Ours is an especially managerial age. The dominant culture is one of technical decision-making and administrative expansion—not only in industry, but in schools, hospitals, public administration, unions, and even in charities. The role of judgment and moral reasoning has been dropped in much of modern organization and replaced with the objectivity of process and metrics. According to famed historian Morgen Witzel:
Where there is judgment there is doubt, and one of the purposes of the new scientific approach to management that emerged after the Second World War was the elimination of doubt.
As a result of this societal project, decision-making has been ceded to technologies themselves, resembling neither the worries of free-market capitalists about an overbearing state, nor of socialists about corporate overlords. Rather, they resemble Terry Gilliam’s Brazil or Franz Kafka’s The Trial, in which the autonomous bureaucracy itself oppresses across sectors of life.
From this vantage point, it is unsurprising that there are anxieties about the so-called “Fourth Industrial Revolution,” which promises to automate whole sectors of the economy with emerging technologies, such as advanced robotics, artificial intelligence, and connected devices. While automation in itself has been a historic good, raising productivity levels and expanding prosperity, the worry is that processes of technological advance won’t be used in the pursuit of societal betterment once they are embedded into technocratic and managerial structures. For humanistic technology to flourish, development and implementation must be guided by values, the very thing that the ascendancy of a value-free, scientific approach to organization deems anathema.
Harnessing the benefits of new technologies, therefore, requires the exercise of good judgment, which has a degree of doubt and a qualitative ethos not present in the modern economy.
Management As Discipline
Before discussing the relevant ways in which modern organization and the dominant managerialism are ill-equipped to make prosperous use of emerging technologies, managerialism must be properly defined. In the context of this piece, management is distinct from leadership, which conjures up images of the charismatic entrepreneur or the visionary political hero.
The manager is that individual, common in modern organizations, whose job is to make sure others are doing their job well. Managers are both feared and maligned; their position serves the role of desire that individuals within organizations strive for, and the role of scorn when someone else fills it. It is the managers who dominate our culture because of their ubiquity, since the daily tasks of planning and organizing teams and departments fall to them. Due to the large presence of this professional class and its influence in decision-making on the margin, rather than grand strategy, it is they who hold significant practical authority in implementing new technologies. As a result, realizing the promises and perils of these technologies is a project which will likely be shaped in large part by the culture of the managerial class.
Management as a modern discipline is fairly new, being necessitated by the Industrial Revolution and first developing in the Anglosphere, but soon spreading across of Europe and beyond. As a result of the increasing organizational complexity that the vast productivity boom of that era provided, there was a demand for individuals with the technical skills of administration. This demand existed in both the nascent corporation and the rapidly expanding civil service. Max Weber, one of the earliest scholars of the rising bureaucratic order dominated by the manager, noted that this structure was “increasingly more precise in calculating the methodical attainment of given practical ends.” Managers were efficient because they went beyond the accounting practices that preceded them. Now it was possible not just to keep account of existing metrics, but also to pursue the quantification of the qualitative through bureaucracy and the managers placed into it.
The role of the manager required increasing precision to justify its increasing authority in organization. Throughout the 20th century, business schools developed their own language of technical rigor. This provided the veneer of an increasingly scientific discipline and invented a new grammar of management. Such trends led to a view that the science of organization was as deterministic and mechanical as that of physics. Following the Second World War, many countries left in place or even expanded the administrative processes initiated to increase the efficiency of the war effort. This was done on the premise that they would retain their efficiency in peacetime. Technically competent managers were increasingly placed to manage schools, hospitals, and the scientific enterprise, rather than the most experienced teacher, doctor, or distinguished scientist. This extended to the structure of the corporation as well, with senior roles increasingly hired out of university management programs as opposed to trained up from the proverbial shop floor.
Scottish philosopher Alasdair MacIntyre has been one of the most scathing critics of this culture. In his 1980 work on moral theory After Virtue, he identified a fundamentally amoral character in management that dissolved genuine social relations. He posits that since managers pursue the most efficient means to a given end, they end up having no connection to the purpose of their work; they are therefore unable to exercise any moral judgment, as doing so is deemed outside their scope. The manager, by the demands of his role, must reduce problems of value or judgment to technical problems with clear empirical solutions. The manager views his role as one of making rational decisions on the basis of evidence, in pursuit of clearly definable and unambiguous goal metrics. In this way, for MacIntyre, management is the perfect embodiment of moral emotivism, which views judgments of value as subjective statements of preference, not subject to rational analysis. In both cases, means may be evaluated as better or worse in relation to a certain end, but the end itself must be regarded as arbitrary. There are no better or worse ends, only different ones.
To MacIntyre, the work environment that a manager creates is one that can easily be adapted to suited ends, one “with effectiveness in transforming raw materials into final products, unskilled labor into skilled labor, investment into profits.” Such an environment demands more routine and process compliant work from laborers. With some exceptions, employees enter into work each morning, do their job, and clock out; teachers teach the test; doctors come to see their patients and charts; and scientists focus on incremental and measurable experimental improvement rather than grand creative leaps. One has only see the degree to which specialization has occurred in recent decades, with workers turned into experts of increasing precision, yet decreasing versatility. The structures of organization are such that workers are deprived of the resilience they need to adapt to different work environments. It is fair to ask in such a situation: what are the impacts on workers as emerging technologies are increasingly implemented?
Technological Organization And The Fracture Of Managerialism
Critiques of managerialism are nothing new. They have dated back to the earliest arrivals of managers into the workforce, well before they became entrenched. Thinkers from Joseph Schumpeter to James Burnham to Rakesh Khurana have variously analyzed its organizational failures. However, the managerialism so synonymous with modern economic life is today being challenged by the effects of technology and automation within the post-industrial economy itself. The growth of digital economies in the early 21st century, as well as a number of technological innovations which are likely to produce similarly large-scale effects in global economic life, are actively diminishing the human role in empirical analysis. As a result, it is goal-oriented judgment—the very aspect of management which the scientific approach sought to minimize—which will increase in value as the unique contribution of human direction to an organization.
One of the problems with many of the analyses conducted on the impacts of automation on the workplace is the view of technology as an external product applied to an environment. To some degree there are products, such as surveillance robotics, that can be purchased outright and put into use. In most cases, however, new technologies must be integrated into an organization: this requires new infrastructure, new rules, and a new culture. Research suggests that these barriers explain why there can simultaneously be a widespread sense of rapid technological advancement and next to no productivity growth across most of the developed world. Technological development is, as a result, embodied in the organizations which sustain it; knowing about artificial intelligence is not enough to predict its implications. It’s necessary to look at the culture and processes that deploy it.
The recent book Prediction Machines provides a useful framework for dissecting decision-making within organizations and the impact artificial intelligence is likely to have. The authors argue that artificial intelligence is fundamentally a prediction technology, and that decisions can be decomposed into the components of prediction and judgment. As the cost of prediction decreases, the value of judgment increases—this is because managers are able to run through more hypothetical “If-Then” scenarios and articulate with greater precision the ideal course of action for their organization. While the book focuses on artificial intelligence specifically, this framework of decision-making components and organizational workflows is nonetheless useful in understanding the general failures of management with respect to the more data-driven high technologies emerging as part of the Fourth Industrial Revolution.
The anatomy of organizational decision-making in Prediction Machines places the role of artificial intelligence into the execution of tasks. Tasks are components of jobs, and jobs are components of the total work output. In each task, there is an input component, a decision component, an action component, and an output. Artificial intelligence, the authors argue, decomposes the decision component of a task into prediction and judgment, increasing the accuracy of the former, and increasing the value of the latter. In fact, this approach can be applied just as well to connected devices and advanced robotics as to artificial intelligence. Different aspects of a task may increase in value with other emerging high technologies, such as advanced robotics increasing the precision of an executed action. However, it is clear in all these cases that the major human input required is to ensure this workflow is consistent with the goals of the organization. The more complex and important the task, the more human judgment is required to ensure that machines are not misaligned with organizational goals.
While strategies are articulated on how to incorporate artificial intelligence into organizations to increase their capacity for judgment, there are reservations to be had on the efficacy of such initiatives given the managerial culture as it now exists. As judgment grows in value in an institutional culture which has historically devalued it—and as the technical components of decision-making are increasingly automated—the likelihood of error increases. The rising possibility of error within rigidly hierarchical organizations is not only the product of new technologies; it has been a concern ever since earlier waves of computing were industrialized. The same dynamics of complexity and disaggregation of the use of information from its collection process and analysis occurred with the normalization of computers and the internet. Similarly, they are now being intensified by the exponential growth of data and algorithmic decision-making. In response a literature of High Reliability Organizations (HROs) has begun to grow, though adaptation has been minimal outside of militaries.
There are five criteria of High Reliability Organizations that enable them to allow for risk prevention and the exercise of good judgment: a preoccupation with failure, reluctance to simplify interpretations, deep situational awareness, commitment to resilience, and flexibility in structure. All of these, to varying degrees, are at odds with the scientific and deterministic bureaucratic culture that modern managerialism has ingrained. Flexibility in structure, for example, requires a decentralized approach:
Any person can raise an alarm and halt operations. When anomalies or near misses arise, their descriptions are propagated throughout the organization, rather than following a fixed reporting path, in the hopes that the person with the right expertise will see them.
Such an organizational culture is rare, especially in the public administrative institutions where the vital functions of statecraft are executed.
Organizations that increasingly seek technical solutions devalue the dynamism that human creativity allows for, which in turn encourages them to cede human decision-making to machines, and simultaneously to make humans more machine-like. The value of intuition is lost, preventing the sorts of risk awareness essential to the exercise of good judgment. Politically speaking, the populist wave has invoked rhetoric denouncing this technocratic view of administration. Modern populism is also more common in societies further along the transition to “post-industrial economies,” where work is increasingly automated, and the functions of management and administration take up an increasingly large chunk of national income. Where jobs reflect individual talents, they are augmented by new technologies, rather than automated by them. But where workers are overspecialized to the point that what they do differs little from a machine, they are more likely to be automated in full, and their intuition discounted. This overspecialization is all too common, and goes against the fundamental basis of a High Reliability Organization.
While there is a growing literature on both human-computer interaction and worker augmentation, the results depend largely on whether organizations adopt a culture which allows technologies to be developed and implemented in a humanistic way. A technocratic culture in which only means are of concern is one chiefly dependent on the quality of its programmers and engineers. Conversely, one in which qualitative work is given value and judgment is prioritized will more likely be one in which machines are tools at the service of enhancing an individual.
The common line of thinking among managers and technocrats is that technologies are neutral tools dedicated to specified ends—in fact, no different than how laborers and capital inputs are generally treated. But this is not the case. Technologies actively impact the broader structure into which they are integrated. Their purpose can only be fully understood by their functional ends within the broader structure. Technologies develop out of certain intentions, are deployed through certain institutional arrangements, and incentivize certain forms of behavior. Much current discussion focuses on the regulation of technologies, such as the surveillance or military applications of artificial intelligence. However, this marginal form of regulation alone will not be sufficient to ensure that new technologies are well-integrated into a culture which actually builds and sustains innovation.
Humanistic Organization And The Ends Of Technology
One of the most fascinating promises of the Fourth Industrial Revolution is the advent of “Smart Cities,” which through digital technologies will become constantly connected and vastly more efficient forms of urban organization. Cities are an interesting form of organization, relying on complex interdependent systems, from public services to local businesses to community groups which help sustain and grow them. The diversity of cities in terms of organizational tasks allows them to integrate the full spectrum of new technologies into their holistic development.
A few simple examples suffice to show how interconnected the result could be. With self-driving cars, the efficiency of traffic would increase, and the greater intensity with which these vehicles could be used would significantly reduce demand for parking space. This newly freed space could be used to make cities denser, allowing more constant interaction and the development of “hubs” where like minded people would congregate and develop new and better ideas. Public administration would need to ensure that these efficiencies are well-integrated, digitizing their citizen-facing infrastructure. Connected devices would proliferate across public infrastructure, vastly increasing the data generated by which private companies would be able to innovate further.
In what was just described, however, nothing was mentioned regarding the ultimate purpose of these gains to efficiency or the forms which these new interconnected structures might take. This is because the technologies themselves only expand material possibility, but the ends to which they are applied cannot be inferred from the techniques they use. Connected cities in China, where the surveillance apparatus has vastly expanded due to the advent of artificial intelligence and advanced robotics, look very different from the smart city project currently underway in Barcelona, Spain.
To understand their differences, however, their cultures, rather than their institutions, must be understood. China’s approach to centralizing control through technology and dictating their citizens’ lives with greater precision fits into a historical cultural value that sees the role of the state as upholding moral order. By way of contrast,the Barcelona project, with its goals of reducing costs and improving the environment, cedes behavioral regulation to a technological environment rather than top-down authority. This fits with a study done by MIT that found that workers prefer taking orders from machines than people, a very technocratic notion of placing our trust in those who are seen as efficient and unbeholden to all-too-human games of social status and intrigue. This is in no way to create a moral equivalency between different value systems, but simply to highlight the cultural bases by which technologies are deployed.
Questions about how the Fourth Industrial Revolution can serve societal welfare, then, are not about merely regulating the particulars or dictating how means are pursued, but rather a broader discourse about the ends to which organizations work. It requires that decision-makers are generalists who understand not just their own tasks, but the social dynamics of their decisions.
There have been attempts to attach managers to their ends, such as through the rise of the shareholder-CEOs in the 1980s, to ensure a commitment to a vision; this development has had mixed results, if not negative ones. While management literature is aware of the problems of bureaucracy, the real world continues to grow in administrative bloat. Even MacIntyre’s scathing critique of management has become one of the most widely cited works in the field of business ethics since its publication in 1980, though its impact is debatable.
The problem is that going against the technocratic impulse of modernity requires answers that are fuzzy and imprecise. It requires workers who both feel connected to their craft, but are aware of their place relative to the wider organization. It requires managers to have a stake in the organization’s goals, to be connected to the social environment in which they operate, and to have a familiarity with their workers’ tasks. It requires that discussion of values can exist alongside and temper technical reports and that intuition be allowed to override data. It requires organizational structures that are less well-defined and processes that are more adaptable. Fundamentally, it requires abandoning the view that decision-making is scientific and deterministic, which is far less palatable to most than particular policy solutions.
The sustainability of a technical, scientific bureaucratic process is in many ways derived from its inscrutability. When a field is made to have the apparent complexity of physics, it restricts participation and dismisses the critiques of those without the relevant technical training as unscientific. The overbearing elitist state or corporate overlord is a more easy-to-identify threat, but resistance to these forms of authority prevents their reach from becoming totalitarian. Bureaucratic creep, however, persists in ways which are much harder to challenge. Unless this is overcome in the transition through the Fourth Industrial Revolution, new technologies will only serve to intensify managerial forms of work and life, as well as their tendency to escape any real form of human direction.