be constructive and relatively unobtrusive—being nudged to leave earlier for work, being advised to take a different route, being told to slow down or speed up. Yet as these nudges accumulate they become interference, encroaching on individual free will and collective spontaneity. Eventually, step by incremental step, we may find ourselves constantly watched over by unseen guardians who record, assess, direct and then applaud or admonish our behaviour.

“The city’s residents are constantly watched and tracked, their behaviour rewarded and penalised”

BF Skinner did not believe in free will and was contemptuous of “autonomous man.” He made social engineering integral to the smooth functioning of Walden Two because, he wrote, this was the only way to guarantee social harmony and wellbeing.

This, remember, is the same Skinner who experimented with multiple ways of re-engineering animal behaviour. He explained, for example, how a pigeon could be programmed to peck a disk up to 10,000 times in order to receive food, through a process of infrequent reinforcement. Initially, the pigeon would receive food every time it pecked the disk, then every other time, then every fifth time—and so on and on until it reached 10,000 pecks. An urban area, with enough live data and machine intelligence, would easily lend itself to countless such experiments. Hands up who wants the city of the future to resemble a Skinner box?

This article has been corrected to more accurately reflect the work of Ant Financial. Previously the author described it as running a "social credit scheme" being piloted "on behalf of the Chinese government." We are happy to make clear that Ant Financial is "piloting a credit scheme with links to the Chinese government but separate from the state’s social credit system."