The Journal of Things We Like (Lots)
Select Page
Jonathan Zittrain, Ubiquitous Human Computing, Phil. Trans. R. Soc. A, vol. 366 no. 1881 3813-3821  (28 October 2008).

A banana usually sells for about 30 cents.  On average, the plantation owner gets 5 of those cents; the shipper, 4 cents; the importer/ripener, 7 cents; and the retailer, 13 cents.  That leaves one penny for the worker who picked the banana.  Fruit economics helps drive the politics of “banana republics:” as the unpaid laborers and netizens at Wikipedia note, such countries are “politically unstable,” “dependent upon” commoditized crops, and “ruled by a small, self-elected, [and] wealthy . . . clique.”  Oligarchs at the top set the direction of society; workers merely play the roles assigned them.  Truth doesn’t matter much; as Paul Krugman noted, one political party promised voters to save money on gasoline by “building highways that ran only downhill.”

Commentators have begun to wonder if the United States is becoming a banana republic.  Nicholas Kristof concludes that “You no longer need to travel to distant and dangerous countries to observe . . . rapacious inequality. We now have it right here at home.”  Chronicling endless financial industry shenanigans, critical finance blogger Yves Smith seems to label every third post “banana republic.”

Wasn’t the internet supposed to solve these problems? Wouldn’t a “wealth of networks” guarantee opportunity for all, as prediction markets unearthed the “wisdom of crowds?”  It turns out that the net, while mitigating some forms of inequality in the US, is accelerating others.  Jonathan Zittrain’s essay “Ubiquitous Human Computing” examines a future of “minds for sale,” where an atomized mass of knowledge workers bid for bite-sized “human intelligence tasks.”  Zittrain explores some positive aspects of the new digital dispensation, but the larger lesson is clear: without serious legal interventions, an expansive global workforce will be scrambling for these jobs by “racing to the bottom” of privacy and wage standards.  This review explains Zittrain’s perspective, applauds his effort to shift the agenda of internet law, and argues that trends untouched on in Zittrain’s essay make his argument all the more urgent.

Exploitation and Alienation Online

Zittrain argues that assembly line-style “division of labor” is becoming more common in mental tasks, ranging from very simple repetitive recognition exercises (“where is the car in this picture?”) to design competitions (“win $1,000 by drawing a new trademark!”).  He states that “We are in the initial stages of distributed human computing that can be directed at mental tasks the way that surplus remote server rackspace or Web hosting can be purchased to accommodate sudden spikes in Internet traffic.”

The resulting distributed labor force offers unparalleled flexibility for CEOs.  While they pursue the vaunted “Four Hour Workweek” of Silicon tycoons, they can avoid making any guarantees of wages to employees–or ask for 80 hour weeks suddenly when business picks up.  In a globally connected world, the cheapest hands are at the ready to perform what “Amazon’s Mechanical Turk enumerate[s as] ‘HITs’ – human intelligence tasks – for sale one unit at a time, from as low as $0.01.”  Once micropayment systems are perfected, pennies from cloud-heaven can rain upon the downtrodden.

Zittrain describes some advantages of this turbocharged division of labor for workers, too.  Operators at one company (LiveOps) “can work whenever they like, wherever they like, for as much or as little time as they like.”  Whereas the traditional employment relationship was like a marriage, with both parties committed to some longer-term mutual project, the digitized workforce seeks a series of hookups. There are plenty of opportunities for the flexibilized worker.

For those saddled a mortgage ball-and-chain, ubiquitous human computing offers less of a blessing.  Aside from a blip of hope in 1990s wage figures, America’s working class has experienced declining compensation since the 1970s.  Establishment journalists were among the first of the “knowledge workers” to experience the same fate, as search engines set up a national and global market for news once delivered locally.  Since similar trends could soon engulf computerized work generally, Zittrain is right to argue that“[m]inimum wage, maximum workinghours, unionization (or at least the ability to know and contact one’s co-workers)” may need to be revisited.  Having discussed “Privacy 2.0” in his book TFOTI, Zittrain also realizes that atomized digital workers need the right to establish a reputation by “building portfolios” if they are to compete effectively for gigs.

Zittrain also worries that “disembodied HITs can deprive people of the chance to make judgments about the moral valence of their work.”  We can imagine a worker figuring out CAPTCHAs in the service of an Iranian intelligence agency or Chinese “fifty cent army” which wants to place hundreds or thousands of messages as comments on blogs.  The atomized HIT is a way of diffusing responsibility in a world where it is already far too hard to pierce the corporate veil, contest trade secrecy claims, or penetrate shadowy government actions.  In response, Zittrain proposes that “harvesters of human mindpower can be encouraged – or perhaps required – to disclose their activities to those who benefit them.”  He also proposes that workers have the opportunity to opt out.  To do that effectively, some entity will need to audit exactly how a company like LiveOps ranks and rates its workforce; otherwise, opting out could be a false choice that simply speeds one’s way to a blacklist.

Why Legal Scholarship Needs Social Theory

Legal scholarship has traditionally focused on discrete doctrinal areas. In intellectual property law, scholars seek to rationalize copyright, trademark, patent, and related doctrines; “cyberlaw” extends to contract, property, and tort online; and privacy experts confront the welter of common law and statutory limits on the accumulation and disclosure of data. While such specialization may promise to “work the law pure” in particular doctrinal bailiwicks, it also risks a tunnel vision that would reinforce trends that few would endorse upon reflection.

Scholars may provide a great service by recognizing trends that burst the seams of extant doctrine.  For example, banking law couldn’t respond to the Panic of 1907 by tweaking extant statutes; it had to struggle fitfully toward an institution like the Federal Reserve.  Similarly, reorganization of work through the internet could make swathes of federal and state employment law irrelevant. It also threatens to transform homes and eviscerate “expectations of privacy.”  Zittrain does not panic, or sensationalize any of these trends.   Rather, he calmly lays out what is happening now, and where it might lead.  The paper originated as a presentation at the World Economic Forum 2008 Annual Meeting workshop, “How Science Will Redraw the Business Landscape of the 21st Century,” and Zittrain manages to present a compelling account of how the day-to-day phenomenology of labor, supervision, and monitoring can be technologically transformed.

As Russell Muirhead has argued, employment relations are not “just work,” in the sense of merely applying oneself to get a wage (only work); they ought to be just (in the sense of fair) work.  Critical internet scholars like Trebor Scholz and Laura Liu have made this case in the realm of digital labor.  Scholz and Liu have focused on the “relationship between labor and technology in urban space, in a context where communication, attention, and physical movement generate financial value for a small number of private stakeholders.”   Zittrain recognizes similar dynamics as a problem for cyberlaw, evincing in his work what William Gibson has called the “eversion” of the internet.  As Gibson observes, “cyberspace has everted. Turned itself inside out. Colonized the physical.”  As the cyber and real become indistinguishable, cyberlawyers will need to influence realms like labor, employment, and consumer protection law. “Ubiquitous Human Computing” is a worthy entrant into these nontraditional cyberfields.

The Tipping Point

Can anything be done?  Diagnosing a social problem is a dangerous game: the issue may be either too trivial to care about or too self-reinforcing and pervasive to be addressed.  Zittrain’s dilemmas of exploitation and alienation may fall into the latter category.   They promise to affect not only the economy, but politics, influencing the “rules of the game” that Zittrain hopes will tame them.

Zittrain mentions in passing the problem of HITs in political campaigns.  If a distributed workforce can crack captchas for spammers, they can also plant comments on blogs, or engage in more creative uses of data to influence public opinion.  Daniel Kreiss recently observed that campaigning has become data-driven; “223 million pieces of personal information” were “provided to Obama’s millions of online and on-the-ground canvassers” during the 2008 campaign.  Data-intensive persuasion will permit new levels of personalization in advertisements, and new demand for a rapid, flexible workforce capable of targeting voters.

The irony of American free speech has reached its apogee in a First Amendment right to lie endorsed by Washington’s Supreme Court, and just about any form of spin and prevarication is allowed in our campaigns.  Unlimited corporate spending also makes the collective will formation a crap shoot. Exxon could have deployed 10% of its 2008 profits to outspend every presidential and senatorial candidate running that year.  Imagine when such interests use “big data” to slice and dice the electorate, with the aid of “minds for sale.”  Impressionable and broke young people might be told they are combing websurfing data to find fellow environmentalists, with their labors really directed to compiling a mailing list of people most likely to be swayed by an ad of Mitt Romney strolling through a meadow.  Big data and personalization will allow candidates to send “Save Medicare” ads to worried seniors and “cut entitlements” ads to Tea Partiers.  Ubiquitous human computing will try to identify every imaginable subgroup: octogenarians particularly worried about Medicare Advantage Plans, angry young men with no dependents, angry old men who can’t stand Medicare Advantage—you get the picture.

Thus, I have little hope that Zittrain’s vision will do much to influence American working conditions: commoditized HITs are far more likely to accelerate current political trends than to be affected by our politics.  Nevertheless, his  proposals should have an impact in other, more advanced political systems, whose governments are committed to shaping economic life to human needs, rather than vice versa.  And if ubiquitous human computing pushes the US one more step toward banana republicdom, at least Zittrain warned us.

Download PDF
Cite as: Frank Pasquale, Banana, JOTWELL (January 14, 2011) (reviewing Jonathan Zittrain, Ubiquitous Human Computing, Phil. Trans. R. Soc. A, vol. 366 no. 1881 3813-3821  (28 October 2008)),