For a couple of months, three small potholes marred my very small street. It turns out that there's a form on the council website: if you upload the address and a picture, they can decide whether to fix it. A few days later, a white ring appeared around each of them, and soon after that they were filled. Great service, you might say, and I would agree with you.
But in The Smart-Enough City, Ben Green points out the limitations of this approach. Yes, my potholes got filled. But at whose expense? Is the metrical simplicity of KPIs based on counting repaired potholes diverting the council's attention from a greater need -- harder to solve and quantify -- elsewhere in the borough? Unless I go to a council meeting I will never know. If getting potholes fixed required a phone call to a harassed council worker, I might grasp the wider picture.
Fixing potholes does not change unfair power structures. And yet: in the early 2000s, when mySociety pioneered this type of service, it was an amazing step forward. No-one wants to say, well, maybe it was a bad idea; it just shouldn't be the only idea.
In The Smart-Enough City Green, a PhD candidate in applied mathematics at Harvard, has written a welcome antidote to most smart city books. Instead of looking at cities through what Green calls 'technogoggles' and technology critic Evgeny Morozov has called 'solutionism', Green advocates starting with citizens' values and the changes they would like to make, and then harnessing technology in the service of those stated goals and principles. Make the city smart enough to serve those values, but don't make 'smart' a goal in itself. It seems so obvious. And yet...
Remove the 'technogoggles'
Early in the book, Green provides a pair of images that make as good a summary of what's wrong with wearing technogoggles as you could want. The first is a rendering of a downtown Boston intersection from MIT's Senseable City Lab, showing how smart autonomous systems can make traffic flow through it smoothly. The second is a picture of the actual intersection on the ground: instead of those clean, white lines and smooth passages for those tiny cars, there are buses, cyclists, and pedestrians -- some in crosswalks making cars wait to turn. You also have, as Green does not note, shops along those sidewalks that would be bankrupted by the disappearance of foot traffic.
Green's tour of the problems with smart cities encompasses predictive policing, algorithmic discrimination, and exclusionary design choices. Among positive examples he cites Brazil's 1980s participatory budgeting. He concludes with a set of five lessons and principles to guide city authorities and concerned citizens. Toronto's ongoing clash pitting citizens and privacy advocates against a partnership that has Google sibling Sidewalk Labs booked to redevelop a section of the waterfront provides a perfect case study.