Narrative responsibility and artificial intelligence

Most accounts of responsibility focus on one type of responsibility, moral responsibility, or address one particular aspect of moral responsibility such as agency. This article outlines a broader framework to think about responsibility that includes causal responsibility, relational responsibility, and what I call “narrative responsibility” as a form of “hermeneutic responsibility”, connects these notions of responsibility with different kinds of knowledge, disciplines, and perspectives on human being, and shows how this framework is helpful for mapping and analysing how artificial intelligence (AI) challenges human responsibility and sense-making in various ways. Mobilizing recent hermeneutic approaches to technology, the article argues that next to, and interwoven with, other types of responsibility such as moral responsibility, we also have narrative and hermeneutic responsibility—in general and for technology. For example, it is our task as humans to make sense of, with and, if necessary, against AI. While from a posthumanist point of view, technologies also contribute to sense-making, humans are the experiencers and bearers of responsibility and always remain in charge when it comes to this hermeneutic responsibility. Facing and working with a world of data, correlations, and probabilities, we are nevertheless condemned to make sense. Moreover, this also has a normative, sometimes even political aspect: acknowledging and embracing our hermeneutic responsibility is important if we want to avoid that our stories are written elsewhere—through technology.

Harm and Helmets: How Harmful Social Costs Fail to Justify Paternalistic Health and Safety Regulations

Mill’s harm principle and the financial externalities of risky behavior are routinely invoked to justify health and safety regulation. However, this approach fares poorly when subjected to theoretical scrutiny. First, it is false: individuals engaging in risky behavior do not harm others. Second, even if risky behavior were harmful to others, the argument from harmful externalities does not imply safety-enhancing policy interventions, at least not without additional appeals to paternalism. Third, focusing on the economic impacts of accidents invites perverse victim-blaming attitudes toward accident victims that undermine democratic values and justice. To improve our moral understanding of health and safety regulation, I sketch a theory of public policy justification grounded in the controversies which attract our attention to paternalistic polices in the first place. On this account, justificatory arguments are plausible if they identify goods that individuals genuinely affirm on their own terms, are sensitive to causal responsibility and imbalances between restraint and protection, and comparatively engage with possible policy alternatives. Illustrating the shortcomings of one dominant approach to public policy justification and reorienting us toward the controversies that policy justifications need to confront reflect two ways that political theory can help enhance justice in public policy design and articulation.

Contribution of causal factors to disease burden: how to interpret attributable fractions

What proportion of the risk in a given population is attributable to a risk factor? The population attributable fraction (PAF) answers this question. “Attributable to” is understood as “due to”, which makes PAFs closely related to the concept of potential impact or potential benefits of reducing the exposure. The PAF is a tool at the border between science and decision making. PAFs are estimated based on strong assumptions and the calculations are data intensive, making them vulnerable to gaps in knowledge and data. Current misconceptions include summing up PAFs to 100% or subtracting a PAF for a factor from 100% to deduce what proportion is left to be explained or prevented by other factors. This error is related to unrecognised multicausality or shared causal responsibility in disease aetiology. Attributable cases only capture cases in excess and should be regarded as a lower bound for aetiological cases, which cannot be estimated based on epidemiological data alone (exposure-induced cases). The population level might not be relevant to discuss prevention priorities based on PAFs, for instance when exposures concentrate in a subgroup of the population, as for occupational lung carcinogens and other workplace hazards. Alternative approaches have been proposed based on absolute rather than relative metrics, such as estimating potential gains in life expectancy that can be expected from a specific policy (prevention) or years of life lost due to a specific exposure that already happened (compensation).

NOVUS ACTUS AND BEYOND: ATTRIBUTING CAUSAL RESPONSIBILITY IN THE CRIMINAL COURTS

Glanville Williams's influential 1989 article on causation, “Finis for Novus Actus?”, addressed two pertinent questions: (1) when, and on what grounds, may a person be judged to bear causal responsibility for harms most immediately brought about by the subsequent action of another person (the locus classicus of the novus actus interveniens doctrine), and (2) how should questions of causation be resolved in cases where the potential cause in question constitutes an omission? This article revisits these questions through an engagement with some of the major causation cases decided in the criminal courts in the past decade. The discussion of these cases is set in the broader context of a critique of H.L.A. Hart and Tony Honoré's influential doctrinal-theoretical framework for findings of legal causation, the autonomy doctrine, on which Williams had built his arguments.

Framing of Armed Conflicts between the Two Nuclear States of South Asia

The current study investigates the framing process through the lens of the causal responsibility and the subject matter adopted by Pakistani print news media during the armed conflicts after the Pulwama assault in February 2019. With the help of the census approach, 282 opinions and editorials were collected from a population of 1,321 published items from six English newspapers. The findings show that Pakistani print media outlets extensively used the individual causal responsibility frame, while the social responsibility frame was used less in numbers. The content analysis reveals that the subject matter of awareness was extensively used in media content during the Pulwama assault and its aftermath. The observations are examined from the perspective of journalistic preferences in selecting specific frames during the framing of regional and global armed conflicts.

When company decisions harm or help the environment: the influence of social context and affective state on moral and causal responsibility attributions

Holding companies accountable for their decisions’ environmental side effects becomes increasingly important in the light of current debates on the climate crisis. The present study investigated a young sample’s (n = 925) causal and moral responsibility attributions to a company’s manager and their subordinate employee, who were either young or middle-aged. The agents jointly made a profit-oriented decision which either harmed or helped the environment. Results were analyzed with respect to the valence of the side effect, the agents’ social role, the agents’ age group, and participants’ affective state. We successfully replicated findings from prior studies, showing that more intention and moral responsibility is ascribed to the manager in case of a negative side effect than in case of a positive side effect, and that the manager deserves more blame for negative side effects than the employee, whereas the employee deserves more praise for positive side effects. Additionally, responsibility attributions varied with the agents’ age group, participants’ positive and negative affective state, and participants’ gender. By incorporating these hitherto neglected factors, the study complements existing theories of responsibility attribution and intergroup perception. The findings further highlight the contribution of affective mechanisms as driving factors of responsibility attributions. Combining moral responsibility research, intergroup research, and affective measures promises to foster the understanding of how and to whom people ascribe blame, praise, and causal responsibility for environmental consequences.

Moral Responsibility and Moral Development in Epicurus’ Philosophy

This chapter shows that Epicurus had a notion of moral responsibility based on the agents’ causal responsibility—as opposed to their ability to act or choose otherwise. Thus, the central question for responsibility was whether the agent was the cause of the action, or was forced to act by something else. Actions could be attributed to agents because it is in their actions that their character manifests itself. As a result, moral development becomes all-important. Thus the chapter discusses evidence for Epicurus’ views on how humans become moral beings and on how they morally improve. Epicurus envisaged a complex web of hereditary and environmental factors as shaping the moral aspects of humans. It results that Epicurean ethics does not have the function of developing or justifying a moral system that allows for the effective allocation of blame. Rather, its function is to give everyone a chance to morally improve.

Theoretical, Measured, and Subjective Responsibility in Aided Decision Making

When humans interact with intelligent systems, their causal responsibility for outcomes becomes equivocal. We analyze the descriptive abilities of a newly developed responsibility quantification model (ResQu) to predict actual human responsibility and perceptions of responsibility in the interaction with intelligent systems. In two laboratory experiments, participants performed a classification task. They were aided by classification systems with different capabilities. We compared the predicted theoretical responsibility values to the actual measured responsibility participants took on and to their subjective rankings of responsibility. The model predictions were strongly correlated with both measured and subjective responsibility. Participants’ behavior with each system was influenced by the system and human capabilities, but also by the subjective perceptions of these capabilities and the perception of the participant's own contribution. A bias existed only when participants with poor classification capabilities relied less than optimally on a system that had superior classification capabilities and assumed higher-than-optimal responsibility. The study implies that when humans interact with advanced intelligent systems, with capabilities that greatly exceed their own, their comparative causal responsibility will be small, even if formally the human is assigned major roles. Simply putting a human into the loop does not ensure that the human will meaningfully contribute to the outcomes. The results demonstrate the descriptive value of the ResQu model to predict behavior and perceptions of responsibility by considering the characteristics of the human, the intelligent system, the environment, and some systematic behavioral biases. The ResQu model is a new quantitative method that can be used in system design and can guide policy and legal decisions regarding human responsibility in events involving intelligent systems.

Moral Reasoning about Aggressive Behavior in Relation to Type of Aggression, Age and Gender in South Korean Pupils

Studies of moral reasoning in relation to aggressive behaviors have paid limited attention to different types of aggression, and have mainly been conducted in Western societies. We describe findings from a study of 157 children, aged 6 or 11 years, from two schools in South Korea. Using a cartoon scenario methodology, we assessed moral reasoning about eight types of aggression: verbal, physical individual, physical group, social exclusion, rumor spreading, breaking one’s belongings, sending a nasty text via mobile phone, and sending a nasty message/email via computer. Four aspects of moral reasoning were assessed: moral judgment, harmfulness, reason for judgment, and causal responsibility. Many significant differences by type of aggression were found, especially for social exclusion (seen as less wrong and harmful, and more the victim’s responsibility), physical group aggression (seen as more wrong or harmful, and a matter of fairness, especially in older children and boys), and cyber aggression (seen more as the aggressor’s responsibility). Older children gave more reasons based on welfare, and fewer “don’t know” responses for reasons and attributions. Gender differences were relatively few, but girls did make more use of welfare in the moral reasoning domain. Findings are discussed in relation to previous research and the cultural context in South Korea.

Moral responsibility for computationally designed products

Computational design systems (such as those using evolutionary algorithms) can create designs for a variety of physical products. Introducing these systems into the design process risks creating a ‘responsibility gap’ for flaws in the products they are used to create, as human designers may no longer believe that they are wholly responsible for them. We respond to this problem by distinguishing between causal responsibility and capacity responsibility (the ability to be morally responsible for actions) for creating product designs to argue that while the computational design systems and human designers are both casually responsible for creating product designs, the human designers who use these systems and the developers who create them have capacity responsibility for such designs. We show that there is no responsibility gap for products designed using computational design systems by comparing different accounts of moral responsibility for robots and AI (instrumentalism, machine ethics, and hybrid responsibility). We argue that all three of these accounts of moral responsibility for AI systems support the conclusion that the product designers who use computational design systems and the developers of these systems are morally responsible for any flaws or faults in the products designed by these systems. We conclude by showing how the responsibilities of accountability and blameworthiness should be attributed between the product designers, the developers of the computational design systems.