We argue that the later Wittgenstein’s philosophy of language and mathematics, substantially focused on rule-following, is relevant to understand and improve on the Artificial Intelligence (AI) alignment problem: his discussions on the categories that influence alignment between humans can inform about the categories that should be controlled to improve on the alignment problem when creating large data sets to be used by supervised and unsupervised learning algorithms, as well as when introducing hard coded guardrails for AI models. We cast these (...) considerations in a model of human–human and human–machine alignment and sketch basic alignment strategies based on these categories and further reflections on rule-following like the notion of meaning as use. To sustain the validity of these considerations, we also show that successful techniques employed by AI safety researchers to better align new AI systems with our human goals are congruent with the stipulations that we derive from the later Wittgenstein’s philosophy. However, their application may benefit from the added specificities and stipulations of our framework: it extends on the current efforts and provides further, specific AI alignment techniques. Thus, we argue that the categories of the model and the core alignment strategies presented in this work can inform further AI alignment techniques. (shrink)

Purpose Fully autonomous self-driving cars not only hold the potential for significant economic and environmental advantages but also introduce complex ethical dilemmas. One of the highly debated issues, known as the “trolley problems,” revolves around determining the appropriate actions for a self-driving car when faced with an unavoidable crash. Currently, the discourse on autonomous vehicle (AV) crash algorithms is primarily shaped by Western ethical traditions, resulting in a Eurocentric bias due to the dominant economic and political influence of the West. (...) However, considering that AV technology will be deployed across diverse cultural and religious contexts, this paper aims to contribute to the discourse by providing an Islamic perspective on programming the response of AVs in the event of an imminent crash. Design/methodology/approach This study proposes a novel methodology based on the Islamic concept of maṣlaḥa for the normative assessment of ethical decisions related to AV programming. Findings Drawing upon the works of classic Islamic jurists, this study highlights two distinct normative visions within Islamic traditions (akin to deontology and consequentialism) concerning the preservation of human lives in the context of AVs. This study explores the shared and divergent elements between Islamic and Western ethical approaches proposed for AVs. Originality/value This pioneering work examines AV crash algorithms from an Islamic perspective, filling a void in the global ethical discourse. This work will also serve an important role to bridge the gap between the theoretical Islamic ethical principles and their practical application in the realm of AVs. (shrink)

Artificially intelligent systems will be used to make increasingly important decisions about us. Many of these decisions will have to be made without universal agreement about the relevant moral facts. For other kinds of disagreement, it is at least usually obvious what kind of solution is called for. What makes moral disagreement especially challenging is that there are three different ways of handling it. _Moral solutions_ apply a moral theory or related principles and largely ignore the details of the disagreement. (...) _Compromise solutions_ apply a method of finding a compromise and taking information about the disagreement as input. _Epistemic solutions_ apply an evidential rule that treats the details of the disagreement as evidence of moral truth. Proposals for all three kinds of solutions can be found in the AI ethics and value alignment literature, but little has been said to justify choosing one over the other. I argue that the choice is best framed in terms of _moral risk_. (shrink)

In 2021, Germany passed the first law worldwide that regulates dilemma situations with autonomous cars. Against this background, this article investigates the permissibility of trade-offs between human lives in the context of self-driving cars. It does so by drawing on the debate about the traditional trolley problem. In contrast to most authors in the relevant literature, it argues that the debate about the trolley problem is both directly and indirectly relevant for the ethics of crashes with self-driving cars. Drawing on (...) its direct normative relevance, the article shows that trade-offs are permissible in situations with self-driving cars that are similar to paradigmatic trolley cases. In scenarios that are unlike paradigmatic trolley cases, the debate about the trolley problem can have indirect normative relevance because it provides reasons against the use of moral theories and principles that cannot account for the trolley problem. (shrink)

Focused on city-scale automation, and using self-driving cars (SDCs) as a case study, this article reflects on the role of AI—and in particular, computer vision systems used for mapping and navigation—as a catalyst for urban transformation. Urban research commonly presents AI and cities as having a one-way cause-and-effect relationship, giving undue weight to AI’s impact on cities and overlooking the role of cities in shaping AI. Working at the intersection of data science and social research, this paper aims to counter (...) this trend by exploring the reverse perspective: how do cities affect the development, and expose the present limits, of SDCs? The contribution of this paper is threefold. First, by comparing urban and nonurban environments and thoroughly examining the relationship between computer vision and city-specific sociality and form, it defines machine autonomy/automation as a function of the sociotechnical milieu in which an AI system operates. Second, and related, the paper problematizes the notion of SDCs as autonomous technologies and the role it plays in envisioning contending policy arrangements and technical solutions for achieving full driving automation. Finally, the article offers insight into a materialist and spatialized understanding of AI—namely, not as an abstract quality susceptible to replication within discrete machines, but rather as a distributed property emerging through embodied interactions among a multiplicity of agents (human, non-human, and technological) within/with their environments. (shrink)

Many researchers from robotics, machine ethics, and adjacent fields seem to assume that norms represent good behavior that social robots should learn to benefit their users and society. We would like to complicate this view and present seven key troubles with norm-compliant robots: (1) norm biases, (2) paternalism (3) tyrannies of the majority, (4) pluralistic ignorance, (5) paths of least resistance, (6) outdated norms, and (7) technologically-induced norm change. Because discussions of why norm-compliant robots can be problematic are noticeably absent (...) from the robot and machine ethics literature, this paper fills an important research gap. We argue that it is critical for researchers to take these issues into account if they wish to make norm-compliant robots. (shrink)

This paper provides a systematic account of how artificial intelligence (AI) technologies could harm nonhuman animals and explains why animal harms, often neglected in AI ethics, should be better recognised. After giving reasons for caring about animals and outlining the nature of animal harm, interests, and wellbeing, the paper develops a comprehensive ‘harms framework’ which draws on scientist David Fraser’s influential mapping of human activities that impact on sentient animals. The harms framework is fleshed out with examples inspired by both (...) scholarly literature and media reports. This systematic account and framework should help inform ethical analyses of AI’s impact on animals and serve as a comprehensive and clear basis for the development and regulation of AI technologies to prevent and mitigate harm to nonhumans. (shrink)

Is ethics a computable function? Can machines learn ethics like humans do? If teaching consists in no more than programming, training, indoctrinating… and if ethics is merely following a code of conduct, then yes, we can teach ethics to algorithmic machines. But if ethics is not merely about following a code of conduct or about imitating the behavior of others, then an approach based on computing outcomes, and on the reduction of ethics to the compilation and application of a set (...) of rules, either a priori or learned, misses the point. Our intention is not to solve the technical problem of machine ethics, but to learn something about human ethics, and its rationality, by reflecting on the ethics that can and should be implemented in machines. Any machine ethics implementation will have to face a number of fundamental or conceptual problems, which in the end refer to philosophical questions, such as: what is a human being (or more generally, what is a worthy being); what is human intentional acting; and how are intentional actions and their consequences morally evaluated. We are convinced that a proper understanding of ethical issues in AI can teach us something valuable about ourselves, and what it means to lead a free and responsible ethical life, that is, being good people beyond merely “following a moral code”. In the end we believe that rationality must be seen to involve more than just computing, and that value rationality is beyond numbers. Such an understanding is a required step to recovering a renewed rationality of ethics, one that is urgently needed in our highly technified society. (shrink)

My starting point in this article is that investigating the ethics of autonomous vehicles through the lens of the trolley problem is not only limited but also unethical. I construct my case by aligning myself with Niklas Toivakainen, who argues against David Gunkel’s reading of Levinasian ethics as an answer to the “Machine Question”. I adumbrate Toivakainen’s critique that the attempt to give a Levinasian face to the machine is an example of a compensatory logic – a way to avoid (...) a deeper exploration of the moral dynamics from which our technologies emerge. I offer an extended reading of Levinas’s formulation of the face of the other to argue that while the machine cannot signify as a face, the machine can announce or anticipate the signification of the face which the self is infinitely responsible to. I reach this conclusion through focusing on the uncertainty and ambiguity of the signification of the face. I then argue that current approaches to the ethics of autonomous vehicles, based on variations of the trolley problem, fail to locate ethical responsibility to the other, precisely in its attempts to evade uncertainty. I demonstrate how an embrace of uncertainty can make the ethics of autonomous vehicles more ethical through a consideration of the concept of shared space and naked streets in the work of the late Dutch traffic engineer Hans Monderman. (shrink)

The dominant paradigm of experiments in the social and behavioral sciences views an experiment as a test of a theory, where the theory is assumed to generalize beyond the experiment's specific conditions. According to this view, which Alan Newell once characterized as “playing twenty questions with nature,” theory is advanced one experiment at a time, and the integration of disparate findings is assumed to happen via the scientific publishing process. In this article, we argue that the process of integration is (...) at best inefficient, and at worst it does not, in fact, occur. We further show that the challenge of integration cannot be adequately addressed by recently proposed reforms that focus on the reliability and replicability of individual findings, nor simply by conducting more or larger experiments. Rather, the problem arises from the imprecise nature of social and behavioral theories and, consequently, a lack of commensurability across experiments conducted under different conditions. Therefore, researchers must fundamentally rethink how they design experiments and how the experiments relate to theory. We specifically describe an alternative framework, integrative experiment design, which intrinsically promotes commensurability and continuous integration of knowledge. In this paradigm, researchers explicitly map the design space of possible experiments associated with a given research question, embracing many potentially relevant theories rather than focusing on just one. Researchers then iteratively generate theories and test them with experiments explicitly sampled from the design space, allowing results to be integrated across experiments. Given recent methodological and technological developments, we conclude that this approach is feasible and would generate more-reliable, more-cumulative empirical and theoretical knowledge than the current paradigm – and with far greater efficiency. (shrink)

Crashes involving self-driving cars at least superficially resemble trolley dilemmas. This article discusses what lessons machine ethicists working on the ethics of self-driving cars can learn from trolleyology. The article proceeds by providing an account of the trolley problem as a paradox and by distinguishing two types of solutions to the trolley problem. According to an optimistic solution, our case intuitions about trolley dilemmas are responding to morally relevant differences. The pessimistic solution denies that this is the case. An optimistic (...) solution would yield first-order moral insights for the ethics of self-driving cars, but such a solution is difficult to come by. More plausible is the pessimistic solution, and it teaches us a methodological lesson. The lesson is that machine ethicists should discount case intuitions and instead rely on intuitions and judgments at a higher level of generality. (shrink)

As Artificial Intelligence (AI) becomes more and more pervasive in our everyday life, new questions arise about its ethical and social impacts. Such issues concern all stakeholders involved in or committed to the design, implementation, deployment, and use of the technology. The present document addresses these preoccupations by introducing and discussing a set of practical obligations and recommendations for the development of applications and systems based on AI techniques. With this work we hope to contribute to spreading awareness on the (...) many social challenges posed by AI and encouraging the establishment of good practices throughout the relevant social areas. As points of novelty, the paper elaborates on an integrated view that combines both human rights and ethical concepts to reap the benefits of the two approaches. Moreover, it proposes innovative recommendations, such as those on redress and governance, which add further insight to the debate. Finally, it incorporates a specific focus on the Italian Constitution, thus offering an example of how core legislations of Member States might contribute to further specify and enrich the EU normative framework on AI. (shrink)

Ethical morality is one of the significant issues in self-driving cars. The paper provides a newer approach to solve the ethical decision problems in self-driving cars until there is no concrete ethical decision to all problems. This paper gives a two-way approach to solve a problem, with first being the mapping of problem to the solution already known or which has a fixed set of solutions and action priorities defined to a problem previously. Now, if no solution is found or (...) mapping is unsuccessful, then the second stage activates, where the solution from Deep Q-learning model is calculated. It estimates the best Q value and returns that solution or action which maximizes the reward at that instance. The reward function is designed with decreasing priorities and acts accordingly, where the users can change or define their priorities if needed. The case study and results show that the solution that is present in the paper will lead to solving ethical morality problems in self-driving cars up to a great extent. (shrink)

Reflective equilibrium —the idea that we have to justify our judgments and principles through a process of mutual adjustment—is taken to be a central method in philosophy. Nonetheless, conceptions of RE often stay sketchy, and there is a striking lack of explicit and traceable applications of it. This paper presents an explicit case study for the application of an elaborate RE conception. RE is used to reconstruct the arguments from Thomson’s paper “Turning the Trolley” for why a bystander must not (...) divert a runaway trolley from five workmen onto one. Analyzing Thomson’s resulting position with the RE-criteria has two main results: Firstly, the adjustment of one of her commitments can be defended. Secondly, no justified position in RE was reached. With respect to RE as a method, the main results from this application are: There is at least one conception of RE that is sufficiently specified to be applicable; the RE criteria put real constraints on the process of justification; and an explicit application of RE has benefits in terms of clarity while at the same time providing guidance for how the justificatory process could be continued. (shrink)

The potential development of self-driving cars (also known as autonomous vehicles or AVs – particularly Level 5 AVs) has called the attention of different interested parties. Yet, there are still only a few relevant international regulations on them, no emergency patterns accepted by communities and Original Equipment Manufacturers (OEMs), and no publicly accepted solutions to some of their pending ethical problems. Thus, this paper aims to provide some possible answers to these moral and practical dilemmas. In particular, we focus on (...) what AVs should do in no-win scenarios and on who should be held responsible for these types of decisions. A naturalistic perspective on ethics informs our proposal, which, we argue, could represent a pragmatic and realistic solution to the regulation of AVs. We discuss the proposals already set out in the current literature regarding both policy-making strategies and theoretical accounts. In fact, we consider and reject descriptive approaches to the problem as well as the option of using either a strict deontological view or a solely utilitarian one to set AVs’ ethical choices. Instead, to provide concrete answers to AVs’ ethical problems, we examine three hierarchical levels of decision-making processes: country-wide regulations, OEM policies, and buyers’ moral attitudes. By appropriately distributing ethical decisions and considering their practical implications, we maintain that our proposal based on ethical naturalism recognizes the importance of all stakeholders and allows the most able of them to take actions (the OEMs and buyers) to reflect on the moral leeway and weight of their options. (shrink)

Robert Nozick famously raised the possibility that there is a sense in which both deontology and utilitarianism are true: deontology applies to humans while utilitarianism applies to animals. In recent years, there has been increasing interest in such a hybrid views of ethics. Discussions of this Nozickian Hybrid View, and similar approaches to animal ethics, often assume that such an approach reflects the commonsense view, and best captures common moral intuitions. However, recent psychological work challenges this empirical assumption. We review (...) evidence suggesting that the folk is deontological all the way down—it is just that the moral side constraints that protect animals from harm are much weaker than those that protect humans. In fact, it appears that people even attribute some deontological protections, albeit extremely weak ones, to inanimate objects. We call this view Multi-level Weighted Deontology. While such empirical findings cannot show that the Nozickian Hybrid View is false, or that it is unjustified, they do remove its core intuitive support. That support belongs to Multi-level Weighted Deontology, a view that is also in line with the view that Nozick himself seemed to favour. To complicate things, however, we also review evidence that our intuitions about the moral status of humans are, at least in significant part, shaped by factors relating to mere species membership that seem morally irrelevant. We end by considering the potential debunking upshot of such findings about the sources of common moral intuitions about the moral status of animals. (shrink)

In this manuscript, we introduce a theoretical model of climate radicalization that integrates social psychological theories of perceived unfairness with historical insights on radicalization to contribute to the knowledge of individuals’ processes of radicalization and non-radicalization in relation to climate change. We define climate radicalization as a process of growing willingness to pursue and/or support radical changes in society that are in conflict with or could pose a threat to the status quo or democratic legal order to reach climate goals. (...) We describe how perceptions of unfairness can play a pivotal role in processes of climate change related radicalization. Without taking any position or judgment regarding climate concerns and associated actions, we suggest that although these behaviors drive many people to participate in peaceful climate protest, they may also lead others to radicalize into breaking the law to achieve their climate goals, possibly in violent ways. This process of climate radicalization, we argue, can be driven by people perceiving certain situations to be blatantly unfair. Specifically, we discuss how radical attitudes and behaviors can be products of perceived unfairness stemming from the past, the future, the immediate social environments of perceivers, as well as those that are spatially distant from them. We further argue that because radicalization processes are shaped by an interaction between individuals and movements, on the one hand, and societal actors and developments, on the other, they tend to develop in non-linear and dynamic ways. We therefore propose that climate radicalization is a dynamic, contingent, and non-linear process, often of an escalating kind, that develops over time, through various interactions between individuals and their contexts, and in which people and groups move back and forth from peaceful protest, through disobedient and unlawful methods, to violent actions. Implications, strengths, and limitations of our model are discussed. (shrink)

What is techno-optimism and how can it be defended? Although techno-optimist views are widely espoused and critiqued, there have been few attempts to systematically analyse what it means to be a techno-optimist and how one might defend this view. This paper attempts to address this oversight by providing a comprehensive analysis and evaluation of techno-optimism. It is argued that techno-optimism is a pluralistic stance that comes in weak and strong forms. These vary along a number of key dimensions but each (...) shares the view that technology plays a key role in ensuring that the good prevails over the bad. Whatever its strength, to defend this stance, one must flesh out an argument with four key premises. Each of these premises is highly controversial and can be subjected to a number of critiques. The paper discusses five such critiques in detail. The paper also considers possible responses from the techno-optimist. Finally, it is concluded that although strong forms of techno-optimism are not intellectually defensible, a modest, agency-based version of techno-optimism may be defensible. (shrink)

One of the moral questions concerning autonomous vehicles (henceforth AVs) is the choice between types that differ in their built-in algorithms for dealing with rare situations of unavoidable lethal collision. It does not appear to be possible to avoid questions about how these algorithms should be designed. We present the results of our study of moral preferences (N = 2769) with respect to three types of AVs: (1) selfish, which protects the lives of passenger(s) over any number of bystanders; (2) (...) altruistic, which minimizes the number of casualties, even if this leads to death of passenger(s); and (3) conservative, which abstains from interfering in such situations. We differentiate between scenarios in which participants are to make their decisions privately or publicly, and for themselves or for their offspring. We aim to answer two research questions: (1) whether the public visibility of the choice of an AV type choice make this choice more altruistic and (2) which type of situation makes it more difficult to choose altruistically: when choosing for society as a whole, when choosing only for oneself, or when choosing only for one’s offspring. Our results show that respondents exhibit a preference for an altruistic strategy for AVs and that it is reinforced when signaled to others. The altruistic preference is strongest when applies to everybody else, weaker when it reflects a solely personal choice, and weakest when choosing for one’s own child. We conclude that a public choice is considerably more likely to pressure consumers into accepting a more socially beneficial solution. (shrink)

The aim of this paper is to better qualify the problem of online trust. The problem of online trust is that of evaluating whether online environments have the proper design to enable trust. This paper tries to better qualify this problem by showing that there is no unique answer, but only conditional considerations that depend on the conception of trust assumed and the features that are included in the environments themselves. In fact, the major issue concerning traditional debates surrounding online (...) trust is that those debates focus on specific definitions of trust and specific online environments. Ordinarily, a definition of trust is assumed and then environmental conditions necessary for trust are evaluated with respect to such specific definition. However, this _modus operandi_ fails to appreciate that trust is a rich concept, with a multitude of meanings and that there is still no strict consensus on which meaning shall be taken as the proper one. Moreover, the fact that online environments are constantly evolving and that new design features might be implemented in them is completely ignored. In this paper, the richness of the philosophical discussions about trust is brought into the analysis of online trust. I first provide a set of conditions that depend on the definition of trust that can be assumed and then discuss those conditions with respect to the design of online environments in order to determine whether they can enable (and under which circumstances) trust. (shrink)

The German Act on Autonomous Driving constitutes the first national framework on level four autonomous vehicles and has received attention from policy makers, AI ethics scholars and legal experts in autonomous driving. Owing to Germany’s role as a global hub for car manufacturing, the following paper sheds light on the act’s position within the ethical discourse and how it reconfigures the balance between legislation and ethical frameworks. Specifically, in this paper, we highlight areas that need to be more worked out (...) in the future either through ethical conventions, corporate measures or legal measures and examine how the law can be incorporated into the existing discourse on the regulation of technologies. Based on this examination, we derive implications for future discourse and elaborate on companies’ responsibilities in developing autonomous driving technologies in an ethical sense. (shrink)

Machine intelligence already helps medical staff with a number of tasks. Ethical decision-making, however, has not been handed over to computers. In this proof-of-concept study, we show how an algorithm based on Beauchamp and Childress’ prima-facie principles could be employed to advise on a range of moral dilemma situations that occur in medical institutions. We explain why we chose fuzzy cognitive maps to set up the advisory system and how we utilized machine learning to train it. We report on the (...) difficult task of operationalizing the principles of beneficence, non-maleficence and patient autonomy, and describe how we selected suitable input parameters that we extracted from a training dataset of clinical cases. The first performance results are promising, but an algorithmic approach to ethics also comes with several weaknesses and limitations. Should one really entrust the sensitive domain of clinical ethics to machine intelligence? (shrink)

There is a concern that the widespread deployment of autonomous machines will open up a number of ‘responsibility gaps’ throughout society. Various articulations of such techno-responsibility gaps have been proposed over the years, along with several potential solutions. Most of these solutions focus on ‘plugging’ or ‘dissolving’ the gaps. This paper offers an alternative perspective. It argues that techno-responsibility gaps are, sometimes, to be welcomed and that one of the advantages of autonomous machines is that they enable us to embrace (...) certain kinds of responsibility gap. The argument is based on the idea that human morality is often tragic. We frequently confront situations in which competing moral considerations pull in different directions and it is impossible to perfectly balance these considerations. This heightens the burden of responsibility associated with our choices. We cope with the tragedy of moral choice in different ways. Sometimes we delude ourselves into thinking the choices we make were not tragic ; sometimes we delegate the tragic choice to others ; sometimes we make the choice ourselves and bear the psychological consequences. Each of these strategies has its benefits and costs. One potential advantage of autonomous machines is that they enable a reduced cost form of delegation. However, we only gain the advantage of this reduced cost if we accept that some techno-responsibility gaps are virtuous. (shrink)

The control principle implies that people should not feel guilt for outcomes beyond their control. Yet, the so-called ‘agent and observer puzzles’ in philosophy demonstrate that people waver in their commitment to the control principle when reflecting on accidental outcomes. In the context of car accidents involving conventional or autonomous vehicles, Study 1 established that judgments of responsibility are most strongly associated with expressions of guilt–over and above other negative emotions, such as sadness, remorse or anger. Studies 2 and 3 (...) then confirmed that, while people generally endorse the control principle, and deny that occupants in an AV should feel guilt when involved in an accident, they nevertheless ascribe guilt to those same occupants. Study 3 also uncovered novel implications of the observer puzzle in the legal context: Passengers in an AV were seen as more legally liable than either passengers in a conventional vehicle, or even their drivers–especially when participants were prompted to reflect on the passengers’ affective experience of guilt. Our findings document an important conflict–in the context of AV accidents–between people’s prescriptive reasoning about responsibility and guilt on one hand, and their counter-normative experience of guilt on the other, with apparent implications for liability decisions. (shrink)

Organizations are making massive investments in artificial intelligence, and recent demonstrations and achievements highlight the immense potential for AI to improve organizational and human welfare. Yet realizing the potential of AI necessitates a better understanding of the various ethical issues involved with deciding to use AI, training and maintaining it, and allowing it to make decisions that have moral consequences. People want organizations using AI and the AI systems themselves to behave ethically, but ethical behavior means different things to different (...) people, and many ethical dilemmas require trade-offs such that no course of action is universally considered ethical. How should organizations using AI—and the AI itself—process ethical dilemmas where humans disagree on the morally right course of action? Though a variety of ethical AI frameworks have been suggested, these approaches do not adequately address how people make ethical evaluations of AI systems or how to incorporate the fundamental disagreements people have regarding what is and is not ethical behavior. Drawing on moral foundations theory, we theorize that a person will perceive an organization’s use of AI, its data procedures, and the resulting AI decisions as ethical to the extent that those decisions resonate with the person’s moral foundations. Since people hold diverse moral foundations, this highlights the crucial need to consider individual moral differences at multiple levels of AI. We discuss several unresolved issues and suggest potential approaches for thinking about conflicts in moral judgments concerning AI. (shrink)

Several technological developments, such as self-service technologies and artificial intelligence, are disrupting the retailing industry by changing consumption and purchase habits and the overall retail experience. Although AI represents extraordinary opportunities for businesses, companies must avoid the dangers and risks associated with the adoption of such systems. Integrating perspectives from emerging research on AI, morality of machines, and norm activation, we examine how individuals morally behave toward AI agents and self-service machines. Across three studies, we demonstrate that consumers’ moral concerns (...) and behaviors differ when interacting with technologies versus humans. We show that moral intention is less likely to emerge for AI checkout and self-checkout machines compared with human checkout. In addition, moral intention decreases as people consider the machine less humanlike. We further document that the decline in morality is caused by less guilt displayed toward new technologies. The non-human nature of the interaction evokes a decreased feeling of guilt and ultimately reduces moral behavior. These findings offer insights into how technological developments influence consumer behaviors and provide guidance for businesses and retailers in understanding moral intentions related to the different types of interactions in a shopping environment. (shrink)

This paper argues against the call to democratize artificial intelligence (AI). Several authors demand to reap purported benefits that rest in direct and broad participation: In the governance of AI, more people should be more involved in more decisions about AI—from development and design to deployment. This paper opposes this call. The paper presents five objections against broadening and deepening public participation in the governance of AI. The paper begins by reviewing the literature and carving out a set of claims (...) that are associated with the call to “democratize AI”. It then argues that such a democratization of AI (1) rests on weak grounds, because it does not answer to a demand of legitimization, (2) is redundant in that it overlaps with existing governance structures, (3) is resource intensive, which leads to injustices, (4) is morally myopic and thereby creates popular oversights and moral problems of its own, and finally, (5) is neither theoretically nor practically the right kind of response to the injustices that animate the call. The paper concludes by suggesting that AI should be democratized not by broadening and deepening participation but by increasing the democratic quality of the administrative and executive elements of collective decision making. In a slogan: The question is not so much whether AI should be democratized but how. (shrink)

After introducing the new field of cultural evolution, we review a growing body of empirical evidence suggesting that culture shapes what people attend to, perceive and remember as well as how they think, feel and reason. Focusing on perception, spatial navigation, mentalizing, thinking styles, reasoning (epistemic norms) and language, we discuss not only important variation in these domains, but emphasize that most researchers (including philosophers) and research participants are psychologically peculiar within a global and historical context. This rising tide of (...) evidence recommends caution in relying on one’s intuitions or even in generalizing from reliable psychological findings to the species, Homo sapiens. Our evolutionary approach suggests that humans have evolved a suite of reliably developing cognitive abilities that adapt our minds, information-processing abilities and emotions ontogenetically to the diverse culturally-constructed worlds we confront. (shrink)

During the last decade there has been burgeoning research concerning the ways in which we should think of and apply the concept of responsibility for Artificial Intelligence. Despite this conceptual richness, there is still a lack of consensus regarding what Responsible AI entails on both conceptual and practical levels. The aim of this paper is to connect the ethical dimension of responsibility in Responsible AI with Aristotelian virtue ethics, where notions of context and dianoetic virtues play a grounding role for (...) the concept of moral responsibility. The paper starts by highlighting the important difficulties in assigning responsibility to either technologies themselves or to their developers. Top-down and bottom-up approaches to moral responsibility are then contrasted, as we explore how they could inform debates about Responsible AI. We highlight the limits of the former ethical approaches and build the case for classical Aristotelian virtue ethics. We show that two building blocks of Aristotle’s ethics, dianoetic virtues and the context of actions, although largely ignored in the literature, can shed light on how we could think of moral responsibility for both AI and humans. We end by exploring the practical implications of this particular understanding of moral responsibility along the triadic dimensions of ethics by design, ethics in design and ethics for designers. (shrink)

We argue that explainable artificial intelligence (XAI), specifically reason-giving XAI, often constitutes the most suitable way of ensuring that someone can properly be held responsible for decisions that are based on the outputs of artificial intelligent (AI) systems. We first show that, to close moral responsibility gaps (Matthias 2004), often a human in the loop is needed who is directly responsible for particular AI-supported decisions. Second, we appeal to the epistemic condition on moral responsibility to argue that, in order to (...) be responsible for her decision, the human in the loop has to have an explanation available of the system’s recommendation. Reason explanations are especially well-suited to this end and we examine whether – and how – it might be possible to make such explanations fit with AI systems. We support our claims by focusing on a case of disagreement between human in the loop and AI system. (shrink)

Philosophical Investigations, Volume 45, Issue 1, Page 91-97, January 2022.

Automated vehicles (AVs) are expected to operate on public roads, together with non-automated vehicles and other road users such as pedestrians or bicycles. Recent ethical reports and guidelines raise worries that AVs will introduce injustice or reinforce existing social inequalities in road traffic. One major injustice concern in today’s traffic is that different types of road users are exposed differently to risks of corporal harm. In the first part of the paper, we discuss the responsibility of AV developers to address (...) existing injustice concerns regarding risk exposure as well as approaches on how to fulfill the responsibility for a fairer distribution of risk. In contrast to popular approaches on the ethics of risk distribution in unavoidable accident cases, we focus on low and moderate risk situations, referred to as routine driving. For routine driving, the obligation to distribute risks fairly must be discussed in the context of risk-taking and risk-acceptance, balancing safety objectives of occupants and other road users with driving utility. In the second part of the paper, we present a typical architecture for decentralized automated driving which contains a dedicated module for real-time risk estimation and management. We examine how risk estimation modules can be adjusted and parameterized to redress some inequalities. (shrink)

When life-saving medical resources are scarce and not everyone can be saved, is the only relevant goal saving the most lives? Or can other factors be considered, at least as tiebreakers, such as how early in life the people we don't save will die or how much future life they are likely to lose? This commentary defends a multiprinciple allocation approach that considers objectives in addition to saving more lives, including preventing early death and preventing harm in the form of (...) lost future life. Particularly compared to an arbitrary, coin-flip tiebreaker, this multiprinciple approach more effectively prevents harm, prioritizes the worst-off, mitigates socioeconomic and racial health disparities, and tracks public values regarding allocation under scarcity—and is legally sound. -/- . (shrink)

We motivate the desire for self-driving and explain its potential and limitations, and explore the need for—and potential implementation of—morals, ethics, and other value systems as complementary “capabilities” to the Deep Technologies behind self-driving. We consider how the incorporation of such systems may drive or slow adoption of high automation within vehicles. First, we explore the role for morals, ethics, and other value systems in self-driving through a representative hypothetical dilemma faced by a self-driving car. Through the lens of engineering, (...) we explain in simple terms common moral and ethical frameworks including utilitarianism, deontology, and virtue ethics before characterizing their relationship to the fundamental algorithms enabling self-driving. The concepts of behavior cloning, state-based modeling, and reinforcement learning are introduced, with some algorithms being more suitable for the implementation of value systems than others. We touch upon the contemporary cross-disciplinary landscape of morals and ethics in self-driving systems from a joint philosophical and technical perspective, and close with considerations for practitioners and the public, particularly as individuals may not appreciate the nuance and complexity of using imperfect information to navigate diverse scenarios and tough-to-quantify value systems, while “typical” software development reduces complex problems to black and white decision-making. (shrink)

In this paper, we argue that solutions to normative challenges associated with autonomous driving, such as real-world trolley cases or distributions of risk in mundane driving situations, face the problem of reasonable pluralism: Reasonable pluralism refers to the fact that there exists a plurality of reasonable yet incompatible comprehensive moral doctrines within liberal democracies. The corresponding problem is that a politically acceptable solution cannot refer to only one of these comprehensive doctrines. Yet a politically adequate solution to the normative challenges (...) of autonomous driving need not come at the expense of an ethical solution, if it is based on moral beliefs that are (1) shared in an overlapping consensus and (2) systematized through public reason. Therefore, we argue that a Rawlsian justificatory framework is able to adequately address the normative challenges of autonomous driving and elaborate on how such a framework might be employed for this purpose. (shrink)

Animal ethicists have been debating the morality of speciesism for over forty years. Despite rather persuasive arguments against this form of discrimination, many philosophers continue to assign humans a higher moral status than nonhuman animals. The primary source of evidence for this position is our intuition that humans’ interests matter more than the similar interests of other animals. And it must be acknowledged that this intuition is both powerful and widespread. But should we trust it for all that? The present (...) paper defends a negative answer to that question, based on a debunking argument. The intuitive belief that humans matter more than other animals is unjustified because it results from an epistemically defective process. It is largely shaped by tribalism, our tendency to favor ingroup members as opposed to outgroup members. And this influence is distortive for two reasons. First, tribalism evolved for reasons unrelated to moral truths; hence, it would at best produce true moral beliefs accidentally. Second, tribalism generates a vast quantity of false moral beliefs, starting with racist beliefs. Once this intuition is discarded, little evidence remains that speciesism is morally acceptable. (shrink)

Almost all participants in the debate about the ethics of accidents with self-driving cars have so far assumed moral universalism. However, universalism may be philosophically more controversial than is commonly thought, and may lead to undesirable results in terms of non-moral consequences and feasibility. There thus seems to be a need to also start considering what I refer to as the “relativistic car” — a car that is programmed under the assumption that what is morally right, wrong, good, bad, etc. (...) is determined by the moral beliefs of one’s society or culture. My investigation of this idea involves six steps. First, I explain why and how the moral universalism/relativism debate is relevant to the issue of self-driving cars. Second, I argue that there are good reasons to consider accident algorithms that assume relativism. Third, I outline how a relativistic car would be programmed to behave. Fourth, I address what advantages such a car would have, both in terms of its non-moral consequences and feasibility. Fifth, I address the relativistic car’s disadvantages. Finally, I qualify and conclude my considerations. (shrink)

In 2017, the German ethics commission for automated and connected driving released 20 ethical guidelines for autonomous vehicles. It is now up to the research and industrial sectors to enhance the development of autonomous vehicles based on such guidelines. In the current state of the art, we find studies on how ethical theories can be integrated. To the best of the authors’ knowledge, no framework for motion planning has yet been published which allows for the true implementation of any practical (...) ethical policies. This paper makes four contributions: Firstly, we briefly present the state of the art based on recent works concerning unavoidable accidents of autonomous vehicles (AVs) and identify further need for research. While most of the research focuses on decision strategies in moral dilemmas or crash optimization, we aim to develop an ethical trajectory planning for all situations on public roads. Secondly, we discuss several ethical theories and argue for the adoption of the theory “ethics of risk.” Thirdly, we propose a new framework for trajectory planning, with uncertainties and an assessment of risks. In this framework, we transform ethical specifications into mathematical equations and thus create the basis for the programming of an ethical trajectory. We present a risk cost function for trajectory planning that considers minimization of the overall risk, priority for the worst-off and equal treatment of people. Finally, we build a connection between the widely discussed trolley problem and our proposed framework. (shrink)

Experimental philosophy is a relatively recent discipline that employs experimental methods to investigate the intuitions, concepts, and assumptions behind traditional philosophical arguments, problems, and theories. While experimental philosophy initially served to interrogate the role that intuitions play in philosophy, it has since branched out to bring empirical methods to bear on problems within a variety of traditional areas of philosophy—including metaphysics, philosophy of language, philosophy of mind, and epistemology. To date, no connection has been made between developments in experimental philosophy (...) and philosophy of technology. In this paper, I develop and defend a research program for an experimental philosophy of technology. (shrink)

The fact that real-world decisions made by artificial intelligences (AI) are often ethically loaded has led a number of authorities to advocate the development of “moral machines”. I argue that the project of building “ethics” “into” machines presupposes a flawed understanding of the nature of ethics. Drawing on the work of the Australian philosopher, Raimond Gaita, I argue that ethical dilemmas are problems for particular people and not (just) problems for everyone who faces a similar situation. Moreover, the force of (...) an ethical claim depends in part on the life history of the person who is making it. For both these reasons, machines could at best be engineered to provide a shallow simulacrum of ethics, which would have limited utility in confronting the ethical and policy dilemmas associated with AI. (shrink)

The ethics of autonomous vehicles has received a great amount of attention in recent years, specifically in regard to their decisional policies in accident situations in which human harm is a likely consequence. Starting from the assumption that human harm is unavoidable, many authors have developed differing accounts of what morality requires in these situations. In this article, a strategy for AV decision-making is proposed, the Ethical Valence Theory, which paints AV decision-making as a type of claim mitigation: different road (...) users hold different moral claims on the vehicle’s behavior, and the vehicle must mitigate these claims as it makes decisions about its environment. Using the context of autonomous vehicles, the harm produced by an action and the uncertainties connected to it are quantified and accounted for through deliberation, resulting in an ethical implementation coherent with reality. The goal of this approach is not to define how moral theory requires vehicles to behave, but rather to provide a computational approach that is flexible enough to accommodate a number of ‘moral positions’ concerning what morality demands and what road users may expect, offering an evaluation tool for the social acceptability of an autonomous vehicle’s ethical decision making. (shrink)

This article addresses a dilemma about autonomous vehicles: how to respond to trade-off scenarios in which all possible responses involve the loss of life but there is a choice about whose life or lives are lost. I consider four options: kill fewer people, protect passengers, equal concern for survival, and recognize everyone’s interests. I solve this dilemma via what I call the new trolley problem, which seeks a rationale for the intuition that it is unethical to kill a smaller number (...) of people to avoid killing a greater number of people based on numbers alone. I argue that killing a smaller number of people to avoid killing a greater number of people based on numbers alone is unethical because it disrespects the humanity of the individuals in the smaller-numbered group. I defend the recognize-everyone’s-interests algorithm, which will probably kill fewer people but will not do so based on numbers alone. (shrink)

This paper looks at philosophical questions that arise in the context of AI alignment. It defends three propositions. First, normative and technical aspects of the AI alignment problem are interrelated, creating space for productive engagement between people working in both domains. Second, it is important to be clear about the goal of alignment. There are significant differences between AI that aligns with instructions, intentions, revealed preferences, ideal preferences, interests and values. A principle-based approach to AI alignment, which combines these elements (...) in a systematic way, has considerable advantages in this context. Third, the central challenge for theorists is not to identify ‘true’ moral principles for AI; rather, it is to identify fair principles for alignment that receive reflective endorsement despite widespread variation in people’s moral beliefs. The final part of the paper explores three ways in which fair principles for AI alignment could potentially be identified. (shrink)

Artificial intelligence plays an important role in current discussions on information and communication technologies and new modes of algorithmic governance. It is an unavoidable dimension of what social mediations and modes of reproduction of our information societies will be in the future. While several works in artificial intelligence ethics address ethical issues specific to certain areas of expertise, these ethical reflections often remain confined to narrow areas of application, without considering the global ethical issues in which they are embedded. We, (...) therefore, propose to clarify the main approaches to AIE, their philosophical assumptions and the specific characteristics of each one of them, to identify the most promising approach to develop an ethical reflection on the deployment of AI in our societies, which is the one based on information ethics as proposed by Luciano Floridi. We will identify the most important features of that approach to highlight areas that need further investigation. (shrink)

This paper takes stock of all the various factors that cause the design-time opacity of autonomous systems behaviour. The factors include embodiment effects, design-time knowledge gap, human factors, emergent behaviour and tacit knowledge. This situation is contrasted with the usual representation of moral dilemmas that assume perfect information. Since perfect information is not achievable, the traditional moral dilemma representations are not valid and the whole problem of ethical autonomous systems design proves to be way more empirical than previously understood.

This paper presents an experimental approach for engaging undergraduate STEM students in anticipatory ethical reasoning, or ethical reasoning applied to the analysis of potential mid- to long-term implications and outcomes of technological innovation. The authors implemented two variations of an approach that integrates three key components—scenario analysis, design fiction, and ethical frameworks—into five sections of an introductory course on the social contexts of science and technology that is required of STEM majors. The authors dub this approach Creative Anticipatory Ethical Reasoning, (...) or CAER. Scenario analysis is a strategy emerging from business consulting for grounded analysis of plausible future trajectories to inform planning. Design fiction is a creative hands-on activity that blends science fiction and design prototyping to facilitate critical thinking with respect to the societal dimensions of a plausible future technology. The authors present the following findings: in each of the variations, students demonstrated significant engagement with CAER and a substantive shift in their conception of what constitutes responsible innovation and ethical conduct in science and technology. Specifically, their integration of ethical reasoning with stakeholder perspectives and scenario analysis reframed technologies, from unproblematic solutions for societal problems to socially-embedded forms of life that might diverge from designers’ intentions. This suggests that CAER could be a useful pedagogical intervention for expanding students’ ethical engagement to consider the potential unintended consequences of technological innovation. (shrink)

Autonomous vehicles —and accidents they are involved in—attest to the urgent need to consider the ethics of artificial intelligence. The question dominating the discussion so far has been whether we want AVs to behave in a ‘selfish’ or utilitarian manner. Rather than considering modeling self-driving cars on a single moral system like utilitarianism, one possible way to approach programming for AI would be to reflect recent work in neuroethics. The agent–deed–consequence model :3–20, 2014a, Behav Brain Sci 37:487–488, 2014b) provides a (...) promising descriptive and normative account while also lending itself well to implementation in AI. The ADC model explains moral judgments by breaking them down into positive or negative intuitive evaluations of the agent, deed, and consequence in any given situation. These intuitive evaluations combine to produce a positive or negative judgment of moral acceptability. For example, the overall judgment of moral acceptability in a situation in which someone committed a deed that is judged as negative would be mitigated if the agent had good intentions and the action had a good consequence. This explains the considerable flexibility and stability of human moral judgment that has yet to be replicated in AI. This paper examines the advantages and disadvantages of implementing the ADC model and how the model could inform future work on ethics of AI in general. (shrink)