Tuesday, December 01, 2020

Automated Law-- Microsoft 's "Insight Computing System,' the Power to Manage Labor, and the Intimacy of Emerging Regulatory Forms


Image From GeekWire


A meeting insight computing system includes a meeting evaluation machine configured to collect quality parameters from meeting quality monitoring devices. The quality parameters each quantify meeting conditions during one or more previously-elapsed meetings and are usable to determine an overall quality score for each of the one or more previously-elapsed meetings. A graphical scheduling interface is configured to facilitate scheduling of an upcoming meeting at a designated meeting time, in a meeting location, and with one or more meeting participants. An insight generation machine is configured to report a meeting insight via the graphical scheduling interface. The meeting insight is based on the meeting time, the meeting location, the one or more meeting participants, and the quality parameters, and includes a recommendation to change one or more of the meeting time, meeting location, and meeting participants to improve a quality score of the upcoming meeting. (Microsoft Patent Application 20200358627; 12 Nov. 2020; abstract)

So reads the fairly innocuous abstract to the patent application filed for Microsoft by its lawyers  for its new “meeting insight computing system.” It builds on its previously developed “Productivity Score” technology, and both are built around data driven metrics for measuring and predicting labor productivity.  But of course, they do more than that--they serve to develop real time systems of data driven assessment and prediction mechanisms to aid employers in controlling not just the way in which they consume labor services they have purchased, but also to assess and affect the way that the individual employee feels through data driven assessment.

Image From GeekWire

Newly surfaced Microsoft patent filings describe a system for deriving and predicting “overall quality scores” for meetings using data such as body language, facial expressions, room temperature, time of day, and number of people in the meeting. The system uses cameras, sensors, and software tools to determine, for example, “how much a participant contributes to a meeting vs performing other tasks (e.g., texting, checking email, browsing the Internet).” (Microsoft patents tech to score meetings using body language, facial expressions, other data).

 It has been described more luridly and perhaps more bluntly: "It's supposed to help bosses make sure work meetings are productive. But a new computer system that can tell when employees are bored has been condemned as an alarming form of 'workplace surveillance'. " (The spy in your computer that knows if you're bored at work). These advances may be different in quality (though that difference is transformative) but they also spring from a more ancient societal reflex--one centered on the control of individuals who sell their labor. That reflex now manifests in this historical era in quite specific ways--blending trends in governance that have become irresistible, and facilitated by tech.

Brief reflections on this further advance of this irresistible trend follow. 

The transfer of formal autonomy from master to servant (I prefer to use the old terminology both to remind my readers of the not so tenuous relation between free and coerced labor, a relationship grounded on a fluid set of understandings of the wall that separates "modern" slavery and "free" labor) that marked the movement to modernity, markets and the end of indenture of the formal sorts (slavery, servitude, serfdom, etc.) posed a problem for those who could once compel but were now required to purchase labor like other commodities required to produce goods or services for the market (or the state).  Purchasing labor inserted market principles into the relations between those who consumed and those who produced labor.  Those market principles included price, accountability and compliance principles.  It required the importation of notions of loyalty (to a certain extent) as well as an obligation against abuse.  To those ends law was sometimes deployed to manage the boundaries of the market for labor by setting limits on pricing (minimum wage and overtime laws), on access to labor (children and before emancipation on the labor of women and other marginalized groups which might be exploited because of their marginal social status), and especially on the key formal element that marked the distinction between servitude and "free" labor--the right to terminate employment and seek employment elsewhere.  That, of course, was at the core of the markets orientation of labor--the principle that markets for labor could not be fully efficient unless labor was free to sell itself essentially as it could (given the conditions of the market). 

 Importantly, law could be used from time to time to manage the outer boundaries of control of labor once it had been purchased.  That posed a significant problem for law and social ordering, of course.  Labor is both an object (service purchased) and the human being delivering the service.  Where one could control labor consumed and deploy it as one liked, emerging principles of human dignity, of the essential role of the individual as an autonomous political actor detached in that role (eventually and more from the middle of the 20th century) militated against the control of the human delivering or producing the labor consumed. 

The boundaries of control have always been murky as political economic models have continued to wrestle with the issues of distinguishing between control of the person (discouraged) and control of the labor sold (over which plenary control was not problematic). Formal law (the badges and incidence of slavery jurisprudence in the United States) and international law principles and norms (the ILO conventions, for example) have been good on exclaiming grand principle but far less successful in actually penetrating the interior relations between a master and her servants 

With respect to that interior relation, the law provides a large space for private ordering. That is, the law permits the administrative and judicial apparatus not just to avert its gaze from the details of the management of that relationship, but also to insert positive principles that inhibit substantial interference with these private relations, or rather create larger spaces for the exercise of discretion by masters in the imposition of rules that affect the character of purchased of labor.  

This is an intimate space, spaces within which the state has delegated substantial authority in the ordering of relationships and their management  As an intimate space it replicates the division of authority between the state as master and others who have been delegated that authority.  These include those relationships between spouses (husband and wife in its ancient form) and with their children, within religious communities (in the United States at least), and within partnerships (economic or civil society partnerships, including those between shareholders and board members in corporations).  Each of these is grounded on the discursive language (and the legal apparatus) of family, of intimacy, and of substantial discretion behind the "privacy" of the barriers that separate these families from others. 

Not that the state cannot intervene: indeed, much of what passes for jurisprudence over the last several centuries in the West has centered effectively on the nature of the relationship between the state and these private spaces. Notions of consent, of privacy (as against the state and sometimes within the family as against others), of discretion and its abuse, of fidelity to the ideals of family organization and operation, and of compliance and accountability have become the language that describes the relationship between webs of law and webs of relationships, between the framework created by law within which these "families" may order their own affairs, and the structures through which those affairs are ordered within these "families." Within families the more intimate setting provided a space for the development of more inmate expressions of control.  Systems of rewards and punishments grounded in deviations from an ideal (either a societal or one specific to a particular family) were more efficient manifestations of coercion than were the formal structures of jurisprudence which remained the domain of the state and increasingly the field of law. 

Likewise, a more intimate space permitted the resort to more intimate mechanisms.  These were manifested in surveillance, in reporting, in accountability, and in the weaving of expectations about the exercise of discretion and its limits. Data driven governance might be understood, then, as a tech based enhancement of an ancient set of core principles about the organization and management of intimate organizational spaces. At the same time these enhanced techniques were more acceptable within these spaces (than in the public sphere) because of the nature of the intimacy of the family (a greater tolerance and expectation for mutual interference in decision making and stringer willingness to account and hold accountable). Within families control based on observation assessed against idealized expectations managed through the institutions of hierarchies of authority evidenced by increasingly authoritative acts of discretion, served as the model around which all intimate space relations were modeled.

Globalization has extended that concept of intimate space beyond these traditional families to others.  The principal extension touches on the global production chain in the field of economic activities.  But it also includes the emergence of the great apparatus of civil society.  Both embody intimate spaces within their institutions.  And both vividly evidence the changing role (or interlinking) between the inside of global production or civil society spaces, and the outside of norms ad regulatory thrusts from public and international organs. Notions of intimacy--and the barriers they create (sometimes by operation of law--for example the laws of the organization of profit and non-profit organizations)--remain a core principle in the organization of these spaces, in the connection between these and public spaces, and in their regulation. 

Technology, though, has dissolved the old borders between these families and the state, and between the realms of law and those of the ordering and exercise of intimate discretion.  Changing expectations--in politics and norms--have created an environment in which the implementation of great ordering principles of human rights and of sustainability now mixes together the ordering mechanisms of state and private space, of public and intimate space. Tech has also changed the way in which both public and intimate spaces may be governed.  Law has been changing both its character and its function in the face of technological innovation and as a response to tremendous societal shifting from cultures of jurisprudence grounded in command, to cultures of management that are manifested in compliance, accountability, and fidelity to an ideal. Most importantly, the traditional informal and data driven discretionary structures of the internal management of families has now spilled out from that space and has become an increasingly important form of regulation of both public and private spaces. 

The characteristics of that approach to regulation is intimate. It  is based on substantial awareness of what "family members" do, and the authority to control those activities to a substantial extent, at least to the extent that society has defined the jurisdictional limits of "family life." Tech has made it possible to extend this intimacy far beyond the classical family, the small partnership, the family or closely held corporation, or the small civil society organization. Technology has made it possible now to treat any organization, or any system, whatever the scale of its operations or membership, as an intimate space and subject to the forms of control traditionally reserved for small intimate spaces in economic, social, and cultural life.  To speak, then, of automated law, of algorithmic regulation, is to do little more than to recognize the fundamental and transformative change in the way in which regulatory space is conceived.  Technology, effectively, has made it possible for apply the rules for the regulation of that most intimate space of the "family" as the basis for regulating any sort of community space--from the smallest enterprise to the largest national state. It acknowledges that the intimate has consumed the formal public spaces of law and has inverted  the authority of the expression of governance. 

To speak of automated law, or of algorithmic governance, then, is to give a name to tech enhanced law of intimacy. At the same it, is served to acknowledge that the locus of power in regulation has shifted form the authority to commend to the power to articulate an ideal state (of behavior, of objectives, etc.) against which conduct can be monitored, assessed and rewarded when it moves toward the ideal and corrected when it deviates form the ideal.  To make this possible, it is necessary to be able to deploy the fundamental authority of intimate spaces to observe all, to judge all, and to intervene toward a communally palable ideal.

Understood as the law of intimate spaces, now expanded to include virtually any regulable space, makes it easier to understand both the substantial investment of enterprises like Microsoft in the technologies of intimacy and in their obsession not merely with the regulation of (now) more intimate spaces, but also in their desire to predict behavior and (more importantly) to nudge it toward ideal states.  Microsoft has done little to hide this, though the language it uses suggests a deep connection with the discursive tropes of the ancien regimes of law and governance. And in the case of Microsoft's "productivity" tech, it also magnifies the importance of the space between labor and management as an intimate one as well as delineating the lines of authority (and responsibility) between master and servant. Microsoft described its productivity scoring tech as part of its comprehensive productivity tools in this way:

For some time now, business leaders have made digital transformation a priority. But when the pandemic hit this spring, adopting and embracing digital technology went from being a matter of importance to one of sheer survival. COVID-19 has catapulted us into the era of digital everything. And to keep up with the pace of change, and build more resilient businesses to meet future challenges, we need to move digital transformation from art to science.

As leaders, it’s our job to make sure people have the tools they need to do their best work. But tools alone are not enough—you also need to help everyone in your organization build the habits that harness the true power of those tools. Until now, it’s been difficult for leaders to get insight into these habits and understand how to help people make the most of the technology they invest in. Today, with the announcement of the general availability of Microsoft Productivity Score, we take an important step forward. Let’s take a look at how Productivity Score can help your people succeed today and in the future. (Microsoft Productivity Score; Oct 2020).

Notice the discursive tropes that are the foundation for the authority of the techniques of intimacy that are the essence of automated and algorithmic law: (1) the ubiquity of tech and its promise; (2) the challenge of a triggering event that points toward new approaches; (3) the connection between both of these and the survival of the "family" (or enterprise, or organization, or nation, or society, or other intimate collective irrespective of scale); (4) the authority of science as a legitimacy enhancer; (5) the description of the ideal (in this case an office suite that can be perfected and that is populated by perfectible individuals from whom perfected labor may be consumed); (6) the allocation of responsibility along patriarchal lines (the pater familias with responsibility and the keeper in trust of the family's wealth) and his children (including spouse) who serve the family and how obey the paterfamilias (in exchange for something of value) ; (7) the core of the responsibility of the master to perfect his servants; (8) and back to the role of tech as a means of enhancing the gaze of the master within his servants (into their "souls" or at last the psychology and morals of their habits which can then be corrected and perfected "scientifically" --that is which once rationalized may be methodically nudged through an intimate knowledge of servant.

The same applies to the even more intimate tools of the "Insight Computing System." But to that intimacy it adds a connection between the interior spaces of the people to be manged with the management of their environment.

Because conventional computerized scheduling systems lack real-world context, users may not be aware that they are attempting to schedule non-optimal meetings, which may result in meetings that are unproductive at best. At worst, such meetings can negatively impact the health of meeting participants who attend, for instance when the temperature, air composition, brightness, noise level, etc., of the meeting space are non-ideal. Conventional meeting solutions similarly do not offer information regarding the efficacy of meetings they are used to schedule and maintain. For instance, many organizations are plagued by overly long, poorly attended, and recurring meetings that could be modified and/or avoided if more information regarding meeting quality was available. (Microsoft Patent Application 20200358627; 12 Nov. 2020; ¶ [0009])
Perfection, then, is possible only when there is an optimal alignment between te internal "environment" of the individual, the external environment in which the individual works, and the objectives of the master which both serve. 

Like social credit systems in China or ratings systems elsewhere (see our discussion here) the intimate regulation of automated law is at its heart a system of ratings from which judgments can be made and corrections (and rewards) distributed. 

For example, based on the plurality of quality parameters, a "quality score" can be calculated for a meeting, with higher quality scores corresponding to better overall meetings. The specific factors that contribute to a quality score may vary from implementation to implementation, and may include how efficient the meeting was, an emotional sentiment expressed by meeting participants, how comfortable the meeting environment was, etc. Information regarding meeting quality may be distributed to members of an organization as meeting trends, for example listing the highest quality meetings, most productive times of day, most popular meeting spaces, etc.(Ibid., ¶ [0010])
To those ends a variety of tech enhances data harvesting and analytics is necessary. These reduce and essentialize behaviors (interior) and conditions (exterior) against an ideal (that itself may be the product of data analytics) which can then be used to predict success or failure, to judge every component of the analytics (for assessment and correction) and to change behavior.

The comprehensive nature of the observation--and the complexity of the data streams necessary to contribute to analytics the objectives of which are to produce "perfect" meetings is worh describing. Microsoft understands that the nature of managing intimacy requires an all around approach to observatin, to judgment, and to the management of the interactions of people with themselves, with each other, and with their environment.
[0015] A wide variety of devices and services may be described as meeting quality monitoring devices. For example, in meeting 100, quality monitoring device 106A is a thermostat recording the current temperature in the meeting room. Air temperature has a bearing on human comfort level and can negatively affect how comfortable and productive the meeting is when too high or too low. Quality monitoring device 106A may additionally or alternatively measure air composition--relative levels of oxygen, carbon dioxide, carbon monoxide, humidity, ozone, etc.,--which can also have a negative effect on meeting participant health and comfort.

[0016] Quality monitoring device 106B is a camera (e.g., visible light camera, infrared camera) that may, for instance, record which of the invited meeting participants attended the meeting, the body language and/or facial expressions of meeting participants, the amount of time each participant spends contributing during the meeting, etc.

[0017] Quality monitoring device 106C is a microphone that may, for instance, detect speech patterns consistent with boredom, fatigue, etc., record how much time each participant spends speaking, record the ambient noise in the meeting location, etc.

[0018] Quality monitoring device 106D is a personal electronic device that may serve as a source of information regarding a specific meeting participant's behaviors. For example, device 106D may track information regarding how much a participant contributes to a meeting vs performing other tasks (e.g., texting, checking email, browsing the Internet), include information regarding a participant's schedule (e.g., how many other meetings they have attended that day), etc. Quality monitoring device 106D may, for example, take the form of a smartphone, desktop, tablet, laptop, and/or any other suitable electronic device. In some cases, data from quality monitoring device 106D may be supplemented with data stored in a remote location, for example a cloud server or database. Furthermore, quality monitoring device 106D may serve as a proxy indicator of a specific user's presence. For example, the current location of quality monitoring device 106D may be tracked via WiFi connections, Bluetooth (or other wireless signal) beacons present in the meeting space, etc. Presence of quality monitoring device 106D in a specific area (e.g., meeting room) may be used to infer that the owner of the device is also present in the specific area. (Ibid.)

Notice the universe of data harvesting devices necessary to manage the intimate space of the meeting and to align the perfection of all of its components to produce value maximizing (value to the master that is) conditions under constant assessment. The universe of tech data harvesting devices include a thermostat, a microphone, a camera, and something described as a personal electronic device (for example a laptop or iPad).  Notice that these devices are then interlinked with other tech that also harvests data.  The personal electronic device, for example, may include Microsoft's productivity score technology, which would then be integrated with its universe of Insight Computing System technologies. 

The ecology of monitoring that generate a complex cluster of inter-related data, then, appears to be targeted to producing the ideal meeting.  But the entirety can be disaggregated and reconfigured to coerce (through systems of reward and punishment--through nudges) any sort of conduct and any environment of utility to the master. "[0047] In some examples, meeting quality scores and/or specific subscores may be gamified. For example, the meeting trends reporting machine may be configured to generate leaderboards indicating the most productive meetings, most valuable meeting participants, etc." (Ibid.). Furthermore, performance incentives may be offered to improve quality scores--for example, participants in the most productive meetings over a period of time may be awarded a prize. One (a master at least) can then achieve the sort of control of the intimate space that is the master-servant relationship (or scaled upward--the state-citizen relationship, the masses-leadership core relationship, etc.) not through the ancient and inefficient mechanics of law, but through the more nuanced system of data driven governance effectuated through the tech enhanced mechanics of equally ancient systems for the management of intimate spaces.  
That--the globalization of the law of intimate spaces through the mechanics and in the language of tech enhanced analytics, the judgment of algorithms, and the consumption of vast amounts of data--points to the future of law-governance in the age of tech.  But it does more than that.  It points to the power to construct, to model, an idealized set of relationships and conditions and to monitor or assess real time deviations from the ideal.  It permits predictive analytics designed to enhance a universally embraced principle of risk avoiding prevention-mitigation-remedy principles as the core objective of law and norms--applied to substantive norms expressed through law. Microsoft's insertions contribute to movement along this trajectory.  It makes it possible not merely to produce simulations (the construction of reality, its ideal, and the ability to analyze deviations) (discussed Data Driven Pandemic and the Ascendancy of Simulated Reality as the New Political Space), it contributes to the ability to use the techniques of intimacy as the basis for regulating conduct--and changing it to suit the needs of those with the power to control both model (ideal) and data-analytics.  And it is a system that, like the family, is intended to grow and adjust, perhaps to become self aware. 
[0068] It will be understood that the configurations and/or approaches described herein are exemplary in nature, and that these specific embodiments or examples are not to be considered in a limiting sense, because numerous variations are possible. The specific routines or methods described herein may represent one or more of any number of processing strategies. As such, various acts illustrated and/or described may be performed in the sequence illustrated and/or described, in other sequences, in parallel, or omitted. Likewise, the order of the above-described processes may be changed. (Ibid.).

No comments: