Sunday, September 17, 2017

Democracy Part 40: From Mass Democracy to The "Wisdom of the Crowd"-- Socializing People for Roles in New Data Driven Governance

(Pix © Larry Catá Backer 2017)

I have started thinking through the issues around social credit generally, and the Chinese Social Credit project. I will be working through the issues and practices that are presented by the emergence of Social Credit theory--both in China (as an indigenous and quite complex set of policies, advances on political theory, and operational challenges), and in the rest of the world. To understand the shaping of law today (and soft law as well) one must understand social credit. To understand social credit, one must understand the evolving structures of the relationships, in law and politics, of the relationships between states, its masses, and the institutions through which it operates.

Index of Social Credit Posts

A social credit system, and its ratings, are grounded in targeted data harvesting, proprietary algorithms, and coordinated incentives and punishments. There may be some glory in worrying about the algorithms that produce the ratings.  Yet algorithms are merely a function of the choices of data to harvest.  And robust harvestable data is likely the most time consuming and expensive part of any data driven system. In this post I consider the way that the data harvesting element of building social credit systems are being socialized within advanced Western liberal democracies.  To that end I consider the effectiveness of modern television programming and especially a new series from CBS, "Wisdom of the Crowd."

(Pix © Larry Catá Backer 2017)

I have been writing about social credit--a Chinese term for a system of governance grounded in the construction of ratings for virtually any activity. The ratings then become an istrument of governance--providing triggers for benefits, incentives, rewards and punishments. Where the ratings are made transparent, it also permits devolution of governance effects form the state to the masses--the ratings of universities or restaurants provide a good example of the use of markets and mass choices grounded in the effects of ratings. Social credit is practiced all over the world.  Outside of China, the drivers of social credit tend to be private enterprises--for everything from rating credit to the rating of the CSR effectiveness of enterprises (e.g., Social Credit in the West: Non-State Rating Systems for CSR Compliance).

But ratings are only part of the story.  A social credit system, and its ratings, are grounded in targeted data harvesting, proprietary algorithms, and coordinated incentives and punishments. Any system grounded in management of behavior requires data.  That data requirement is actually of three sorts.   

The first is to identify the data that is useful for harvesting (and conversely to identify data that may be ignored. Those choices are ideological and political--far more so than the production of the algorithms through which data is proceed into useful governance objects. It requires consensus on the identity of "facts" and its distinction from conclusion and opinion.  That is easier said than done--consider a system that requires data identifying weeds--the concept of weed is both arbitrary and a function of ideologies of domestic landscaping that varies from place to place. Other "facts" become more difficult--"race", "ethnicity" and "religion" for example.

The second, is to collect data in ways that do not compromise their utility (double counting, fictitious or fabricated facts). This is also more difficult than it might appear.  When the United State produced a government social credit system for public schools grounded in ratings based on standardized test taking, it was not long before it became necessary to devote resources to policing against fraud as school districts, whose prestige and income would depend on their performance rating, sought to "manage facts" in their favor. (e.g., Atlanta Scandal Could Be the Tip of a 'Test Cheating Iceberg').

The third, is to collect data.  Data does not flow like a river from its source to its end, to be drawn on by those who approach its banks to the melodies of Ma Vlast.  Data must be extracted from its sources.  Its production may be compelled--usually by the state through disclosure regimes).  It may be bought, usually by private actors (and by buying the price can be anything frpom access to information (reports and the like in exchange for a name and e-mail) to the actual purchase of informality for money. Or like deer in the forest, it may be hunted, or like truffles in that same forest, collected. One might hire truffle pigs or other intermediaries, but data harvesting is an active endeavor in every case.

But what if one could induce the masses to voluntarily provide data?  Restaurant reviews provide a good example.  It is in people's interest to share experiences about dining in particular places (producing data) as long as they might be able to access the rating that result from the aggregation of that data contribution (after processed through an algorithm into a rating (and made available as raw data). The same theory drives the data collection of data consolidators like TripAdvisor. These sites provide benefits to both data provider and data harvester.  The data harvester can sell a service enriched by the ratings that can be posted which are themselves the product of its customer's willingness to provide data about the products that the data harvester sells.

These closed loop systems of private social credit work well enough.  But is it possible to socialize the masses, or even mass democracy as a collective, to embrace this pattern of data disclosure beyond these immediately self serving closed loop systems? Would it be possible for the state to develop systems for the enforcement of laws (criminal and regulatory) that depends on intelligence by inducing the masses to serve as positive contributors of data necessary for enforcement or regulation?  The answer, in Western liberal democracies, may depend on the ability of the great culture management machinery of Western society--its television, movies and other related media--to develop a narrative in which such activity is naturalized within Western culture (cf., Leigh H. Edwards, The Triumph of Reality TV: The Revolution in American Television (Praeger, 2013)). 

One might get the sense of this possibility, of the socialization of norms that naturalize a societal obligation to actively and affirmatively provide data (to the state or to a private enterprise exercising state or market functions) buy considering the premise of anew television show that is set to debut on the CBS television network.  That show, "Wisdom of the Crowd," is based on a premise that is irresistible  but also challenging for contemporary (or traditional) social ordering.
Written by Humphrey, Wisdom of the Crowd is based on the Israeli format of the same name. Inspired by the notion that a million minds are better than one, it centers on a tech innovator who creates a cutting-edge crowd-sourcing hub to solve his own daughter’s murder, as well as revolutionizing crime solving in San Francisco.

Humphrey executive produces with Keshet’s Avi Nir, Alon Shtruzman, Peter Traugott and Rachel Kaplan, Dror Mishani and Shira Hadad. CBS Television Studios, where Humphrey is under an overall deal, produces through his Algorithm Entertainment banner in association with Universal Television, where Keshet Studios has a deal.  (HERE)
There is a certain self consciousness about the way that the television series may contribute to societal narratives and the normalization of behavior.
“Honestly, the idea scared the hell out of me,” said EP/showrunner Ted Humphrey who worked with Keshet in adapting the Israeli series for a network audience. “In the last year we’ve all encountered the ying and yang of the internet in the way it can be used as a vehicle for human collaboration.” Humphrey said crowdsourcing is beginning to rear its head among police forces across the country, and the show, in addition to its human drama will deal with the online means of how it can be flawed and helpful in solving crimes. (‘Wisdom Of The Crowd’ Will Capture The Controversies Of Crowdsourcing – TCA).
 The story is pitched as a self-help narrative (“'The courts can only do so much. The rest is up to us.' CBS has just released the first preview for the upcoming TV series Wisdom of the Crowd." (here)). And yet the more interesting narrative is the legitimacy of creating a platform through which data mat be harvested through voluntary disclosure because it may be the right thing to do. That right thing to do may involve the advancing of justice (this is a crime solving drama to be sure).  But it can be expanded to almost anything (e.g., find and catch lawbreakers of any kind, expose corporate misconduct, expose the administrative misconduct of officials, catch your neighbor's dog defecating in your yard, etc.).

The premise isn't a leap into unknown territory. Western mass culture has been herded toward the acceptance of the value of this responsibility for some time.  That herding comes form two quite distinct directions.  In the first there is entertainment and pleasure in real time participation in reality and talent shows and contests.  These include the call in features of shows like The Voice with a call in or online voting feature for contestants.  Mass society has been trained to share their views--that is to provide data, as an aspect of participatory mass entertainment.In the second, there is participation in the exaction of justice against criminals--a premise already quite close to that of "Wisdom of the Crowd". Shows like America's Most Wanted illustrate this approach. "In this series, John Walsh, the father of a murdered child hosts this show that illustrates crime stories which have lead to the capture of hundreds of fugitives from the law. With as much luridness and accuracy as possible, various crimes are dramatically recreated with an appeal for any viewer with information as to the crime and the perpetrators to call the show and the authorities and help the cause of justice. " (source here).

The difference is that Wisdom of the Crowd introduces algorithm to data collection to make the probability if meeting objectives more likely (it is indeed interesting that Mr. Humphrey's company is named Algorithm Entertainment). the production company was named   It is a story of the value of regulatory governance as a partnership between the state and its polity.  It suggests the importance of state private cooperation in the construction of real time governance systems that draw on law but are grounded in management and administration to secure behavior norms--reward the good, punish the bad and enhance social order to the benefit of all. But in Western fashion, it will also expose some of the difficulties and challenges.  But those difficulties and challenges will likely touch on administrative discretion, and on the algorithm to be effective.  It will be interesting to see the extent to which it celebrates a change of societal norms that encourages ordinary individuals to part with information as a social duty.  And that, more than anything else, will provide a strong grounding for any social credit, rating, or regulatory management system grounded in  targeted data harvesting, proprietary algorithms, and coordinated incentives and punishments.

No comments: