Clockwork City, Responsive City, Predictive City and Adjacent Incumbents
This is the in-depth version of my column posted in Dezeen on 30 October 2014, around the impact of predictive analytics on cities. This version was posted at cityofsound.com on 3 November 2014, and particularly uses new transport startups as the pivot for its arguments, as transport (or transit, or mobility) is a fundamental aspect of city services currently being transformed, disrupted and contested through such dynamics. The arguments get usefully tangible when we’re looking at Uber, Lyft, Bridj alongside MTA, Transport for London and MBTA. This also features a bit of a Q&A with Bridj CEO Matt George — I’ve posted a fuller version of that separately. I now realise that, completely unintentionally, this is a follow-on to a piece on ‘transport informatics’ I posted around seven years ago.
For the last 150 years or so, we’ve run our cities like clockwork.
I don’t mean that as a compliment, a suggestion of flawless efficiency. Just that we’ve designed, planned and run our cities based on regulated industrial rhythms, bound to pre-digital engineering and organisations, and we still do.
We expect a rush hour at the beginning and end of work-week days, and planners intensify mass transit at these times along major arteries, usually into the middle of cities via a form of ‘hub-and-spoke’ model. Citizens must move towards the nearest nodes in that network — the bus stop, the metro station — rather than their actual origin or destination, and these must necessarily be organised along averages of demand.
These patterns are in-part derived from mass industrialisation, and its physical impacts, and the 20th century urban planner’s instinct to separate functions like retail, offices, housing and industry into different zones of the city.
These days, however, not only are we now trying to create ‘mixed-use’ urban environments, dissolving zones left, right and centre, but many of our patterns of working are fragmenting — whether that’s through zero hours contracts or the burgeoning freelance sector — as are many other patterns of living, generally.
But those clockwork patterns run deep. Few western cities look like a Lowry painting anymore. The factories have gone, the workers have gone, the tramlines that delivered them have often gone too. Yet traffic still tends to runs along those now-buried lines, even though the route’s raison d’être has long since departed.
Our bureaucracies are also based on processes that would not be all that unfamiliar to the Dickensian clerk hunched over reams of paper in stygian gloom, shuffling applications, plans, appeals, and accounts back and forth. Those processes have sometimes been digitised, yet the Estonian president Toomas Hendrik Ilves has suggested that the shift from digital to paper is not just a shift of format, for the same processes; he says means completely ”redesigning government and how you interact with people.” This probably applies to to municipal governance more than national or federal, given their remit in actually delivering services, yet few, if any, cities have redesigned accordingly.
City policies and services are still derived from planning around averages and rounding down, informed by a muddy stew of samples, snapshots, censuses and ideologies, and delivered en masse, one size near-enough fits all, and then spotily measured or policed, to see what actually happened. Sometimes.
There are two ways this may about to change, as cities begin to address Ilves’s challenge — by enabling services on-demand, and by using data to predict the need for services in a particular place at a particular time, with great precision (allegedly.)
Although replete with thorny issues, a potent combination of predictive analytics with responsive services may prove irresistible to politicians and policymakers, and indeed could deliver genuinely positive transformative services. But this air of inevitability makes it all the more necessary for us to stop and think about what kind of city we might want, and who takes advantage of these new dynamics.
John Lanchester tells a story in “How to Speak Money” — a book I recommend to any designer or architect wishing to understand the context their work is produced in — about the way Ancient Egypt worked. In short, everything hinged on the annual inundation of the Nile flood plain: the society itself, arguably the most stable the world has ever seen, its cultural artefacts, such as their calendar, their understanding of seasons, their taxation system, and of course their agricultural cycles — all were directly linked to the Nile’s flood. The priesthood of Egyptian society, drawing from rich mythologies, performed complex rituals to divine the nature of the flood, and thus the harvest, each year.
But Lanchester reveals how the priests actually did it: they were cheating. Unbeknownst to the population, they had a ‘nilometer’, a device to predict the level of flood water. Based on measuring stations secreted in temples, the nilometer captured the flow of the river, plotted it against various markers and combined with flood records dating back centuries, enabling relatively accurate predictions of that year’s harvest — its success, or the likelihood of disaster. Lanchester suggests “the nilometer was an essential tool for control of Egypt. It had to be kept secret by the ruling class and institutions, because it was a central component of their authority”.
Not that Herodotus would’ve put it like this, but this is perhaps the first example of predictive analytics in urban governance.
Predictive analytics is the ability to deliver services for *future* events, before the need has manifested itself, based on the accretion of ‘big data’ about past events, increasingly derived from a sensor-rich environment. Although urban planning and policy has always been a form of prediction — sometimes combined with agency to make it true — this is an order-of-magnitude shift in data gathering and number crunching; and so, in turn, in the purported accuracy of what can be predicted.
The same principles are in play across an increasingly wide range of services. Based on the data they have, Amazon can apparently guess what you’re about to buy before you know you want to buy it, and will move it nearby accordingly (they filed a patent for ‘anticipatory shipping’ earlier this year.)
Yet when people talk about predictive analytics and cities, they often turn to Chicago’s rats. Combining many previously disparate data sets, such as calls to the city’s 311 service related to garbage or broken water mains, over geospatial databases — aka maps — has enabled the city to predict where infestations are likely to occur before they do. Just as Mr. Rat has paid the deposit on a new place and picked up his keys, he turns the corner to find a Chicago Department of Streets and Sanitation officer leaning against the door, whistling.
This idea, and its impacts, could fundamentally change the way we think about running cities generally. Using similar techniques, and digging ever deeper into urban data, Chicago could soon predict which buildings are most likely to catch fire next, where vacant properties will occur, or where planning violations might be likely.
More controversially, Chicago’s datasets include a ‘heat list’ of those 400 or so individuals in the city that are most likely to be involved in a violent crime — it enables police to pre-emptively, well, ‘address’ those on the list, which some see as an exaggerated extension of ‘Stop-and-Frisk’.
(You may be familiar with the concept from the film ‘Minority Report’. Incidentally, isn’t it odd that this single film — which, after all, is not a terribly good — has dominated the popular imagination about interaction design for the last decade, thanks to John Underkoffler’s work on gestural interfaces for the movie; and now it may dominate the conversation around system design for the next decade. Almost as if the filmmakers could predict the future.)
The ‘heat list’ has come under much scrutiny in Chicago (and LA and New York, where similar software has been deployed) but few realise that these systems, like PredPol, are increasingly in use everywhere, from Seattle to Kent. Here the difference between pest control and policing should be clear, but you can imagine the attraction to a certain kind of politician — “Sir, you can eradicate crime, nipping it in the bud before it happens” — irrespective of the enormous ethical, and indeed constitutional, issues. Once the list exists, it’s difficult to put that particular genie back in the bottle. What if you don’t act upon the list and someone dies? Equally, what if you do act upon the list? And how exactly do you do that?
(Interesting to note how recent this is; there was little direct mention of such techniques in the talk by Michael Downing, Deputy-Chief of Counter Terrorism for LAPD at Postopolis LA, despite a lot of tech in his talk. The antecedents were visible, perhaps, but could a deep culture of practice have emerged in that time? Besides, at some point, once things are scaled and pass certain thresholds, they become something else altogether: just as Jane Jacobs said a city is not just a big village — it is an entirely different condition — predictive policy is probably not just exaggerated stop-and-frisk.)
Perhaps less immediately contentious is the use of predictive algorithms in public transport. If transport agencies can ‘scrape’ data from the surfaces of their city’s interactions, they can build models of behaviour that enable them to predict where the demand for transport is needed, before anyone asks for it. In other words, instead of walking to the bus stop, the bus stop comes to you.
These dynamics underpin Bridj, a transport startup in Boston. Bridj uses patterns of transport use, combined with social media analytics and apps, in order to send its fleet of buses to where there is demand for a fleet of buses — on the fly. It is largely post-timetable, post-route and its founders have just attracted around $4m in seed funding.
Techcrunch reports, “Bridj leads to sub-10 minute wait times, as well as much faster commutes than its passengers are used to. … Bridj routes can cut commute times in half, with 20-minute rides compared to city transit routes that would typically take 45 minutes. It also generally picks up customers closer to their homes or businesses.” (Note many non-US cities achieve sub-10 minute wait times anyway; of course, services like Bridj emerge in a certain local context.)
Bridj seems a sharp approach, sitting neatly between the approaches of mass transit and private car ownership — as the name suggests. It could work as a form of “relief valve” for the Boston MBTA, and equally would enable those awkward transverse routes, threading together the city outside of the ‘hub-and-spoke’ model that are simply too expensive for public transport to cover.
I spoke to Bridj’s CEO Matt George about their service. George suggests they’re additive to cities, rather than subtractive, but working in a different way:
“On one end of the spectrum you have low-cost low-flexibility services like traditional mass transit. On the other end of the spectrum is high-cost and high-flexibility options like owning a car or using Uber. We are looking to be a third option — moderate flexibility and moderate price — that we think captures most of the needs of city travellers. “
Where have we seen on-demand transit before, outside the limited horizons of US tech culture? Interestingly enough, many of the more useful reference points, positive and negative, for these developments occur in more informal urban environments.
Keiichi Matsuda tells me about the gloriously-festooned buses that careen around Medellin, essentially occupying a legal grey area as well as often unpredictable street routes, and sometimes, the pavement. In Nairobi, the equally gaudy matatu buses are fighting off Google NFC-enabled smart cards, preferring to transact in cash — whilst matatu may observe a form of bus-stop, they move through traffic at full pelt as if autonomous vehicles (though with a rather different safety record.)
Even highly-regulated cities like New York have what The Verge called “a shadow transportation network” of dollar vans and buses, serving areas like Chinatown, or particular communities. Often unlicensed, sometimes the police apparently turn a blind eye, and even welcome their presence; other times, not so much.
These services fill in the cracks and gaps of the formal transit networks in a broadly similar way to Bridj, yet based on driver knowledge, instinct and ‘small data’, if we can call it that. The fact they don’t scale doesn’t really matter, and in some senses they are more legible, local and, well, likeable than an Uber, say. (Incidentally, in one of its few genuinely innovative moves to date, Uber has also started using predictive analytics — Bayesian, since you ask — to predict pick-up points.)
Yet the bus-stop that moves to you via predictive data is something else again: a broader idea of a Responsive City. (Though it does also remind me of the time I informed the council of an unnamed Northern English city about Barcelona’s responsive litter bins; “They just tell you when they need emptying, so you don’t have to send out the bin-men when they don’t!” One of the councillors muttered, “It’s all very well having responsive bins, lad. Problem is, we don’t have responsive bin-men.”)
Both the matatu and the metro could be derailed by startups like Bridj, Lyft, Uber, Relayrider et al, however. None of these startups have a primary aim of putting public transit agencies out of business — least of all Bridj, who George says have a clear mission of working in the gaps — yet the adjacent space they play in is close enough to destabilise those incumbent agencies, given the way that startups play.
This is ‘The Problem of the Adjacent Incumbent’.
It could be that Uber et al seriously destabilise existing public transport agencies, simply by working an adjacent patch to them, mopping up bits of their business without having to work with their constraints like universal service.
Here, we might describe universal service as the delivery of a consistent transport service at a consistent, affordable price to everyone in the city, irrespective of net worth or location. (I would personally frame this as a wonderful ambition for mobility services in a city.) Transport for London, for example, has to deliver that, whilst innovating; Uber, for example, simply gets to pick off the low hanging fruit in the middle of town, moving freely as a 21st century mobility business, floating across taxi-likes, hire-cars, delivery services and ultimately minibuses and privately-owned cars generally. Given half a chance, however, Uber’s trajectory will envelop most mobility in the city, not merely taxis.
Google is not the same as Nissan (or equivalent), but its self-driving cars may begin to place it very close indeed. So they are also adjacent. Thus, Google (or equivalent) do not have to build a 20th century automobile business to replace one — they can simply build a self-driving car industry, in the far smaller numbers such a service requires. Like Uber, they exploit the redundancy in private car use via a Responsive City combination of on-demand and just-in-time.
Similarly, yet in a different line of work, Airbnb is adjacent to hotel chains like Hilton and Hyatt, yet in theory not the same thing at all. (In theory.) Airbnb does not have to build a 20th century hotel industry to replace one — they unlock pseudo-hotel rooms via software rather than laying bricks. It enables a Responsive City approach to urban fabric.
All these initiatives could indirectly destabilise incumbents that are not theoretically “in their sights”, simply through their powerful dynamics.
For tech startup culture plays by its own rules, and is a little careless and untidy about how it does so.
Richard Barbrook and Andy Cameron described those rules as ‘Californian Ideology’ dynamics, and they typify tech startups. In this context, it means such ‘disruptive innovators’ may try avoid other ‘constraints’ over and above universal service’s equitable agenda — such as aspects of local legislation (especially around workers’ rights) or local taxation (eg.), just as their venture capital-fuelled trajectories mean that they may instigate temporary price wars in order to gain early market domination, subsequently raising fees later. These dynamics create issues. (Indeed, Uber has just suffered its first strike, alongside reports of long working hours, difficult working conditions, low wages, an exploitation of a wider economic context, and even mishandled retweets from drivers. This, and their ongoing tussles with city regulators, may be why Uber hired Obama’s campaign manager. Suggestions that these services self-regulate themselves are a little off the mark as a result, and all this means that the phrase ‘sharing economy’, and its connotations of trust, is a terrible bit of ‘newspeak’ for something like Uber. Uber is a large venture capital-backed corporation. The only thing you truly share with the driver on an Uber ride is the attempt to locate your destination.)
These are what economists would call ‘externalities’, which is usually econo-code for something not terribly good. (Do read Cameron Tonkinwise’s informed take-down of the “sharing economy” more generally, noting also the last vestiges of possibility there, as well as the externalities.)
Matt George is aware that Bridj is playing with slightly different dynamics than those transit agencies:
“What we offer is the ability to start with a clean slate. We are an incredibly high-paced technology company who focuses exclusively on creating a better demand-responsive mass transit system. There are some small agencies that we think can implement better demand response, however, most agencies are trying to simply maintain the service they have, and have little capacity to try new things.“
Though Bridj’s service is designed to complement rather than replace, the lack of a level playing field means that unhelpful externalities could arise either way. The challenge is to sketch out the externalities and mitigate against them, to enable and yet also shape such innovation.
Another example: Urban Engines is a newish startup which also collates data from transport operations, and then uses an understanding of how congested the network is, in real-time, in order to offer small payments to people to ride later, shifting the load across the network better. A good idea, at first glance. But on second thoughts, what does this market-based approach do to the core idea of public transport?
Such an offer is all very well for knowledge workers with what Will Hutton called ‘time sovereignty” — the ability to choose when to work, and to move your day around. They can now get paid for exploiting their own time! But what if you’re a low-paid service worker on a zero hour contract, having to get a bus from the outer suburbs at 5am in order to clean a CBD Walmart store before it opens? You hardly have the choice to displace your travel, and so you pay whatever price the network demands.
Urban Engines is trying to solve a Real Urban Problem many of us are familiar with (over-congested mass transit) and, speaking as a regular London Underground user, many of us desperately do want it solved. Yet the way in which this is done could either reinforce public good, generating a cohesive civic spirit — or rent it asunder. Currently it’s another example of brute-force market dynamics being deployed into areas which have traditionally had rather more subtle toolkits at hand. It is potentially socially-divisive as a result.
Equally, a recent NYT account of Uber use in Los Angeles suggests it has little to offer as a public transport innovator across a wide range of citizens. In fact, it suggests it’s for well-off kids on nights out in a few highly populous parts of an American city where, typically, the car has been king — unlike most other cities outside the USA. This is a long way from an innovative public transport offer — and from its claims to be able to really shift the mobility dynamics of a city, as with their recent statement about removing one million cars from London’s roads. Frankly, as with Google’s self-driving cars, swapping private cars for a different flavour of private cars is unlikely to transform much, compared to the possibilites of other modes, from bikes to buses.
Yet Uber is frequently held up as a key contemporary urban innovation, including by policymakers and others in European cities with excellent public transport. The lustre is extraordinary. Equally, in terms of ‘following the money’, watch how Uber’s founders talk about the potential market size. One way or another, Uber is seen as a future of mobility in general, despite its inability to deliver to the strategic goals of the incumbents it may destabilise.
So now would be a good time to pause to think through — to design through — the unforeseen implication of predictive analytics and responsive city services. What might we gain and what might we lose? It doesn’t mean we have to remain trapped in our clockwork cities; just that we need to try to unpick the unforeseen and adjacent. We might want to hang on to some of these precarious incumbents, clockwork or not.
Perhaps predictive analytics applied to crime can cause the rate of violent crime in a city to plummet? Yet might the way it does this also shred social fabric? Perhaps Airbnb enables redundant space to become temporarily valuable. But perhaps in doing so it puts big and small hotel chains out of business, leaving only a largely unregulated offer in its place, and further reinforcing the idea of home as financial commodity rather than, well, home. Perhaps predictive analytics applied to transport creates a nifty little service like Bridj, but when it is combined with Uber, Lyft, RelayRider, Urban Engines et al as well, all swirling around the city with those new dynamics, perhaps it destabilises public transport to the extent that a universal service is no longer viable? These implications are adjacent, slippery, opaque, and laden with assumptions about the positive effects of data-driven services on the city.
(Incidentally, that most acute social commentator, South Park, just skewered all these new transport startups. Wacky Races indeed.)
Modern service dynamics
Uber is popular partly because, on a few levels, it is doing an excellent job. Leaving aside their premium service, where status is an indicator and so not particularly relevant to an equitability argument, Uber’s ability to eat into the taxi business is due to their understanding that a good user experience is a differentiator. People expect it, quite rightly.
In this case, that means the service should be responsive, location-aware, personalised, well-designed and reflexive (in other words: cars should be quick to arrive on-demand; locations of passenger, vehicle and destination handled easily via GPS-enabled smartphone; payment should be seamless; both user interface and driver should be courteous and effective, and the car should be clean, attractive and potentially low-emission; and customers might want to leave feedback about their experience, and consider the service level an explicitly valuable part of the transaction.)
These, and related, qualities may typify how many people expect most systems in general to behave now. Technically, Uber is a sketch of how urban mobility should be, to a large extent. The vehicle comes towards you, when you need it; the transactions, digital and otherwise, are seamless and reliable; the route is bespoke to your needs, and adjustable in real time.
I doubt customers care much whether it’s Uber — again, outside of their premium service customers. It simply provides a reasonable user experience at a reasonable price (surge pricing aside). That is not actually innovative in terms of service. While good user experience is not easy, it is at least well-understood now. It takes rigour, but not necessarily breakthrough innovation. It differentiates nonetheless, compared to existing transit offers — and that means something (leverage, in fact.)
But Uber’s UX could be improved; it does the job, but no more. Their primary innovation is in skirting ‘constraints’ such that they can play with market dynamics. (Indeed, the roughly contemporaneous Hailo does essentially the same thing, though within admittedly self-imposed constraints. Set up partly by London cabbies, Hailo has now been forced to pull out of North America due to being caught in the crossfire of a deregulated price war between Uber and Lyft, according to the FT.)
Given that, there is every chance for those qualities of 21st century service to be adopted by public agencies. Seoul recently ‘banned’ Uber, following the lead of an increasing number of cities such as Brussels and Berlin, and announced they plan to simply set up their own version. What is to stop any reasonably well-funded urban transit agency from developing their own high-quality digital services; essentially replicating, or even improving upon, the user experience of Uber and equivalent? Uber is a good UX, a suite of lawyers and lobbyists, and almost insane levels of ambition. A strong city government digital team could also achieve the technical standards (it’s strong craft work, but that’s a known-known, and well understood practice); and they don’t need the lawyers and lobbyists because they are the legislation.
So the stumbling block is ambition, after 30 years of being told they can’t innovate.
Yet the UK’s Government Digital Services has shown that it is entirely possibly to substantially improve upon private sector digital practice, never mind replicate it, from within the public sector. Now apply this to public transport. (This relates to an earlier Dezeen column asking whether, or how, public agencies can invent or adopt such service innovations.)
In this way, local services can be tuned to local needs—and only a small fraction of the population frequently takes public transport across multiple cities (and if they did, this is likely to be within a regional economy; hence the ‘Oyster for the North’ conversations going on across British northern cities.) There is a mismatch between the easy replicability of code and the distinct differences of cities, yet cities are increasingly able to adopt and adapt code-based services.
And adaptation is important, as there’s a limit to how far we can stretch the idea of California. Uber emerges from a particular place, with what we could describe as a largely uncivilised approach to public transport, and a sclerotic approach to regulating and operating the taxi business. These conditions are not the same everywhere, particularly outside the US, and Uber’s mileage may vary as a result, despite their hefty capital. That also leaves an opportunity for local variants, adaptations.
Local culture is what a city is, and is also damaged by an adjacent homogenous global offering. As I wrote in my Dezeen column on Uber, their movement, as with Amazon, can be seen as another cultural blitzkrieg, obliterating difference and leaving high-quality homogeneity in its wake. With clothes and coffee it’s a shame, but not that big a deal. However, when it ploughs into a core urban service like mobility, it rewrites the stories that make a city.
Although taxis are a form of privatised transport, they remain part of the city’s civic infrastructure, part of their character. As architect and teacher Robin Boyd wrote, “taxi-men teach the visitor a lot about their towns, intentionally and unintentionally.” Boyd was able to to demarcate Sydney culture from Adelaide culture based on whether the cabbie opens the door for you. I recall scribbling a drawing of a Stephen Holl building I wanted to visit in Beijing, as my only way of communicating my desired destination to the taxi driver. (See also ‘Three taxis’.)
Uber makes transactions easier, but what we gain from a seamless UI, and the convenience of the global currency of apps, we lose from the possibility of understanding a place through a slightly bumpier “seamful” experience. What would a local UI be like?
Similarly, Google’s self-driving car prototype is predicated largely on their map data, as well as their sensors, and happens to have been tested on the bit of the world they happen to have mapped in almost microscopic detail ie. around Menlo Park, California. Will Google’s car even function in places where it is uneconomic to map? (Note that other autonomous vehicle prototypes preference sensors over maps, such as Oxford University’s RobotCar — could we even say that is a more equitable approach?)
Uber says that regulations are outdated as they were written before we carried smartphones around. This is true. The question is what one does about it, and further, what ideologies underpin such decision-making. As it says on the Uber blog, “Uber is fundamentally a marketplace”, and exonerates itself from much collateral damage by portraying Uber as being simply a technology platform for connecting riders and drivers.
The “fundamental interconnectedness of all things” in today’s complex urban dynamics means that there is no such ‘simple platform’. So the impact of Predictive City and Responsive City is not simply in the development of new services, but how they interact with the existing services in the city, or create value in the city. It’s also about what they stand for. These advances coud have outcomes either way, of course — so a larger question I would have, over and above their short-term market share-grabs, is about the possibility for enriching the idea of the city as a public good.
When one of the primary challenges facing our cities is inequality, that must surely be a core concern. But is it? The value that Uber generates, outside of their generic if professional user experience for ‘riders’, is leached from the city. The low-paid jobs, in the form of the drivers, remain the city — though the Californian Ideology dynamic tends to suggest that wages drop until they are replaced by automation, and in this case autonomous taxis — whereas the high-paid jobs (the coders, but really the handful of investors) exist only in California, ultimately.
Compare that with a scenario of a similar service — again given that the core service features are entirely obvious, and replicable — deployed by a local public transport agency or by a local startup employing locals in a range of valuable jobs, from coding to investing to driving, as well as paying taxes in the city? Better service, and better for the city too?
Put bluntly, there is, or was, little technically to stop Transport for London doing a version of Uber before Uber did, or instead of Uber. Perhaps even now. The issue is political will and belief systems, rather than the ability to design quality UX or code predictive analytics. But the impact of TfL not doing that could be profound: if I spend a pound with Uber in London, under the trajectory above, most of that pound increasingly disappears out of the city to a few deep pockets in California (and almost all of that pound does, under autonomous mobility.) If I spend a pound with Transport for London, most of that pound stays in the city and its surrounds (TfL reinvests its revenue in its services, locally.) I’m using Transport for London purely as an example here, but I know which scenario I’d prefer, as a citizen of London.
The motor car and the traffic jam
There are numerous possibilities for positively using big data, and data-driven services in the city. The Clockwork City is from another age, and much of it is no longer fit-for-purpose in terms of addressing complex, interdependent 21st century issues and opportunities. Many of Chicago’s early advances show huge promise, and services like Bridj could revolutionise mass transit, actually finding new patterns in-between ‘mass’ and personal. Imagine if it was combined with the data that Citymapper already has, for instance? Who wouldn’t want bus-stops that were more at-hand, with buses arriving just-in-time?
So the design challenge is to better understand and shape the ideologies underpinning network urbanism, to combine a sophisticated understanding of rich urban data with a holistic, collaborative approach to design and governance, to see that data is a material that must be fashioned into particular services for particular places.
An excited Harvard Business Review notes that:
“It’s still pretty amazing that we can use analytics to predict the future. All we have to do is gather the right data, do the right type of statistical model, and be careful of our assumptions.”
Be careful of our assumptions, indeed. A model is still a model, at the end of the day. The financial crisis of 2008–2009 (and beyond) has already given us a potent example of poor predictive analytics in action — in that case, when assessing how likely mortgage customers were to repay their loans.
When we look at a picture of, say, subways and sewers, and their tangled knots of pipes, we should really note that the picture “n’est pas une pipe” — it’s just a picture. The map is not the territory; not even if that map is as richly detailed as the kind of deep behavioural data our cities may soon generate. Whilst the notion of ‘path dependency’ obsesses urbanists, a city is more than just the sum of its previous behaviours; just as a former violent criminal might not ever be violent again.
Fundamentally, with this Predictive City in mind, the sheer unpredictability of cities is not only part of their charm, but a vital lesson about the possibility of change. As Oscar Wilde said, “every saint has a past, and every sinner has a future.” Predictive analytics, if applied with a carelessness Lady Bracknell would recognise, has no time for such subtleties.
Such approaches needn’t necessarily be detrimental — they could be highly informative, in the context of more holistic, collaborative, imaginative approaches to designing and running cities — but only if, unlike the priests of Ancient Egypt and their secret nilometers, we are openly discussing the potential pitfalls — the possibility that ironing out unpredictability also irons out difference, and the possibility of change itself — and sketching out richer versions of our urban future.
There’s a saying in design futures circles — when you invent the motor car, you also invent the traffic jam. You gain; you might also lose. But no-one envisages the traffic jam because … HEY LOOK A MOTOR CAR!
We might borrow the informal dynamics of Nairobi’s matatus or New York’s dollar vans and wrap them up in scalable code structures, in the gleaming business models of tech start-ups. Yet in doing so we potentially destabilise not just the matatu but the metro as well, and we stumble towards whatever invisible version of a traffic jam is contained within these scenarios.
Ironically, it may not be a traffic jam — perhaps software can fix that, after all — but what will it be?