Adalet NCAT01 D Industrial Electrical And Explosionproof Enclosure ...
Big data: accumulation and enclosure
Transcript of Big data: accumulation and enclosure
1
Big Data: accumulation and enclosure
Jodi Dean
On May 1, a working group instructed by US President Barak
Obama to undertake a 90 day study of big data and privacy issued
its report. The group was headed by the President’s Counselor,
John Podesta, and included senior administration officials such
as the Secretaries of Commerce and Energy. The report’s cover
letter opens dramatically: “We are living in the midst of a
social, economic, and technological revolution.” The White House
blog linking to the report similarly pronounces:
While big data presents new challenges, it also
presents immense opportunities to improve lives; the
United States is perhaps better suited to lead this
conversation than any other nation on earth. Our
innovative spirit, technological know-how, and deep
commitment to values of privacy, fairness, non-
discrimination, and self-determination will help us
harness the benefits of the big data revolution and
encourage the free flow of information while working
2
with our international partners to protect personal
privacy.i
When I first read the term ‘revolution,’ I rolled my eyes.
This is ideology on stilts, and not least because the US is
significantly behind the EU with respect to data protection
regulation.ii Replace big data in the above quote with Web 2.0,
the internet, and personal computing and you have the basic form
for the techno-optimistic soundbites of the last forty years.
As I carried out this little thought experiment, though, I
realized that I was missing the point. These sentences from the
White House are not just hype. They are true. We have been
experiencing ongoing technological revolutions (be sure on Marx
slide):
The bourgeoisie cannot exist without constantly
revolutionizing the instruments of production, and
thereby the relations of production, and with them the
whole relations of society. Conservation of the old
modes of production in unaltered forms, was, on the
contrary, the first condition of existence for all
earlier industrial classes. Constant revolutionizing of
3
production, uninterrupted disturbance of all social
conditions, everlasting uncertainty and agitation,
distinguish the bourgeois epoch from all earlier
ones.iii [change slide, #11]
Capital as a class can do nothing but seek new avenues for
profit, new opportunities for accumulation. It engages in
creative destruction, looking for new markets, develops new
commodities, struggles with labor over wages and hours, and tries
to find new ways to extract value. So the White House
announcement of a revolution needs to be taken at its word,
recognized as a statement of class struggle, and understood as
made in the interest of capital as a class.
This leads, then, to the questions I will focus on today:
1. How does big data benefit capital as a class? What is
its role in class war?
2. How should we understand the emphasis on individual
privacy that comes up in so many discussions of big data?
The second question connects to ideology. Since big data is
always accompanied by crazy, unfathomable numbers, always
associated with algorithms and complexity and technologies
4
interacting in ways beyond our understanding and control, it’s
odd that the policy question of big data would be concentrated
into the singular figure of the individual and her privacy. [at
slide #16] What does the individual have to do with so much
bigness? How can the individual and her wants or choices count at
all in a setting measured in exabytes and zettabytes and where
multiple data sources flow together into pools and lakes that can
be algorithmically filtered to reveal patterns answering
questions she never imagined someone could ask? Why is privacy
the name for the issue big data poses?
The Podesta report on big data and privacy recognizes this
mismatch when it says that “the most common privacy risks today
still involve ‘small data’—the targeted compromise of, for
instance, personal banking information for purposes of financial
fraud.”iv Moreover, a second report on big data from the
President’s Council of Advisors on Science and Technology (PCAST)
issued at the same time as the Podesta report makes it clear that
individuals no longer have a reasonable expectation of privacy
when it comes to personal data. The PCAST report asserts that,
given the large benefits of “near-ubiquitous data collection” as
5
well as how data collection fuels “an increasingly important set
of economic activities,” thinking we can limit data collection is
the wrong approach. Technology’s and economy’s own imperatives
determine our options. For example, we can’t even know what data
is stored on an individual much less verify that any data has
been destroyed. So this report recommends that regulatory focus
be placed not on data collection or analysis but on use: does the
use of data harm an individual or class of individual?v
Given these open acknowledgements of privacy’s supercession,
the White House’s emphasis on personal privacy seems like
displacement. Obama called for the study of big data and privacy
in a speech on January 17, 2014. That speech was primarily
concerned with the National Security Agency and the controversy
that erupted in the wake of the outpouring of information leaked
by Edward Snowden regarding NSA surveillance (slide #25). The
leaks included evidence of the PRISM program (whereby the NSA
directly accesses data held by Google, Apple, Facebook, Skype,
Microsoft and others), evidence that the NSA collected internet
metadata in bulk from 2001 to 2011, evidence of the NSA’s storing
of metadata on millions of web users, evidence that it regularly
6
monitors the mobile phones and internet traffic of foreign
leaders, diplomats, and corporations, evidence that is collects
data from Angry Birds and has infiltrated massive multiplayer
online games, and more. In his January 17th speech, Obama
addressed concerns with the bulk collection of phone metadata,
trying both to reassure “ordinary Americans” that the government
was not examining their phone records and to justify what the NSA
was doing in the interest of keeping Americans safe and “not
repeating the tragedy of 9/11.” Incidentally, using 9/11 to
justify massive surveillance was included on an NSA list of
talking points, obtained by Al Jazeera through a Freedom of
Information Act request.
In the January 17th speech Obama issued a new presidential
directive on overseas surveillance that said that the US does not
collect information to suppress dissent nor does collect
information to try to provide a competitive advantage for US
companies. NSA documents leaked by Snowden had suggested
otherwise. Some of the documents treat political and legal
opponents to the US drone program as national security threats.
Further, the NSA spied on negotiators in advance of the 2009
7
climate summit, which ended in disarray after the US tried to ram
through an agreement most other countries rejected. vi The NSA
targeted financial transactions in the Visa and SWIFT networks.
It regularly sweeps up data from foreign corporations such as the
giant Brazilian oil conglomerate, Petrobras. Drawing from
documents leaked by Snowden, a Brazilian television report
“emphasized that the company controlled vast quantities of data
on Brazil’s offshore oil fields.”vii The NSA treats securing
access to fossil fuels and securing an economic advantage over
Japan and Brazil as primary strategic objectives.viii [slide #37]
Given NSA activities, it isn’t surprising that the Podesta group
putting together the report on big data and privacy included the
Secretaries of Commerce and Energy. Given the description of the report
and its goal of protecting personal privacy, however, it is surprising that
there was no one in the working group from, say, the Civil Rights
section of the Justice Department, or the American Civil
Liberties Union, or the Center for Constitutional Rights. This
disconnect is thus one indication of displacement.
The Snowden leaks are the background against which Obama
announced that he had instructed Podesta, along with PCAST (the
8
President’s Council of Advisors on Science and Technology), to
undertake two separate reviews of big data and privacy. Rather
than really focusing on privacy, he is trying to build trust. He
wants to reassure “ordinary Americans” that they don’t need to
worry about the NSA. And he is trying to do this by shifting
focus onto something Americans understand, namely, their own
personal activity and self-interest. Personal privacy takes
pressure off the NSA, displacing it onto the private sector where
data can be linked to consumer benefit and economic growth. The
report put together by Podesta’s group, “Big Data: Seizing
Opportunities, Preserving Values,” is explicit here: “This report
largely leaves issues raised by the use of big data in signals
intelligence to be addressed through the policy guidance that the
President announced in January.”ix It looks instead at big data
as an economic resource, albeit one that brings with it potential
risks. With the shift in focus, big data becomes something quasi-
natural, generated through everyday use of everyday items in
everyday tasks. It is also inevitable – an outgrowth of
technological development. Understood as natural, growing, and
inevitable, big data is, potentially, a resource for immense
9
social good. Because it can reveal patterns and “deliver incisive
results,” big data can drive new economic opportunities. The
report insists: “No matter how serious and consequential the
questions posed by big data, this Administration remains
committed to supporting the digital economy and the free flow of
data that drives innovation” (pg. 9). The report takes as its aim
encouraging “the use of data to advance social good, particularly
where markets and existing institutions do not otherwise support
such progress,” while at the same time protecting core values
(which will have to adapt).
The language of privacy in the Obama administration’s
discussion of big data displaces concern from surveillance to
social good and from government in its national security role to
government in its economic role. Even as energy and commerce
underlie both, concern is condensed in the figure of the
individual whose privacy is to be protected. How this individual
is imagined should tell us something about the ideological field
of which big data is a part. As I will show, the individual is
rarely a worker, figured in face via the constitutive absence of
workers. The preoccupation with privacy thus not only echoes the
10
refrain of US liberal individualism but does so symptomatically,
indexing the proletarianized individual trapped in communicative
capitalism and unable to do anything but perpetually give herself
up to the algorithms of big data analytics. She produces for
capital whether or not she receives a wage. More fundamentally,
that concern with big data is condensed into the individual form
gives us insight into communicative capitalism’s enclosure and
expropriation of the common via capture and analytics. Access to
the common, to the flows and relations of sociality, is what big
data promises. And it can only deliver on that promise by
pulverizing it into and tracing it through the ever more granular
movements of the individual. To paraphrase Karl Polanyi, the
individual fiction, “therefore, supplies a vital organizing
principle” (76), that is, the fiction of what is valued and
protected even as it is generated as mechanism for embedding into
the social substance capitalism’s means of expropriation.
1. Communicative capitalism [slide #48/ change to 49, monopoly
banker]
Communicative capitalism designates that form of late
capitalism in which values heralded as central to democracy take
11
material form in networked communications technologies. Ideals of
access, inclusion, discussion and participation come to be
realized in and through expansions, intensifications and
interconnections of global telecommunications. In communicative
capitalism, capitalist productivity derives from its
expropriation and exploitation of communicative processes. This
does not mean that information technologies have replaced
manufacturing; in fact, they drive a wide variety of mining,
chemical, and biotechnological industries. Nor does it mean that
networked computing has enhanced productivity outside the
production of networked computing itself. Rather, it means that
capitalism has subsumed communication such that communication
does not provide a critical outside. Communication serves
capital, whether in affective forms of care for producers and
consumers, the mobilization of sharing and expression as
instruments for “human relations” in the workplace, or
contributions to ubiquitous media circuits.
Marx’s analysis of value in Capital helps explain how
communication can be a vehicle for capitalist subsumption. In his
well-known discussion of the commodity, Marx considers how it is
12
that different sorts of goods can be exchanged with one another.
His answer is human labor; understood as quanta of time, labor
renders different goods commensurable with one another. But how
is this possible? Why would an hour of mining labor be
commensurate with an hour of farming labor? The answer to this
question involves the fundamentally social character of labor.
What is common to different kinds of human labor is that they are
all labor in the abstract, components of the larger homogeneous
mass of human labor. Labor, and hence value, is, inextricable
from the relations of production and reproduction constitutive of
society. Products of labor are “crystals of this social
substance, common to them all,” that is to say, values.
Communicative capitalism seizes, privatizes, and attempts to
monetize the social substance without waiting for its
crystallization in products of labor. It doesn’t depend on the
commodity-thing. It directly exploits the social relation at the
heart of value. Social relations don’t have to take the fantastic
form of the commodity to generate value for capitalism. Via
networked, personalized communication and information
13
technologies, capitalism has found a more straightforward way to
appropriate value.
One of the clearest expressions of communicative
capitalism’s direct exploitation of the social substance is
Metcalfe’s Law: “The value of a communications network is
proportional to the square of the number of its users.” The basic
idea is that the more people using a network, the more valuable
it is. The truth in Metcalfe’s Law is its association of value
with the communicative network itself. Value is a property of the
relations, the links, between and within pages. Google’s PageRank
algorithm, for example, is one of most successful information
retrieval algorithms because it takes linking into account,
mining and extracting common knowledge PageRank puts to use the
fact that networked communications are the form of capitalism’s
subsumption of the social substance to its terms and dynamics.
Matteo Pasquinelli thus argues that “Google is a parasitic
apparatus of capture of the value produced by common
intelligence.” He treats the prestige that PageRank attends to
(and reflexively enhances) in terms of the “network value” of any
given link. “Network value” describes a link’s social relations:
14
How many other links is it related to? Are those links related to
other links? How many? Google captures this value, the link’s
social substance, its place within a general system of social
relations. [slide #56, yoda, change to next].
And this leads us back to big data. Communicative capitalism
subsumes everything we do. It turns not just our mediated
interactions but all our interactions into raw material for
capital.x Financial transactions, interactions that are caught on
video or are photographed, gps location data, RFID tags, as well
as , soon, the data generated by the multiple small ubiquitous
sensors in what is called the internet of things encloses every
aspect of our life into the data form. A few years ago we might
have understood this as a communicative commons. Now, with the
absorption of a wide array of forms of unstructured data into
massive data pools, it’s clear that we are dealing with something
even more all encompassing. Big data is the capitalists’ name for
this material that Marx understood as the social substance.
2. Data dispossession
15
Two metaphors stick out in big data rhetoric: data as oil
and data as gold, that is to say, fuel and money, energy and
commerce, something that powers and something that circulates.
Both metaphors link data with the opportunity for immense profits
and the rush for these profits that ensues.xi Both treat data as
a natural resource. And both highlight the fact that this natural
resource has to be worked on, whether refined and processed out
of a deluge substance or identified and mined from within a
dense, useless, mass. The oil metaphor first appeared in 2006.xii
Gold is later, around 2012, in the context of reports
acknowledging the challenges executives were facing in drawing
actionable information from their massive stores of unstructured
data.xiii
The oil and gold metaphors are telling in that they identify
big data as the natural resource on which communicative
capitalism relies. This resource, produced by all in common, is
being seized, enclosed, and privatized in a new round of
primitive accumulation. Everyone probably recalls Capital’s famous
Part VIII where Marx discusses the “historic process of divorcing
the producer from the means of production.” This process
16
involved the forcible enclosure of the commons. Landlords,
assisted by the law, expropriated for themselves what had
belonged to the people in common. Property based on the labor of
the owner is thereby replaced with property based on capitalist
ownership. In Marx’s words, “the pigmy property of the many” is
concentrated into the huge property of the few.”
David Harvey rightly points out that far from existing
outside of capitalist processes as some sort of origin, the
practices associated with primitive accumulation coexist with
capitalism. He thus emphasizes accumulation by dispossession,
associating various schemes of privatization, financialization,
and commodification with a new enclosure of the commons.xiv
Dispossession, rather than happening all at once, is an ongoing
process.
No one will deny the ongoingness of data dispossession.
Sometimes it is blatant: the announcement that our call will be
monitored for quality assurance, the injunctions to approve again
Apple’s privacy changes, the necessity of renewing passwords and
credit card information. Sometimes the ongoingness is more
subtle, naturalized in maps, GPS signals, video surveillance, and
17
the RFID tags on and in items we purchase. And sometimes the
ongoingness is completely beyond our grasp, as when datasets are
combined and mined so as to give states and corporations
actionable data for producing products, patterns, and policies
based on knowing things about our interrelations one to another
that we do not know ourselves. Here the currents of lives as they
are lived are frozen into infinitely separable, countable, and
combinatory data-points.
There is a, however, strangeness to data dispossession that
differentiates it from the dispossession that accompanies debt,
privatization, foreclosure. It’s not as if we no longer have our
location when our location data is sold to advertisers in a real-
time auction. We still have our names and email addresses when we
provide them in exchange for access to a website. [slide #72]
It’s not even that we somehow lose control over our names,
addresses, and other identifying information – such control has
always been a myth that treats markers that pinpoint us for state
and capital as crucial to a similarly fantastic vision of deep,
unique, and authentic individual self. We are always already
deeply imbricated with others such that their thoughts and
18
feelings, desires and drives are inseparable from our own. Where
we search for uniqueness, we always find the Other. More
prosaically, rumors and gossip circulated long before the
internet. The dispossession of big data, then, is not about
control of our individual identities.
Rather we are dispossessed of a certain kind of temporality
and a certain kind of being together. These are expropriated from
us and put to alien use. Two kinds of temporality are
expropriated: the momentary and the futural. We lose the momentary
because everything leaves a storable trace. Moving through space
with a mobile phone creates data. Touching a screen, looking at a
screen, creates data. Rather than a time of instants, we have a
time of permanents. In this time, mistakes, errors, and lies
coexist with corrections. Learning, falsifiability, become
attributes of systems, of algorithms, rather than dimensions of
meaning. We lose a dimension of futurity in that a primary mode of
data analysis is predictive: the search for patterns is in order
to predict, and intervene in, the future, even if that future is
only microseconds ahead, as happens in high frequency trading, or
days and weeks ahead, as in the case with the NSA climate summit
19
spying. In each case, data analytics try to eliminate surprise,
the very possibility that something could happen inseparable from
futurity. I should add that the shortening of the future into
ever smaller bits of time may also be part of data dispossession.
I say this because of the failure of big data to result in
significant long-term predictions. Although Google has been
credited with being able to predict flu trends, it over-estimated
US cases in 100 out of 108 weeks. Relying on CDC data on actual
cases yielded better results.xv An MIT scientist claimed to be
able to use Twitter to predict social unrest, but his evidence
was retroactive. And, the National Consumer Law Center looked at
Silicon Valley backed start-ups using big data to generate new
loan products for people with poor credit history. According to
an article in the Wall Street Journal it concluded that “big data
doesn’t make a big difference.”xvi
Data dispossession likewise changes our modes of being
together in that now our being together is for the private
benefit of another. In the words of a 2014 World Economic Forum
report on big data, “Our collective discussions, comments, likes,
dislikes, and networks of social connections are now all data,
20
and their scale is massive.”xvii Sociality – and not just person
to person but persons to animals, environment, and things – is
enclosed, analyzed for past patterns and held for future ones in
the interest of squeezing out some competitive advantage.
Communication, culture, and care are seized and tagged. We can do
nothing that is not already for capital. To invoke words from
Karl Polanyi, data dispossession separates “the people from power
over” our own communicative life. And the way that this matters
is not individual but collective, our common power, exercised in
multiple, fluid, indirect, and uncertain ways, over the relations
we create in common.xviii
Big data pushes for data driven analytics, prediction, and
decision-making for everything but its own adoption. That is to
say, even as the big data mantra is that with the proper
analytics big data can provide real, actionable knowledge, the
move to big data is itself exempt, unbacked by either big data or
analytics that prove its benefits. Its promises remain
provisional, speculative, the matter of multiple small
experiments and an enormous push by governments and the
information and telecommunications sector. A skeptical Wall Street
21
Journal report as recent as 2012 notes the failure of an early
Pentagon big data test. Battlefield commanders were provided
“exquisite situational awareness,” that is, information about
everyone’s whereabouts. “What they found was that just giving bad
generals more information didn't make them good generals; they
were still bad generals, just better informed.”xix Nevertheless,
the installation of big data as an economic force isn’t itself
subject to big data, much like the emergence of an economy rooted
in commodity production wasn’t itself the product of markets.
3. Capital accumulation
Understood as the basic resource of the knowledge economy,
big data has the characteristic of being self-renewing. It’s
inexhaustible, co-extensive with the reproduction of social life.
It reaches through and beyond work, even beyond the reproduction
of workers, into the social substance itself. The Podesta report
asserts that big data can “help create entirely new forms of
value.xx”
It’s worth considering this claim closely. Nothing in the
report would lead one to conclude that its authors have in mind
22
something like the collective benefit that accrues from common
modes of being or the strength that emerges out of pulling
together to address the global challenges of climate change and
economic inequality. Rather, they seem instead to have in mind
capitalist value or value that leads to capital accumulation.
That it is taking a “new form” could mean that it is value that
exceeds the wage relation and even the property relation, which
is what I’ve been arguing in terms of communicative capitalism’s
direct expropriation of the social substance in the form of big
data. But this probably isn’t what they have in mind either. They
probably are just thinking of new opportunities for capital
accumulation. A contribution to the World Economic Forum Global
Information Technology Report 2014 on the rewards and risks of
big data invokes value in this sense when it describes the
potential for gains of 14.4 trillion dollars in added value in
the commercial sector over the next ten years:
This opportunity exists in the form of new value
created by technology innovation, market share gains,
and increasing competitive advantage. It translates
into an opportunity to increase global corporate
23
profits by approximately 21 percent, driven by
improvements in asset utilization (reducing costs and
improving capital efficiency), employee productivity
(improved labor efficiency), supply chain logistics
(eliminating waste and improving process efficiency),
customer experience (adding more customers), and
innovation (reducing time to market).xxi
Value here is a matter of global corporate profit, of capital
accumulation by the capitalist class. It accumulates from cutting
the labor force (“reducing costs” and “supply chain logistics” in
their terms), squeezing the remaining workers (“improved labor
efficiency”), trying to get people to spend more money, and
becoming more competitive. This last benefit is necessarily short
term. Even if big data gives a competitive advantage to early
adopters, as it becomes standard, that competitive edge will
diminish; this is the case in the adoption of any technology.
Consider, though, a couple of the big data experiments that
have generated corporate value. Bank of America put tracking
sensors on ninety workers and discovered that the most productive
workers frequently engaged with their colleagues. The bank
24
started mandating group breaks and saw a ten percent productivity
increase.xxii UPS installed sensors as well as GPS in its trucks
in an effort to increase efficiency and control costs. Data on
more than 200 elements is collected, including truck speed,
number of times the truck is put in reverse, driver seat belt
use, the length of time a truck is idling.xxiii It can now decrease
fuel consumption while increasing the number of deliveries per
truck.xxiv A Forbes article on workforce analytics indicates the
importance of big data for “controlling labor costs,” which for
the healthcare, education, and service industries is upward of 50
percent of operating budget. Many such enterprises “track the
time employees arrive, what they do at work, when they leave for
breaks, the times they call in sick, schedule details, personal
information and much more,” writes the author, Bill Barlow.
Workforce analytics lets a company use this information “to
optimize its labor force by scheduling the right mix of full-
time, part-time and temporary labor on a variety of schedules.”xxv
Another way to make the same point: big data increases worker
precarity as it enables companies to do more with less.
25
Approached in terms of class struggle, big data looks like
further escalation of capital’s war against labor. If earlier
ways of automation displaced industrial workers, big data
portends the displacement of post-industrial or knowledge
workers. It should come as no surprise, then, that education and
health care, two of the last remaining sites of intensive, face-
to-face relatively high-paying labor are often singled out in
media, industry, and governmental discussions of big data. An
excellent study, “The Future of Employment,” by Carl Benedikt
Frey and Michael A. Osborne, explains why this is the case.xxvi
Historically, those tasks could be computerized which followed
clear, rule-based routines. Too many variables meant too hard to
program. More recently, though, there have been technological
breakthroughs wherein non-routine tasks could be turned into
well-defined problems. And this becomes possible because of data.
As Frey and Osborne write:
Data is required to specify the many contingencies a
technology must manage in order to form an adequate
substitute for human labor. With data, objective and
quantifiable measures of success of an algorithm can be
26
produced, which aid the continual improvement of its
performance relative to humans … As a result
computerization is no longer confined to routine tasks
that can be written as rule-based software queries, but
is spreading to every non-routine task where big data
becomes available.”xxvii
Examples include the development of voice recognition capacities
that enable call centers to replace people with algorithms,
Google’s driverless cars, as well as developments in robotics.
They include as well the kind of knowledge work previously seen
as invulnerable. Massive amounts of data enable the automation of
an array of decision-making tasks: medical diagnoses and
treatment, fraud detection, legal services, ad design, purchase,
and placement, stock-trading. The struggle over education,
particularly with MOOCs and efforts to “personalize” student
learning by conducting ever more of it on screens, is still in
its early stages. Frey and Osborne find that “47 percent of total
US employment” is at high risk of being automated within the next
two decades.xxviii I’ll add that they put most jobs in healthcare
and education at relatively low risk for automation insofar as
27
they require assisting and caring for others, persuasion, social
perceptiveness, and manual dexterity. Perhaps not surprisingly,
CEOs do not also to appear to be at risk.
The value in and of big data is for capital, not for the
people from whom it is expropriated. One of the contributions to
the World Economic Forum 2014 report is explicit on this point.
The authors, Peter Haynes and M-H Carolyn Nguyen, note that “the
greater the role that data play in the global economy, the less
the majority of individuals will be worth.” In fact, “this could
mean that a data-driven economy may become a contracting
economy.”xxix Although Haynes and Nguyen propose various schema
for remunerating people for their data, when they say that “the
concept of fair value exchange no longer exists,” they imply that
the train has already left the station (ship has sailed, horse is
out of the barn?). Many of us already give away massive amounts
of data, and “corporations are making significant profits as a
result because their cost of materials is essentially zero.”xxx
They quote Jaron Lanier, “The dominant principle of the new
economy, the information economy, has lately been to conceal the
value of information…. We’ve decided not to pay most people for
28
performing the new roles that are valuable in relation to the
latest technologies. Ordinary people ‘share,’ while elite network
presences generate unprecedented fortunes.”xxxi
4. The individual
I should try, if I can, to get back to the Obama
administration’s reports on big data and privacy. They don’t have
anything to say about the impact of big data on workers. The
Podesta report presents individuals primarily as consumers
(secondarily as citizens) whose access to free goods on the
internet is subsidized by the marketing and advertising
industries (pg. 50). These consumers are primarily at risk of
discriminatory uses of their data, and the report recommends that
expertise be developed to be able to determine if this occurs.
The PCAST report, however, is different. The second chapter
of the report sketches “examples and scenarios” of the benefits
of big data. It begins with an unstructured list of some big data
experiments. These include differential pricing (the practices
that construct prices according to consumer rather than item or
service), license plate readers that alert police if a vehicle
29
associated with an outstanding warrant passes, and inferences
regarding a person’s emotional state based on subtle changes in
facial coloration.
The report then groups its examples and scenarios around
health, education, and the “special status of the home,”
emphasizing the benefits of big data while also acknowledging
potential risks to privacy. The healthcare section highlights
“personalization.” It associates “personalization” with
researchers’ access to millions of health records as well as the
hypothetical use of a mobile device connected to “a personal
assistant in the cloud” capable of detecting Alzheimer symptoms.
Through the device, data about the patient’s medications,
symptoms, and habits is produced and uploaded into the cloud.
Personalization should thus not be confused with personal. There
is nothing personal here. Instead of interacting with a person,
who would need to be paid, a patient uses a mobile device and a
cloud assistant like Siri or OK Google. The report mentions Siri
and OK Google directly, perhaps because Google’s Executive
Chairman, Eric Schmidt, was a member of the big data and privacy
working group, along with a senior advisor to the CEO of
30
Microsoft and the founder and managing director of Zetta Venture
Partners, a venture capital fund focused on analytics.
The education section is interested in “innovative pedagogy,
but not teaching. Teachers and professors are completely missing.
What appears instead are MOOCs and the opportunity to compile
longitudinal data on millions of students. The pedagogical
innovation seems to be turning education into an opportunity for
data production.
This will include not just broad aggregate information
like grades, but fine-grained profiles of how
individual students respond to multiple new kinds of
teaching techniques, how much help they need to master
concepts at various levels of abstraction, what their
attention span is in various contexts, and so forth. A
MOOC platform can record how long a student watches a
particular video; how often a segment is repeated, sped
up, or skipped; how well a student does on a quiz; how
many times he or she misses a particular problem; and
how the student balances watching context to reading a
text . . . With this information, it will be possible
31
not only to greatly improve education, but also to
discover what skills, taught to which individuals at
which points in childhood, lead to better adult
performance in certain tasks, or to adult personal and
economic success.xxxii
In the absence of teachers, this looks like a transfer of wealth
from education to technology. It also looks like an
intensification of surveillance. Every moment of a child’s
education becomes part of a data stream feeding an assessment of
adult performance. Schools and districts become data farms,
providing an unending supply of harvestable data.
The section on the home draws out the data to be produced by
the internet of things: air quality and environmental sensors,
motion detectors, commodities signaling their presence,
“omnipresent audio and video collection.” The home appears as a
site of multiple appliances intensifying the dependence and
vulnerability of the in-data-ed person. On the one hand, the
appliances seem to enhance their occupants’ quality of life:
Netflix, Amazon, Microsoft Kinect, and Google Glass all make an
appearance as the occupants are imagined playing games and
32
consuming media. The report asks: “If the camera in your
television knows what brand of beer you are drinking while
watching a football game, and knows whether you opened the bottle
before or after the beer as, who (if anyone) is allowed to sell
this information to the beer company, or to its competitors?”(pg
16). On the other hand, the occupants are positioned as
vulnerable to intrusive controls: environmental monitors that
detect heroin smoke and introduce possibilities of police
notification and insurance cancellation, landlords that make the
presence of multiple sensors requirements for renters. The home
is a data factory. To live in it is to produce data, whether one
wants to or not.
In his 2007 book, iSpy, Mark Andrejevic uses “digital
enclosure” as a term for “the creation of an interactive realm
wherein every action and transaction generates information about
itself.”xxxiii He describes the processes whereby people submit to
monitoring in order to enjoy the benefits of using a mobile
phone, making a purchase on the internet, or digitally recording
a television show. And he draws the powerful connection between
early twentieth century workplace management techniques using
33
monitoring to increase worker efficiency and early twenty-first
century forms of domestic monitoring to increase consumption.
Andrejevic writes, ‘Convincing members of the public to submit to
this type of monitoring, or rendering surveillance so passive and
invisible that they may not recognize it, remains the challenge
of the twenty-first century.”xxxiv The PCAST report adds a third
option: making data production and monitoring a condition of
being so that choice and consent no longer make sense as options.
Conclusion
I’ve argued that the erasure of the worker in the Obama
administration’s recent reports on privacy and big data is
symptomatic of a focus on capital accumulation from the
perspective of capital as well as on a view of data as a common
resource that, under communicative capitalism, we can’t not
produce. This leads me to a final element in the PCAST report:
Taylor Rodriguez. Taylor Rodriguez figures in the future scenario
that closes the discussion of events and scenarios. She is the
one worker who appears in the report and so gives us insight into
how the privacy and big data working group imagines workers. In
34
the scenario, she doesn’t appear in a specific workplace. She’s
going on a business trip. She seems to be quite successful:
She packed a bag the night before and put it outside
the front door of her home for pickup. No worries that
it will be stolen: The camera on the streetlight was
watching it; and, in any case, almost every item in it
has a tiny RFID tag . . . Nor is there any need to give
explicit instructions to the delivery company, because
the cloud knows Taylor’s itinerary and plans . . . At
the airport, Taylor walks directly to the gate – no
need to go through any security . . . There are no
boarding passes and no organized lines. Why bother,
when Taylor’s identity (as for everyone else who enters
the airport) has been tracked and is known absolutely?
When her known information emanations (phone, RFID tags
in clothes, facial recognition, gait, emotional state)
are known to the cloud, vetted and essentially
unforgeable?xxxv
The report acknowledges that this might seem creepy to us. But it
concludes nonetheless that in this world of “the cloud and its
35
robotic servants … major improvements in the convenience and
security of everyday life” have been possible.
Competition in communicative capitalism is intensifying.
Corporations look for advantages, using big data and workplace
analytics to squeeze out new efficiencies so they can get more
with fewer employees. Advertisers try to get closer and closer to
the customers, using data from social media, credit cards, search
patterns and click streams to dream up ways to get attention,
make sales. In the finance sector, big data drives high frequency
trading and other questionable banking practices. New big data
efforts in health and education create even more opportunities
for data production, reaching behind the backs of paid labor as
they render ever more practices into generative opportunities for
the data resource necessary for trillions of dollars of new
capital accumulation. Taylor Rodriguez looks like one of the
winners in this economy. Served by hordes of personal digital
assistants, everything flows smoothly, seamlessly. The real
workers making her life possible –picking up the bag, driving the
car, staffing the airport, flying the plane, not to mention the
i http://www.whitehouse.gov/blog/2014/05/01/findings-big-data-and-privacy-working-group-reviewii http://www3.weforum.org/docs/WEF_GlobalInformationTechnology_Report_2014.pdf pg.76iii Marx, The Communist Manifestoriv P. 21 Podesta reportv http://www.whitehouse.gov/sites/default/files/microsites/ostp/PCAST/pcast_big_data_and_privacy_-_may_2014.pdf xii-xiiivi http://www.theguardian.com/environment/2014/jan/30/snowden-nsa-spying-copenhagen-climate-talksvii http://www.nytimes.com/2013/09/09/world/americas/nsa-spied-on-brazilian-oil-company-report-says.html?_r=0viii NYT Nov 2 2013 No Morsel Too Miniscule for All-Encompassing NSA, Scott Shane http://www.nytimes.com/2013/11/03/world/no-morsel-too-minuscule-for-all-consuming-nsa.htmlix Podesta report pg 9x WEF report 35, data as resource, new input like labor and capitalxi http://tech.fortune.cnn.com/2013/09/04/big-data-employment-boom/xii Clive Humby 2006, said “Data is the New Oil” adopted in 2011 by World Economic ForumData is the new oil, Michael Palmer, http://ana.blogs.com/maestros/2006/11/data_is_the_new.htmlxiii Kronos report, “data+analysis+action=gold” http://www.kronos.com/showAbstract.aspx?id=12884902528&rr=1&sp=y&LangType=1033&ecid=ABEA-56QT5Sxiv Harvey, “The New Imperialism,” Socialist Register 75xv http://www.theguardian.com/technology/2014/mar/27/google-flu-trends-predicting-fluxvi http://online.wsj.com/news/articles/SB10001424052702304732804579425631517880424xvii http://www3.weforum.org/docs/WEF_GlobalInformationTechnology_Report_2014.pdf pg3xviii Polanyi 234xix “Big data’s big problem: little talent,” The Wall Street Journal http://online.wsj.com/news/articles/SB10001424052702304723304577365700368073674xx Podesta 37xxi Pg 38
http://www3.weforum.org/docs/WEF_GlobalInformationTechnology_Report_2014.pdfxxii Rachel Emma Silverman, “Tracking Sensors Invade the Workplace,” http://online.wsj.com/news/articles/SB10001424127887324034804578344303429080678xxiii http://www.automotive-fleet.com/channel/green-fleet/article/story/2010/07/green-fleet-telematics-sensor-equipped-trucks-help-ups-control-costs/page/1.aspxxxiv https://www.google.com/search?q=sensors+in+ups+big+data&rlz=1C1GGGE_enUS484US484&oq=sensors+in+ups+big+data&aqs=chrome..69i57.5770j0j4&sourceid=chrome&es_sm=122&ie=UTF-8xxv ‘Work Smarter: how big data can boost labor performance” http://www.forbes.com/sites/ciocentral/2012/07/05/work-smarter-how-big-data-can-boost-labor-performance/xxvi http://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf
xxvii Frey and Osborne 15-16xxviii Frey and Osborne 38xxix “Rebalancing Socio-Economic Symmetry in a Data-drive economy,” Peter Haynes andM-H. Carolyn Nguyen, WEF report http://www3.weforum.org/docs/WEF_GlobalInformationTechnology_Report_2014.pdf pg. 70xxx WEF 69xxxi 69xxxii PCAST 14xxxiii Mark Andrejevic, iSpy (University of Kansas 2007) 2.xxxiv 91xxxv PCAST 17