PART ONE
Design in Society
1
Designing Ethics in Large-scale Socio-technical Systems
Jeffrey Chan
Introduction: Designing Ethics?
In what ways are design and ethics connected? Philosopher Glenn Parsons outlines three major ways.1 Firstly, ethics exists in the form of rules and norms to regulate conduct in design practices. Typically, ethics is framed in the form of normative theories.2 For instance, utilitarianism or deontological ethics are applied to examine issues and dilemmas in design practices in order to evaluate their impacts and moral significance. This process produces ethical knowledge, which then guides responsible action. Secondly, ethics is also fundamentally concerned with what the designer creates. To rely on a drastic example here: should a designer elect to design a thermonuclear weaponâa weapon that exceeds all conceivable military purposes and armed only with the singular objective of annihilating large cities?3 Perceived this way, this fundamental concern of ethics in design approximates what Horst Rittel described as the paramount question of âwhat to design and what not to design.â4 Finally, and in a reversal of directionality from design to ethics, the new possibilities engendered by design may transform existing ethical values. An instance of this is how technological design could normalize inequality, where a desirable advancement for a technologically privileged group is attained at the expense of an undesirable regression for other social groups that are either less technologically savvy or have little access to advanced technologies.5 This reality has been routinely observed in how premium technologies (for example, fiber optic network cable)6 that are embedded in new networksâwhich are also only connected to selected users and places while bypassing social groups and places that cannot afford them7 âhave hardened differences and intensified inequality in many smart cities.8
But is there also a fourth connection? Can ethics be designed? Or stated differently, how can ethics be changed or transformed by design? To visualize the possibilities implied by these questions, a threefold distinction is first necessary. Firstly, ethics often prompts design.9 The discovery of a genuine ethical problem is often closely followed by figuring out what one ought to doâto devise a plan of action, which is to design. This is then Caroline Whitbeckâs provocative analogy of âethics as design.â10 The target of design here is likely a safer, or a more equitable, solution to an engineering problem. And as argued, this new design may also prompt changes in ethics. Even so, these changes, if any, are merely the unintended consequencesârather than the original and intended goalsâof the design intervention. For example, Temple Grandin, professor of animal science, has spent many years designing slaughterhouses that are more âethicalâ so that livestock could be less stressed before their final moments.11 But her design has unintentionally raised an acute awareness on the rampant cruelty of the livestock industry that, in turn, has advanced food ethics.
Secondly, moral features could be built into the design, or what has been referred to as âdesigning in ethics.â12 As a recent development of institutional design,13 âdesigning in ethicsâ argues that ethical relations are often mediated by institutions and technologies, and for this reason, their design can constrain possible and available ethical choices. In other words, it may be possible to design these institutions and technologies in ways that they are more likely to produce an ethical outcome. Here, consider the automated emergency hotline system: if the bot on the other end of the call has determined that the emergency at hand is not urgent by its triage logics, and furthermore could only offer preset options in response, then what ought the caller do? This caller, who is confounded by the many shades of any genuine real-life emergency, has little discretion either to persuade or to override the botâs automated judgment.14 In this example, could the algorithm be redesigned to also include a moral âfail-safe,â which will redirect the call to a human emergency responder among other comparable alternatives?
Thirdly, designing ethics is far more foundational; it aims to alter the moral ecology, which is an interdependent system comprising of reinforcing norms, values, and institutions. Here, the key variable for intervention is always humanity, and the outcome is inevitably a new human condition. Unlike âdesigning in ethics,â which is designing the mediating institutions and technologies so that they are more likely to produce an ethical outcome, designing ethics is closer to the creation of a new set of moral codesâsetting into place a new outlook for the human society that first jettisons existing moral codes and then fundamentally recalibrates what is good and bad, right and wrong. Throughout recorded history, entire moral ecologies have been periodically overhauled, for instance, during the Enlightenment, or more violently overturned, as witnessed in the Third Reich. While these examples are instructive evidence that moral ecologies exist, and moreover that sociopolitical turmoil tends to follow their destabilization, they neither offer any decisive insight on how new moral ecologies could be specifically designed with as little social destruction as possible, nor how to design them at scales far smaller than entire societies.
Research Questions, Methods, and Contributions
Admittedly, the primary question of âcan ethics be designed?â and the conceptual terrain raised by this question has yet to be systematically broached in either design or ethics. On the one hand, design has been defined as the act of devising plans to change certain existing situations into preferred ones.15 Conceivably, one could design an artifact, a process, a service, and even an organization.16 However, what does the design of ethics entail substantively or procedurally? On the other hand, traditional ethics h as frequently been concentrated on the problems of individual conduct, the moral qualities of human goals, obligations, and aspirations.17 In this way, ethics calibrates the relational quality between persons as moral beings.18 If so, then to what extent, and in what ways, is ethics even amenable to design? In nearly 3,000 years of Occidental ethics, perhaps the closest proxy has been the utopiaâor what Karl Popper recasts in the language of design as âutopian engineering.â19 Instead of building a new society fit for people to live in, utopian engineering tries to mold people to conform to an ideal society, usually conceived by a single individual or a small group of visionaries.20 Nevertheless, utopiasâfor example, the community of New Harmony as envisioned by Robert Owen (1771â1858)âtend to fail because, by necessity, they have to stabilize and control the dynamic processes that were once mobilized to build them.21 And because the utopia is essentially an attempt to arrest change, which is unceasing for any complex system, the utopia tends to demand many improvisations in order to remain in an immutable stateâimprovisations that in turn lead to even more surprises that would ultimately derail its original ideal.22 For these reasons, even if utopian engineering might suggest a concrete possibility for designing ethics, this form of large-scale social design is neither its most effective nor its most ethical benchmark.
Considering the untenable example of utopian engineering, this chapter instead broaches the idea of designing ethics by examining two contemporary design case studies of moral significance. They are, namely, dockless bikeshare programs and the controversy surrounding autonomous vehicles. These two cases appear to demonstrate the necessity for designing ethicsâalthough in diametrically different ways. While the former case considers how ethics could be designed by modifying the artifact as well as the moral ecology of the dockless bike system, the latter case suggests the need to directly program autonomous vehicles (or self-driving cars), ex-ante, for moral reasoning. In sum, these two cases begin to demonstrate the relevance of designing ethics in large-scale socio-technical systems today.
The Dockless Bikeshare Program: Designing the Moral Ecology
The first bikeshare program (henceforth, BSP), Witte Fietsen (White Bikes), was launched in 1965 in Amsterdam. However, it is in recent years that BSP has experienced rapid growth and now finds adoption in cities around the world.23 As a rough indication of the rapid growth of the BSP, thirteen cities operated a BSP in 2004 but by 2014, 855 cities had a BSP.24 This number is surely set to grow even further based on the trend of new cities adopting the dockless BSP. The dockless system is the fourth generation of BSP, which is the most recent evolution of the bikeshare system after three successive generations of BSP: namely, the first-generation Witte Fietsen (White Bikes) BSP, the second-generation large-scale bikeshare launched in Copenhagen in 1995, and the third-generation BSP characterized by dedicated docking stations where one could retrieve and return the bicycles.25 In the fourth-generation dockless BSP system, users first download a smartphone app, which subsequently unlocks the bicycle. And at the end of a userâs journey, this user relies on the same app again to lock the bicyc...