Chapter 6 – Intro and table of contents – Chapter 8
Chapter 7 – The Reality Business
The Prediction Imperative
While for a time, it was thought that the future of the internet was a virtual reality that would replace the real world, the vision of Eric Schmidt and other Surveillance Capitalists is that the internet will weave itself into the very fabric of our normal reality. This solves an important issue: having the data is no longer enough in order to generate surveillance revenues. The prediction models need to be better. And morphing into everyday life essentially enables the prediction products to become more accurate. First, by enabling economies of scope: data can be extracted from a wider array of surfaces (more surfaces than in the online world, like in cities, in our bodies, in our beds, etc…) and it is extracted in more depth (ever more personal and intimate data). Second, by enabling economies of action: the ability to force the prediction models to become true, by nudging users and getting them to behave in a certain way that fulfills the prediction.
In that process, extraction becomes intertwined with execution, producing a modern “means of behavioral modification,” with the commercial goal of transforming a business’ “guaranteed levels of performance” into “guaranteed outcomes.”
The tender Conquest of Unrestrained Animals
Zuboff focuses in this section on R. Stuart MacKay, who applied biomedical telemetry research on animal populations in the Galapagos in the 1960s, in a way that enabled him to collect and analyze large amounts of data about said populations without their knowledge, and therefore without them modifying their behavior (like they would in a zoo, for example). And by the early 21st century, the system has been scaled up, with a continuous inflow of data about animal populations around the world, with a twist: scientists surveil both animal herds AND individuals. MacKay had also dreamed of making the data flow bi-directional, with an ability to influence the animals’ behavior for specific goals (preservation, reproduction, etc…). Zuboff argues that surveillance capitalism modifies this system in two ways: the goals are related to profits, and the subjects are not anymore limited to non-human creatures.
Zuboff then switches things up to address the MIT MediaLab, and its work on embedding sensors into our physical world – ubiquitous computing in its purest form. The scientists, artists, and designers there imagine a world where billions of sensors capture a wealth of information every instant, which could then be organized in a similar way to how Google organized the web’s information. The crawlers are now sensors, and the new application layer (the browser of the digital world) needs to be invented. But once that will be done, humans would be that much closer to fusing into their surroundings. That is, if surveillance capitalists don’t grab all of this data, of course.
Surveillance Capitalism’s Realpolitik
In the world of ubiquitous computing, surveillance capitalism firms are also advancing unencumbered by policy or governments. They are left with a new rhetoric to impose on the world: dark data. Any data that has been captured but not yet analyzed is dark, unknown, scary… And it becomes a given that its purpose is to come into the light by stepping out of the unknown: the notion of dark data provides the justification for complex machine intelligence systems to analyze behaviors that were never intended for public life. And by undergoing this analysis, the processed and transformed bits of data are stripped of any of their social, moral, political undertones. In the name of better predictions and profits, individual elements that should be attached to real people and complex stories become a dump of cold data devoid of any context.
Certainty for Profit
We have now reached the second state of things predicted by Hal Varian (chief economist @Google), where surveillance capitalists use the data captors to extract dark data, and analyze it to draw new conclusions, which enables them to monetize new sectors of our lives. And tech companies are not the only ones: they are joined by all types of “traditional” businesses, car insurers being at the forefront. Using telematics to monitor the behavior of drivers, those companies can derive predictions on who will crash and need to trigger their insurance cover, and therefore can adapt their premiums accordingly: they move from profiling using historical data and statistics to an individual monitoring system that uses dark data for better prediction quality. Thus, they eliminate large swaths of risks (offering cheap contracts to bad drivers), and increase profitability.
Even beyond that, those telematic capabilities enable economies of action, by enabling insurers to nudge users. When a driver has been driving safely, they can get rewarded with a discount for next year’s premium. And of course, some nudges can be punitive, to counter supposed “bad behavior.” Other even more invasive areas are prime candidates for such systems, like the health insurance world for example (“you have not exercised in 2 weeks, next month’s premium will be 15% higher!”). While those types of surveillance are still deemed too invasive for privacy by most consumers, consultants’ advice is to gamify, and propose positive nudges that are too good to pass on… And in the relatively short term, we could end up with marketplaces of data generated by insurance companies and car manufacturers, which would be sold to the usual suspects of surveillance capitalism. Unless of course the latter go generate the data themselves (with Google Maps or Android Auto, for example).
Executing the Uncontract
So this recuperation of the dark data is done for the benefit of the surveillance capitalists, with the user basically not even aware of what is happening to them, and in ways that can be incredibly disruptive to their life. Around the mid-20th century, futurists like Herman Kahn painted a bleak future for the year 2000, where a computer that knows and surveils us would become “indifferent to the fate of those who stand in his way rather than brutal.” And such a system was commonly painted as a hell-ish nightmare. Yet, here we are! The new form of social contract, posits Zuboff, is an uncontract, where, through economies of action (nudging for better predictions), social interactions are replaced with machine action dictated by economic imperatives.
Zuboff argues that most supporters of ubiquitous computing are framing it as inevitable, in continuity with hundreds of years of utopianism. Like Marx and other proponents of planet-wide utopias before them, Silicon Valley promotes the inevitable rise of sensors as if it was a matter of faith. “Everyone in the world will be connected,” repeat Schmidt, Zuckerberg, et al. But when were you last asked if you wanted to live in a world where you are constantly connected to “the system?” It seems that “the internet of things is all push, no pull,” most consumers do not want these devices, they are being sold to enable tech companies to continue their growth.
Men Made It
We forget that behind an inevitable system, there are humans. This is because we have come to not question technology, because it will necessarily raise our quality of life and bring humanity to complete its quest for Progress. Zuboff relies on Langdon Winner’s thoughts on technological determinism: “no one ever bothered to ask if there were other possibilities.” However, she opposes the whole exposé she wrote so far in the book: men and women created the monster.
To the Ground Campaign
All the concepts exposed in this chapter coalesce in a main breeding ground, where surveillance capitalists have started heavily investing: smart cities. Google, and its Sidewalk Lab entity, want to control wifi systems, cameras, traffic flow mechanisms, and many more elements of the bustling life of a city, all focused on flowing data. One of the goals is to also enact economies of action, by running algorithms on the data to design the city itself. And just like for individuals, without necessarily asking for anyone’s consent before rezoning a block, or making a street pedestrian… Another main issue is that cities then end up depending heavily on those companies’ algorithms for some of their income (parking, tolls, commercial rental income…). Zuboff ends the chapter musing on whether cities will become the next area of battle for surveillance capitalists, where they will seek to further their agenda while trying to wiggle free of most regulations in place.