A review of Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for the Future at the New Frontier of Power (Profile Books, 2019)
In 2014, a new toy hit the shops, My Friend Cayla. This doll was unlike other dolls: responding to its playmate’s voice; programmable with the names of family and pets; and, through its Bluetooth internet connection, giving spoken responses to questions.
But Cayla had its own agenda: collecting data from the child's speech for targeted advertisements including product placement in its speech. The spy-doll was banned in Germany, but “the smartest friend you’ll ever have” is still available in the US.
The central focus of Shoshana Zuboff’s new book is that Cayla is but a foot-soldier of surveillance capitalism, harvesting data from a subjugated public. Websites now search you, maps follow you and TVs watch you.
This book mainly examines Google/Alphabet and the more narrowly focused social network corporation Facebook, along with a host of smaller tech companies, arguing that they are progressing beyond gathering data on individuals’ activity online to more via smartphones and the internet of things.
More speculatively, Zuboff argues these companies’ monopolisation of knowledge is creating a new undemocratic power over the individual which she calls “Big Other” (fusing Orwell’s Big Brother with the behaviourist psychologist B F Skinner’s view that human subjects should be viewed as detached “others”).
Google’s ever increasing array of user-interfaces (Google Maps, Gmail, Google Chrome, Streeview and the dominant smartphone operating system Android) creates the unprecedented ability to gather data on its users. Google’s self-defined mission is to use this data to pre-emptively satisfy each user’s needs.
Without regard for privacy, data is scraped from voice activated household gadgets (like Cayla) searching for key words (“like”, “want”, “love”) to keep tabs on your preferences. Google’s users are not its customers but rather its raw material, supplying behavioural data which is the basis of its business. When Google was launched in 1998 it had no clear business plan. Only after 2000’s dot-com crash did it develop the capacity to record users’ activity across the internet and beyond as the basis for selling advertising.
When Zuboff theorises more grandly, her critique of surveillance capitalism is both overly elaborate and half-baked. The book offers no understanding of what capitalism is.
She sees state regulation as allowing good market capitalism to thrive and suppressing the excesses of free enterprise, such as the bad capitalism of surveillance capitalism. While Google and Facebook are viewed as the product of neo-liberal, unregulated, markets, good capitalism has “reciprocities” with state regulation, ensuring workers get something back (the term and broad framework comes from the mid-twentieth century social-democratic theorist, Karl Polyani). Her anti-capitalism is rhetorical. She appears to want capitalism with privacy laws, appropriate state regulation and responsible corporations. That capital rules our lives in the workplace is faded out from her picture in favour of speculations about capitalism ruling us by marketing nudges.
Yet her claims are sweeping: “established capitalist ‘laws’ such as competitive production, profit maximisation, productivity and growth … now operate in the context of a new logic of accumulation that also introduces its own distinctive laws of motion”, this being “the foundational framework for a surveillance-economic order” where “the ‘means of production’ now serve ‘the means of behavioural modification’.”
She sees Google’s users being exploited for their behavioural raw material as dispossession – a renewed but vampiric capitalism.
But the tech sector is only ten per cent of even the US economy, and the so-called surveillance capitalists are a minor part of this sector. And Zuboff does not even analyse closely how the “ surveillance capitalists” make their money, namely through advertising.
New forms of advertising are unlikely to transform capitalism: companies that sell advertising are reliant on other industries to make goods which (with the advertisers’ help) are sold at a profit. Adverts themselves cannot create new profit.
Zuboff touches on the possibility that the surveillance economy may intervene more directly in the production process through big-data and machine learning, but does not develop this.
The argument that offering free content to sell advertising constitutes exploitation of the user/viewer is not new. There is some (poor) Marxist theory that argues that people who watch TV adverts are performing unpaid labour for capital (Dallas Smythe) and this has also been applied to the internet (Christian Fuchs).
A better way of understanding Google and Facebook is that they produce use-value (a search engine is certainly useful, and many people find social media so) to create an audience for the advertising they sell. Marx saw activities like advertising as unproductive labour – they do not directly produce surplus value, but only allow the advertising capitalist to scoop up a bit of the surplus value produced elsewhere. Advertising is a cost of circulation, a price the commodity-producing capitalist may have to pay to sell their products and “realise” surplus-value in the form of money-in-the-bank profits.
It would be possible to analyse how the internet has transformed advertising. It can be more focused. It is affordable for smaller producers. It has squeezed press and commercial broadcasting revenues. It may move advertising from an old model of using imagery, suggestion and psychology to create new or artificial desires to a new(ish) model that attempts to nudge the consumer into parting with their money.
The need to sell advertising affects the nature of the use-value offered as free content on the internet (which explains much about social media), but Zuboff touches on this only tangentially.
Instead, Zuboff builds up a speculative theory that the huge amounts of data being gathered on individuals’ behaviour, combined with the ability of machine-learning to identify usable patterns in this mass, constitute an existential threat to human agency.
She calls the resulting undemocratic power of “Big Other” “instrumentarian” power, a term that is never fully defined, but would appear to combine pragmatic goal-directed practice (instrumentalism) with the desire to control the individual of totalitarianism. She depicts a society where human behaviour is conditioned by impersonal machine-learning systems that shape our behaviour for no purpose other than economic gain.
Zuboff’s speculative dystopia obscures the problematic present, but this is where we should start. We used to navigate with paper maps bought as a commodity. We now use free maps on smartphones paid for by advertising (both on the map and other adverts using the information we have surrendered about our journey). While the information on the paper map is relatively neutral, we now use cartography conditioned by economic interests.
While this might seem trivial, it is scaleable. Google are developing urban spaces where people and their “needs” are tracked and predicted by machine-learning. The resulting information supplied to us is created by matching our behavioural data with the interests of the companies paying Google. Such a Google city would, in a sense, privatise both public space and personal life. As Google’s Dan Doctoroff stated, “We expect to make a lot of money from this.”
This does not, however, constitute the kind anti-political authoritarian power that Zuboff suggests. It warps our knowledge of our environment in the interests of commodity-producing capitalists who pay for the services we use via our smartphones. Zuboff speculates about this warping becoming full-on control, although she can furnish few examples of such control.
Several times she uses the example of leased cars refusing to start if the payments are not kept up - hardly the harbinger of a new authoritarian system and not the kind of coup that she is seeking to demonstrate.
She suggests that the incipient technology of insurance companies gathering data on policy holders’ driving habits is a threat to liberty. Zuboff’s objections are of an anti-social libertarian type - what if I am late and have to drive recklessly, I don’t want my insurance premium to go up!
It may be that such surveillance will push onto the individual costs of their own (bad) decisions in an undesirable way, creating a more atomised, self-regarding and fragmented society.
People with unhealthy lifestyles could be forced pay more for their health insurance, although Zuboff has a tendency to blame the surveillance (the cost of health insurance going up) rather the capitalism (a market in health care and health insurance). Her implied solution is a regulated market, spreading costs between individuals with different lifestyles (very Obamacare).
This vague social liberalism contains sharp fragments of right-wing individualism. She poses the problem of the use of data analysis to predict human behaviour not as one of it being skewed and used by commercial interests, but rather as one of the whole business being antithetical to individual freedom. Similarly, she opposes in principle the collectivism of the “hive mind” which Zubbof sees as being created by online forums, and ignores its potential for new forms of collectivism.
There is much good material in this book, but the idea that capitalism is transformed by such surveillance is overblown. If Zuboff’s speculative dystopia has a purpose it is that it will be to encourage more careful and considered research into the cultural, political and economic consequences of the gathering of private data for commercial gain.
Although the digital world is distorted by profit, its technology does extend human potential. Another internet is possible.