This paper shares results from a pedagogical experiment that assigns undergraduates to “cheat” on a final class essay by requiring their use of text-generating AI software. For this assignment, students harvested content from an installation of GPT-2, then wove that content into their final essay. At the end, students offered a “revealed” version of the essay as well as their own reflections on the experiment. In this assignment, students were specifically asked to confront the oncoming availability of AI as a (...) writing tool. What are the ethics of using AI this way? What counts as plagiarism? What are the conditions, if any, we should place on AI assistance for student writing? And how might working with AI change the way we think about writing, authenticity, and creativity? While students (and sometimes GPT-2) offered thoughtful reflections on these initial questions, actually composing with GPT-2 opened their perspectives more broadly on the ethics and practice of writing with AI. In this paper, I share how students experienced those issues, connect their insights to broader conversations in the humanities about writing and communication, and explain their relevance for the ethical use and evaluation of language models. (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)

There has been an increased focus within the AI ethics literature on questions of power, reflected in the ideal of accountability supported by many Responsible AI guidelines. While this recent debate points towards the power asymmetry between those who shape AI systems and those affected by them, the literature lacks normative grounding and misses conceptual clarity on how these power dynamics take shape. In this paper, I develop a workable conceptualization of said power dynamics according to Cristiano Castelfranchi’s conceptual framework (...) of power and argue that end-users depend on a system’s developers and users, because end-users rely on these systems to satisfy their goals, constituting a power asymmetry between developers, users and end-users. I ground my analysis in the neo-republican moral wrong of domination, drawing attention to legitimacy concerns of the power-dependence relation following from the current lack of accountability mechanisms. I illustrate my claims on the basis of a risk-prediction machine learning system, and propose institutional and project-specific solutions to mitigate domination. (shrink)

This book explicitly addresses policy options in a democratic society regarding cognitive enhancement drugs and devices. The book offers an in-depth case by case analysis of existing and emerging cognitive neuroenhancement technologies and canvasses a distinct political neuroethics approach. The author provides an argument on the much debated issue of fairness of cognitive enhancement practices and tackles the tricky issue of how to respect preferences of citizens opposing and those preferring enhancement. The author persuasively argues the necessity of a laws (...) and regulations regarding the use of cognitive enhancers. He also argues that the funds for those who seek cognitive enhancement should be allocated free of charge to the least advantaged. The work argues that the notion of autonomy has been mistakenly associated with the metaphysical concept of free will, and offers a political definition of autonomy to clarify how responsibility is implicitly grounded in the legal and political system. As such, this book is an essential read for everyone interested in neuroethics, and a valuable resource for policy makers, as well as scholars and students in philosophy, law, psychiatry and neuroscience. (shrink)

This paper explores the role and resolution of disagreements between physicians and their diagnostic AI-based decision support systems. With an ever-growing number of applications for these independently operating diagnostic tools, it becomes less and less clear what a physician ought to do in case their diagnosis is in faultless conflict with the results of the DSS. The consequences of such uncertainty can ultimately lead to effects detrimental to the intended purpose of such machines, e.g. by shifting the burden of proof (...) towards a physician. Thus, we require normative clarity for integrating these machines without affecting established, trusted, and relied upon workflows. In reconstructing different causes of conflicts between physicians and their AI-based tools—inspired by the approach of “meaningful human control” over autonomous systems and the challenges to resolve them—we will delineate normative conditions for “meaningful disagreements”. These incorporate the potential of DSS to take on more tasks and outline how the moral responsibility of a physician can be preserved in an increasingly automated clinical work environment. (shrink)

Virtue ethics has many times been suggested as a promising recipe for the construction of artificial moral agents due to its emphasis on moral character and learning. However, given the complex nature of the theory, hardly any work has de facto attempted to implement the core tenets of virtue ethics in moral machines. The main goal of this paper is to demonstrate how virtue ethics can be taken all the way from theory to machine implementation. To achieve this goal, we (...) critically explore the possibilities and challenges for virtue ethics from a computational perspective. Drawing on previous conceptual and technical work, we outline a version of artificial virtue based on moral functionalism, connectionist bottom–up learning, and eudaimonic reward. We then describe how core features of the outlined theory can be interpreted in terms of functionality, which in turn informs the design of components necessary for virtuous cognition. Finally, we present a comprehensive framework for the technical development of artificial virtuous agents and discuss how they can be implemented in moral environments. (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)

Given the complexity of teams involved in creating AI-based systems, how can we understand who should be held accountable when they fail? This paper reports findings about accountable AI from 26 interviews conducted with stakeholders in AI drawn from the fields of AI research, law, and policy. Participants described the challenges presented by the distributed nature of how AI systems are designed, developed, deployed, and regulated. This distribution of agency, alongside existing mechanisms of accountability, responsibility, and liability, creates barriers for (...) effective accountable design. As agency is distributed across the socio-technical landscape of an AI system, users without deep knowledge of the operation of these systems become disempowered, unable to challenge or contest when it impacts their lives. In this context, accountability becomes a matter of building systems that can be challenged, interrogated, and, most importantly, adjusted in use to accommodate counter-intuitive results and unpredictable impacts. Thus, accountable system design can work to reconfigure socio-technical landscapes to protect the users of AI and to prevent unjust apportionment of risk. (shrink)

In this paper, we propose a methodology to maximize the benefits of interdisciplinary cooperation in AI research groups. Firstly, we build the case for the importance of interdisciplinarity in research groups as the best means to tackle the social implications brought about by AI systems, against the backdrop of the EU Commission proposal for an Artificial Intelligence Act. As we are an interdisciplinary group, we address the multi-faceted implications of the mass-scale diffusion of AI-driven technologies. The result of our exercise (...) lead us to postulate the necessity of a behavioural theory that standardizes the interaction process of interdisciplinary groups. In light of this, we conduct a review of the existing approaches to interdisciplinary research on AI appliances, leading to the development of methodologies like ethics-by-design and value-sensitive design, evaluating their strengths and weaknesses. We then put forth an iterative process theory hinging on a narrative approach consisting of four phases: definition of the hypothesis space, building-up of a common lexicon, scenario-building, interdisciplinary self-assessment. Finally, we identify the most relevant fields of application for such a methodology and discuss possible case studies. (shrink)

In this paper, I shall set out the pros and cons of assigning legal personhood on artificial intelligence systems under civil law. More specifically, I will provide arguments supporting a functionalist justification for conferring personhood on AIs, and I will try to identify what content this legal status might have from a regulatory perspective. Being a person in law implies the entitlement to one or more legal positions. I will mainly focus on liability as it is one of the main (...) grounds for the attribution of legal personhood, like for collective legal entities. A better distribution of responsibilities resulting from unpredictably illegal and/or harmful behaviour may be one of the main reasons to justify the attribution of personhood also for AI systems. This means an efficient allocation of the risks and social costs associated with the use of AIs, ensuring the protection of victims, incentives for production, and technological innovation. However, the paper also considers other legal positions triggered by personhood in addition to responsibility: specific competencies and powers such as, for example, financial autonomy, the ability to hold property, make contracts, sue. (shrink)

Ensuring the trustworthiness of autonomous systems and artificial intelligence is an important interdisciplinary endeavour. In this position paper, we argue that this endeavour will benefit from technical advancements in capturing various forms of responsibility, and we present a comprehensive research agenda to achieve this. In particular, we argue that ensuring the reliability of autonomous system can take advantage of technical approaches for quantifying degrees of responsibility and for coordinating tasks based on that. Moreover, we deem that, in certifying the legality (...) of an AI system, formal and computationally implementable notions of _responsibility_, _blame_, _accountability_, and _liability_ are applicable for addressing potential responsibility gaps (i.e. situations in which a group is responsible, but individuals’ responsibility may be unclear). This is a call to enable AI systems themselves, as well as those involved in the design, monitoring, and governance of AI systems, to represent and reason about who can be seen as responsible in prospect (e.g. for completing a task in future) and who can be seen as responsible retrospectively (e.g. for a failure that has already occurred). To that end, in this work, we show that across all stages of the design, development, and deployment of trustworthy autonomous systems (TAS), responsibility reasoning should play a key role. This position paper is the first step towards establishing a road map and research agenda on how the notion of responsibility can provide novel solution concepts for ensuring the _reliability_ and _legality_ of TAS and, as a result, enables an effective embedding of AI technologies into society. (shrink)

The European Union Commission’s whitepaper on Artificial Intelligence proposes shaping the emerging AI market so that it better reflects common European values. It is a master plan that builds upon the EU AI High-Level Expert Group guidelines. This article reviews the masterplan, from a culture cycle perspective, to reflect on its potential clashes with current societal, technical, and methodological constraints. We identify two main obstacles in the implementation of this plan: the lack of a coherent EU vision to drive future (...) decision-making processes at state and local levels and the lack of methods to support a sustainable diffusion of AI in our society. The lack of a coherent vision stems from not considering societal differences across the EU member states. We suggest that these differences may lead to a fractured market and an AI crisis in which different members of the EU will adopt nation-centric strategies to exploit AI, thus preventing the development of a frictionless market as envisaged by the EU. Moreover, the Commission aims at changing the AI development culture proposing a human-centred and safety-first perspective that is not supported by methodological advancements, thus taking the risks of unforeseen social and societal impacts of AI. We discuss potential societal, technical, and methodological gaps that should be filled to avoid the risks of developing AI systems at the expense of society. Our analysis results in the recommendation that the EU regulators and policymakers consider how to complement the EC programme with rules and compensatory mechanisms to avoid market fragmentation due to local and global ambitions. Moreover, regulators should go beyond the human-centred approach establishing a research agenda seeking answers to the technical and methodological open questions regarding the development and assessment of human-AI co-action aiming for a sustainable AI diffusion in the society. (shrink)

In their paper “Algorithms for Ethical Decision-Making in the Clinic: A Proof of Concept,” Meier et al. present the design and preliminary results of a proof-of-concept clinical ethics algor...

The three main kinds of theory in normative ethics, namely, consequentialism, deontology, and virtue ethics, are often presented as the ‘palette’ from which we may choose, or use as a starting point for an investigation. However, this way of doing ethics and philosophy, by the palette, may be leading some of us astray. It has led some to believe that all that there is to ethics, and to ethics of AI, is given in terms of these already devised petrified categories (...) of theory. It has also led others to abandon normative ethics and philosophy altogether and to resort to descriptive methods that are then used to justify action. I wish to argue that (1) we should not abandon traditional philosophical approaches, but (2a) this does not entail that the petrified palette should constitute the beginning of our philosophical investigations. Further, (2b) I recommend a non-methodological approach in which it is instead radical questions that spur these investigations, which arise through consideration of the practical actions (potential or otherwise) of machines and their programmers. (shrink)

This article discusses the topic of algorithmic personalization and the creation of the so-called “filter bubble” effect, which is often understood as one of the most problematic influences of artificial intelligence on democratic social order. The author suggests that focusing on the issue of information diversity, which had far-reaching effect on the empirical research that tried to quantitatively measure and systematically prove the existence of the filter bubbles, was the wrong starting point for the discussion on the application of algorithmic (...) personalization. It has drawn our attention away from the deeper issue: habit. Habitual adaptation of algorithmic personalization, namely, stands in essential contradiction to the ideal image of a non-adaptive public sphere, upon which democratic societies should be based. Focusing on habitual adaptation could also explain why users addictively stick to certain kinds of information, even if they are not caught in an isolated chamber. (shrink)

AI technology is capturing the African imaginations as a gateway to progress and prosperity. There is a growing interest in AI by different actors across the continent including scientists, researchers, humanitarian and aid organizations, academic institutions, tech start-ups, and media organizations. Several African states are looking to adopt AI technology to capture economic growth and development opportunities. On the other hand, African researchers highlight the gap in regulatory frameworks and policies that govern the development of AI in the continent. They (...) argue that this could lead to AI technology exacerbating problems of inequalities and injustice in the continent. However, most of the literature on AI ethics is biased toward Euro-American perspectives and lack the understanding of how AI development is apprehended in the Global South, and particularly Africa. Drawing on the case study of the first African Master’s in Machine Intelligence program, this paper argues for looking beyond the question of ethics in AI and examining AI governance issues through the analytical lens of the raciality of computing and the political economy of technoscience to understand AI development in Africa. By doing so, this paper seeks a different theorization for AI ethics from the South that is based on lived experiences of those in the margins and avoids the framings of technological futures that simplistically pathologize or celebrate Africa. (shrink)

In popular discussions, the nuances of AI are often abridged as “the algorithm”, as the specific arrangements of machine learning, deep learning and automated decision-making on social media platforms are typically shrouded in proprietary secrecy punctuated by press releases and transparency initiatives. What is clear, however, is that AI embedded on social media functions to recommend content, personalize ads, aggregate news stories, and moderate problematic material. It is also increasingly apparent that individuals are concerned with the uses, implications, and fairness (...) of algorithmic systems. Perhaps in response to concerns about “the algorithm” by individuals and governments, social media platforms utilize transparency initiatives and official statements, in part, to deflect official regulation. In the following paper, I draw from transparency initiatives and statements from representatives of Facebook and TikTok as case studies of how AI is embedded in these platforms, with attention to the promotion of AI content moderation as a solution to the circulation of problematic material and misinformation. This examination considers the complexity of embedded AI as a material-discursive apparatus, predicated on discursive techniques—what is seeable, sayable, knowable in a given time period—as well as the material arrangements—algorithms, datasets, users, platforms, infrastructures, moderators, etc. As such, the use of AI as part of the immensely popular platforms Facebook and TikTok demonstrates that AI does not exist in isolation, instead functioning as human–machine ensemble reliant on strategies of acceptance via discursive techniques and the changing material arrangements of everyday embeddedness. (shrink)

It is a truism that, due to human weaknesses, human soldiers have yet to have sufficiently ethical warfare. It is arguable that the likelihood of human soldiers to breach the Principle of Non-Combatant Immunity, for example, is higher in contrast tosmart soldierswho are emotionally inept. Hence, this paper examines the possibility that the integration of ethics into smart soldiers will help address moral challenges in modern warfare. The approach is to develop and employ smart soldiers that are enhanced with ethical (...) capabilities. Advocates of this approach think that it is more realistic to make competent entities (i.e., smart soldiers) become morally responsible than to enforce moral responsibility on human soldiers with inherent (moral) limitations. This view somewhat seeks a radical transition from the usual anthropocentric warfare to arobocentricwarfare with the belief that the transition has moral advantages. However, the paper defends the claim that despite human limitations, the capacity ofethically enhanced smart soldiersfor moral sensitivity is artificial and unauthentic. There are significant problems with the three models of programming ethics into smart soldiers. Also, there are further challenges from the absence of emotion as a moral gauge, and the problems of apportioning responsibility in case of mishap from the actions or omissions of smart soldiers. Among other reasons, the paper takes the replacement of human soldiers as an extreme approach towards an ethical warfare. This replacement predicates ethical complications that outweigh the benefits from the exclusive use of smart soldiers. (shrink)

This project uses machine learning and computer vision techniques and a novel interactive visualization tool to provide street-level characterization of urban spaces such as safety and maintenance in urban neighborhoods. This is achieved by collecting and annotating street-view images, extracting objective metrics through computer vision techniques, and using crowdsourcing to statistically model the perception of subjective metrics such as safety and maintenance. For modeling human perception and scaling it up with a predictive algorithm, we evaluate perception predictions across two points (...) in time separated by economic changes in the urban core of Raleigh, North Carolina, in the aftermath of the 2008 Great Recession. We hypothesize specific socioeconomic processes can be substantially reflected in the built environment of cities and, thus, render themselves visible at the street level. This paper describes the process of incorporating subjective visual ratings across two datasets of temporally separated street-view images, an algorithm, and a visualization tool. This work serves as a case study for utilizing AI and visualization techniques in a richer characterization of urban spaces that includes both objective metrics such as income (that operates at a broader scale) and subjective metrics such as perception of individuals (that operates at a narrower scale at specific locations). We outline an interdisciplinary methodology to test this hypothesis in streetscape data from Raleigh, NC, from 2008 to 2020. We describe the results of training algorithms that utilized image features with crowdsourced human perception ratings. We provide a comparison of the results with income data. The analysis and interpretation of this comparison provide insight into the challenges and opportunities for using AI technology in characterizing changes in urban environments. One challenge is the ability of human domain experts to interpret the output of algorithms through manipulation and to integrate these results into their workflow. This is addressed with a novel interface designed for interactive analysis and visualization. We conclude with a discussion of some of the benefits and limitations of integrating AI models in the human expert’s decision-making process in the presence of both subjective and objective metrics. (shrink)

It is widely recognized that the Web contributes to user polarization, and such polarization affects not just politics but also peoples’ stances about public health, such as vaccination. Understanding polarization in social networks is challenging because it depends not only on user attitudes but also their interactions and exposure to information. We adopt Social Judgment Theory to operationalize attitude shift and model user behavior based on empirical evidence from past studies. We design a social simulation to analyze how content sharing (...) affects user satisfaction and polarization in a social network. We investigate the influence of varying tolerance in users and selectively exposing users to congenial views. We find that (1) higher user tolerance slows down polarization and leads to lower user satisfaction; (2) higher selective exposure leads to higher polarization and lower user reach; and (3) both higher tolerance and higher selective exposure lead to a more homophilic social network. (shrink)

Consequential historical decisions that shaped transportation systems and their influence on society have many valuable lessons. The decisions we learn from and choose to make going forward will play a key role in shaping the mobility landscape of the future. This is especially pertinent as artificial intelligence (AI) becomes more prevalent in the form of autonomous vehicles (AVs). Throughout urban history, there have been cyclical transport oppressions of previous-generation transportation methods to make way for novel transport methods. These cyclical oppressions (...) can be identified in the baroque and modernist periods, and a third oppression may occur in the contemporary period. To explore the idea of a third oppression, we focus on the bicycle and outline the history of cycling to understand how historical mode oppression unfolded. We then present several social and political factors that contributed to the oppression of cycling and share recommendations for how to avoid future oppressions including political, social, and design actions for researchers and policymakers to take. This paper argues that priorities for AI-enabled mobility and cyclist needs be advanced in proportion to the extent that they contribute to societal goals of urban containment, public realm, and proximal cities. Additionally, future mobility evolutions should prioritise mobility justice and mode choice over inducing a singular transportation method. (shrink)

In this paper, we outline a new method for evaluating the human impact of machine-learning applications. In partnership with Underwriters Laboratories Inc., we have developed a framework to evaluate the impacts of a particular use of machine learning that is based on the goals and values of the domain in which that application is deployed. By examining the use of artificial intelligence in particular domains, such as journalism, criminal justice, or law, we can develop more nuanced and practically relevant understandings (...) of key ethical guidelines for artificial intelligence. By decoupling the extraction of the facts of the matter from the evaluation of the impact of the resulting systems, we create a framework for the process of assessing impact that has two distinctly different phases. (shrink)