September 12, 2022

Blog All Dog-eared Pages: Undoing Optimization by Alison B. Powell

Undoing Optimization by Alison B. Powell is an excellent primer on "smart cities". It starts by looking at the evolution of them, then dives deep into problems with their default framing, and what we could do which would be better.

I loved how this book opened. Such strong opening paragraphs which brought the subject to life with a simple anecdote. Here are the sections I highlighted whilst reading it...

Page 1

Then we hurry across, because this signal doesn't usually last long enough for a child to walk across the road before it switches back to letting the cars, trucks, and buses through. When we get to the other side, my daughter asks, "How do they decide when the lights should change? Is there a person somewhere who does it, or is it a computer? Do they make them change at different times when there are not so many cars?"

Page 38

"We are now selling the smart city. It is an economic development tool. ... Wi-Fi is becoming the order of the day. It is like having better sidewalks—if you ask a city if they would rather have bad sidewalks or better sidewalks, they will always say they want better sidewalks. The overall principle doesn't change."

Reminds me of my thoughts on "superfast broadband" from back when I was invited to Council tech strategy meetings...

Page 55

This overall process of optimization narrows the frame for citizenship, individuals are perceived as consumers who can be nudged to change their behaviour based on predictions extracted from data they share.

Page 62

For example, smartphone apps can suggest straightforward walking routes to unknown destinations, make it easy to find restaurants similar to those you have already eaten at, or predict accurately when the bus will arrive. Of course, the logic of predictability and, indeed, possibility have deeper social consequences—because the big data produced in smart cities requires "small analytics" that, by aggregating, parsing, cleaning, and ordering data, enact particular social assumptions. This cleaning, ordering, and parsing is needed to make a cybernetic feedback system return optimal results, but the process removes most of the traces of friction and difference that are part of the reality of urban life.

Page 73

machine-learning techniques require a certain amount of stability in order to generate predictions.

Page 74

It is relatively straightforward to optimize transportation or the collection of recycling but more difficult to optimize volunteering, knowing neighbors, or creating local capacity to take care of people in a crisis.

Page 91

But seeing the city as only a marketplace reduces citizens to consumers and makes data advocates the critics of an administrative state that technology industry pundits like O'Reilly think should be replaced with more efficient platform-based marketplaces. For example, in the years after the founding of the ODI, it began to advocate for opening up commercial data as well as government data and to act less as an advocate and more as a networked broker of data-related discussions and connections, employing commercial and business justifications to advocate for businesses to release data that could be examined and reused by others. This shift over time illustrates how it is often easier for open data to be understood and justified in relation to innovation and open markets than it is to conceive of it as a shared or collective resource.

Page 95

Faced with the reality of class-based and racial discrimination, data do not automatically generate the kind of stories that garner attention and data that questions the decisions of the powerful can still be ignored.

Page 96

Being a civic intermediary for published open data is a significant responsibility, requiring advocates to strike a balance between the risks of publishing and the benefits of transparency. Organizations, including governments, can be defensive about making data publicly available and worried about misinterpretation. An interviewee said, "The way you deal with that is make sure you publish all the context that goes with it. So, you describe how it's been collected. You want to help people understand where there are limits in using the data, ways that you can usefully use it or ways that you shouldn't use it."

Reminds me of my well-worn Usman Haque quote: "It's not about making data public, it's about the public making data."

Page 100

The contracting standards that Open Digital Service Co-operative created include standards intended to prevent data from suddenly becoming the property of a different company after the takeover of a government service-provision contract. The detail-oriented excavation of the issues underlying the brokerage of data in smart cities shows the high level of political and technical skill required to carry public interest requirements into the heart of the function of a platformed government. For community advocacy using data to be successful, that data needs to be accessible, comparable, and usable, and it must remain in the public realm.

Page 127

For data to be used to address issues of justice, processes need to be put in place to make the measurements actionable and, furthermore, to link the collection of data to the problem defined in the first place. This is a more complex process than simply collecting data or even working through it in a hackathon; it depends on the community advocacy and capacity building that were part of the effort at building the commons but that otherwise don't fit the city-as-platform framework.

Page 137

Geese were originally misrecognized as shopping carts in water quality sensing applications in Oxford. Thinking about misrecognized animals also helps us to start thinking about the way that people are misrecognized and about how the emergent properties of "design" strategies for data-optimized smart cities also cause misrecognition.

Page 163

The tensions in datafication show that power and agency are always at work in influencing who can speak, be heard, or act in relation to things that matter in the places they live.

Page 171

The smart city in its data-based version measures everything and optimizes the processes that can be best represented through measurement. There are civic efforts to influence the consequences of such datafication by challenging the knowledge meant to be held by the data or the ontological power of organizing reality that smart cities promise, and these efforts can shift ways of thinking and acting. However, they still reinforce and reiterate the idea that a city is a system to be measured and made knowable and that data accrued as part of this measurement is a material that can be gathered and stocked.

Page 177

It means following a strategy of minimizing, rather than maximizing, this kind of data, and it means seeking to employ decision-making strategies that may appear to be more costly on the surface but that leave space for different kinds of knowledge, as well as for data to decay over time, for frictions to be identified and addressed, and for different forms of democratic participation and accountability, including but not limited to data audit, sensing citizenship, and autonomous networking. It means leaving room for determining what can be known, claimed, and acted upon outside, against, and within the data.

Posted by Adrian at 01:44 PM | Comments (0) | TrackBack