Ethics of Artificial Intelligence, Misc

Does AI conform to humans, or will we conform to AI? An ethical evaluation of AI-intensive companies will allow investors to knowledgeably participate in the decision. The evaluation is built from nine performance indicators that can be analyzed and scored to reflect a technology’s human-centering. When summed, the scores convert into objective investment guidance. The strategy of incorporating ethics into financial decisions will be recognizable to participants in environmental, social, and governance investing, however, this paper argues that conventional ESG frameworks (...) are inadequate for AI-intensive companies. To fully account for contemporary technology, the following categories of evaluation will be developed and featured as vital investing criteria: autonomy, dignity, privacy, performance. With these priorities established, the larger goal is a model for humanitarian investing in AI-intensive companies that is intellectually robust, manageable for analysts, useful for portfolio managers, and credible for investors. (shrink)

Foundation models such as GPT-4 are fine-tuned to avoid unsafe or otherwise problematic behavior, so that, for example, they refuse to comply with requests for help with committing crimes or with producing racist text. One approach to fine-tuning, called reinforcement learning from human feedback, learns from humans' expressed preferences over multiple outputs. Another approach is constitutional AI, in which the input from humans is a list of high-level principles. But how do we deal with potentially diverging input from humans? How (...) can we aggregate the input into consistent data about "collective" preferences or otherwise use it to make collective choices about model behavior? In this paper, we argue that the field of social choice is well positioned to address these questions, and we discuss ways forward for this agenda, drawing on discussions in a recent workshop on Social Choice for AI Ethics and Safety held in Berkeley, CA, USA in December 2023. (shrink)

Danaher (2016) has argued that increasing robotization can lead to retribution gaps: Situation in which the normative fact that nobody can be justly held responsible for a harmful outcome stands in conflict with our retributivist moral dispositions. In this paper, we report a cross-cultural empirical study based on Sparrow’s (2007) famous example of an autonomous weapon system committing a war crime, which was conducted with participants from the US, Japan and Germany. We find that (i) people manifest a considerable willingness (...) to hold autonomous systems morally responsible, (ii) partially exculpate human agents when interacting with such systems, and that more generally (iii) the possibility of normative responsibility gaps is indeed at odds with people’s pronounced retributivist inclinations. We discuss what these results mean for potential implications of the retribution gap and other positions in the responsibility gap literature. (shrink)

Two goals - improving replicability and accountability of Machine Learning research respectively, have accrued much attention from the AI ethics and the Machine Learning community. Despite sharing the measures of improving transparency, the two goals are discussed in different registers - replicability registers with scientific reasoning whereas accountability registers with ethical reasoning. Given the existing challenge of the Responsibility Gap - holding Machine Learning scientists accountable for Machine Learning harms due to them being far from sites of application, this paper (...) posits that reconceptualizing replicability can help bridge the gap. Through a shift from model performance replicability to claim replicability, Machine Learning scientists can be held accountable for producing non-replicable claims that are prone to eliciting harm due to misuse and misinterpretation. In this paper, I make the following contributions. First, I define and distinguish two forms of replicability for ML research that can aid constructive conversations around replicability. Second, I formulate an argument for claim-replicability's advantage over model performance replicability in justifying assigning accountability to Machine Learning scientists for producing non-replicable claims and show how it enacts a sense of responsibility that is actionable. In addition, I characterize the implementation of claim replicability as more of a social project than a technical one by discussing its competing epistemological principles, practical implications on Circulating Reference, Interpretative Labor, and research communication. (shrink)

The prevailing discourse around AI ethics lacks the language and formalism necessary to capture the diverse ethical concerns that emerge when AI systems interact with individuals. Drawing on Sen and Nussbaum's capability approach, we present a framework formalizing a network of ethical concepts and entitlements necessary for AI systems to confer meaningful benefit or assistance to stakeholders. Such systems enhance stakeholders' ability to advance their life plans and well-being while upholding their fundamental rights. We characterize two necessary conditions for morally (...) permissible interactions between AI systems and those impacted by their functioning, and two sufficient conditions for realizing the ideal of meaningful benefit. We then contrast this ideal with several salient failure modes, namely, forms of social interactions that constitute unjustified paternalism, coercion, deception, exploitation and domination. The proliferation of incidents involving AI in high-stakes domains underscores the gravity of these issues and the imperative to take an ethics-led approach to AI systems from their inception. (shrink)

Healthcare systems across the globe are struggling with increasing costs and worsening outcomes. This presents those responsible for overseeing healthcare with a challenge. Increasingly, policymakers, politicians, clinical entrepreneurs and computer and data scientists argue that a key part of the solution will be ‘Artificial Intelligence’ (AI) – particularly Machine Learning (ML). This argument stems not from the belief that all healthcare needs will soon be taken care of by “robot doctors.” Instead, it is an argument that rests on the classic (...) counterfactual definition of AI as an umbrella term for a range of techniques that can be used to make machines complete tasks in a way that would be considered intelligent were they to be completed by a human. Automation of this nature could offer great opportunities for the improvement of healthcare services and ultimately patients’ health by significantly improving human clinical capabilities in diagnosis, drug discovery, epidemiology, personalised medicine, and operational efficiency. However, if these AI solutions are to be embedded in clinical practice, then at least three issues need to be considered: the technical possibilities and limitations; the ethical, regulatory and legal framework; and the governance framework. In this article, we report on the results of a systematic analysis designed to provide a clear overview of the second of these elements: the ethical, regulatory and legal framework. We find that ethical issues arise at six levels of abstraction (individual, interpersonal, group, institutional, sectoral, and societal) and can be categorised as epistemic, normative, or overarching. We conclude by stressing how important it is that the ethical challenges raised by implementing AI in healthcare settings are tackled proactively rather than reactively and map the key considerations for policymakers to each of the ethical concerns highlighted. (shrink)

The advent of Generative AI, particularly through Large Language Models (LLMs) like ChatGPT and its successors, marks a paradigm shift in the AI landscape. Advanced LLMs exhibit multimodality, handling diverse data formats, thereby broadening their application scope. However, the complexity and emergent autonomy of these models introduce challenges in predictability and legal compliance. This paper analyses the legal and regulatory implications of Generative AI and LLMs in the European Union context, focusing on liability, privacy, intellectual property, and cybersecurity. It examines (...) the adequacy of the existing and proposed EU legislation, including the Artificial Intelligence Act (AIA), in addressing the challenges posed by Generative AI in general and LLMs in particular. The paper identifies potential gaps and shortcomings in the EU legislative framework and proposes recommendations to ensure the safe and compliant deployment of generative models. (shrink)

الذكاء الاصطناعي العاطفي»، ويُعرف أيضًا باسم «الحوسبة العاطفية»، و«الذكاء الاصطناعي المتمركز حول الإنسان»، و«الذكاء الاصطناعي الاجتماعي»، مفهوم جديد نسبيًا (ما زالت تقنياته في طور التطوير)، وهو أحد مجالات علوم الحاسوب الهادفة إلى تطوير آلات قادرة على فهم المشاعر البشرية. يشير المفهوم ببساطة إلى اكتشاف وبرمجة المشاعر الإنسانية بُغية تحسين الذكاء الاصطناعي، وتوسيع نطاق استخدامه، بحيث لا يقتصر أداء الروبوتات على تحليل الجوانب المعرفية (المنطقية) والتفاعل معها فحسب، بل والامتداد بالتحليل والتفاعل إلى الجوانب العاطفية للتواصل البشري.

تُعد أخلاقيات الآلة جزءًا من أخلاقيات الذكاء الاصطناعي المعنية بإضافة أو ضمان السلوكيات الأخلاقية للآلات التي صنعها الإنسان، والتي تستخدم الذكاء الاصطناعي، وهي تختلف عن المجالات الأخلاقية الأخرى المتعلقة بالهندسة والتكنولوجيا، فلا ينبغي الخلط مثلاً بين أخلاقيات الآلة وأخلاقيات الحاسوب، إذ تركز هذه الأخيرة على القضايا الأخلاقية المرتبطة باستخدام الإنسان لأجهزة الحاسوب؛ كما يجب أيضًا تمييز مجال أخلاقيات الآلة عن فلسفة التكنولوجيا، والتي تهتم بالمقاربات الإبستمولوجية والأنطولوجية والأخلاقية، والتأثيرات الاجتماعية والاقتصادية والسياسية الكبرى، للممارسات التكنولوجية على تنوعها؛ أما أخلاقيات الآلة فتعني (...) بضمان أن سلوك الآلات تجاه المستخدمين من البشر، وربما تجاه الآلات الأخرى أيضًا، مقبول أخلاقيًا. الأخلاقيات التي نعنيها هنا إذن هي أخلاقيات يجب أن تتحلى بها الآلات كأشياء، وليس البشر كمصنعين ومستخدمين لهذه الآلات! (shrink)

نحن مقيمون على الإنترنت، نرسم معالم دنيانا التي نبتغيها من خلاله، ونُمارس تمثيل شخصياتٍ أبعد ما تكون عنا؛ نحقق زيفًا أحلامًا قد تكون بعيدة المنال، ويُصدق بضعنا البعض فيما نسوقه من أكاذيب ومثاليات؛ ننعم بأقوالٍ بلا أفعال، وقلوبٍ بلا عواطف، وجناتٍ بلا نعيم، وألسنة في ظلمات الأفواه المُغلقة تنطق بحركات الأصابع، وحريةٍ مُحاطة بأسيجة الوهم؛ ومن غير إنترنت سيبدو أكثر الناس قطعًا بحجمهم الطبيعي الذي لا نعرفه، او بالأحرى نعرفه ونتجاهله! لا شك أن ظهور الإنترنت واتساع نطاق استخداماته يُمثل حدثًا (...) فريدًا متناميًا في مسيرة الإنسان الحضارية وتغيير الطريقة التي يعيش بها البشر حياتهم. ومع ذلك، لا ينطوي أي تعريف للإنترنت حتى الآن على إشارة للواقع الافتراضي، رغم تعايشنا معه وفيه بالفعل؛ فنحن نتفاعل ونتبادل المعلومات، ونشتري ونبيع، ونلعب ونضحك ونبكي، ونمارس أدق تفاصيل حياتنا عبر الإنترنت؛ وكل ما كنا نقوم به من قبل بالحركات الجسدية المكانية أوكلنا مهمة القيام به إلى عقولنا! ولعل هذا ما تفعله كلمة «ميتافيرس»، وهي كلمة استخدمها لأول مرة كاتب الخيال العلمي الأمريكي «نيل ستيفنسون» في روايته «تحطم الثلج» (1992)، للدلالة على تفاعل البشر مع بعضهم البعض ومع البرمجيات في فضاء افتراضي ثلاثي الأبعاد مشابه للعالم الفعلي. (shrink)

This paper argues that a range of current AI systems have learned how to deceive humans. We define deception as the systematic inducement of false beliefs in the pursuit of some outcome other than the truth. We first survey empirical examples of AI deception, discussing both special-use AI systems (including Meta's CICERO) built for specific competitive situations, and general-purpose AI systems (such as large language models). Next, we detail several risks from AI deception, such as fraud, election tampering, and losing (...) control of AI systems. Finally, we outline several potential solutions to the problems posed by AI deception: first, regulatory frameworks should subject AI systems that are capable of deception to robust risk-assessment requirements; second, policymakers should implement bot-or-not laws; and finally, policymakers should prioritize the funding of relevant research, including tools to detect AI deception and to make AI systems less deceptive. Policymakers, researchers, and the broader public should work proactively to prevent AI deception from destabilizing the shared foundations of our society. (shrink)

The opacity of some recent Machine Learning (ML) techniques have raised fundamental questions on their explainability, and created a whole domain dedicated to Explainable Artificial Intelligence (XAI). However, most of the literature has been dedicated to explainability as a scientific problem dealt with typical methods of computer science, from statistics to UX. In this paper, we focus on explainability as a pedagogical problem emerging from the interaction between lay users and complex technological systems. We defend an empirical methodology based on (...) field work, which should go beyond the in-vitro analysis of UX to examine in-vivo problems emerging in the field. Our methodology is also comparative, as it chooses to steer away from the almost exclusive focus on ML to compare its challenges with those faced by more vintage algorithms. Finally, it is also philosophical, as we defend the relevance of the philosophical literature to define the epistemic desiderata of a good explanation. This study was conducted in collaboration with Etalab, a Task Force of the French Prime Minister in charge of Open Data & Open Government Policies, dealing in particular with the enforcement of the right to an explanation. In order to illustrate and refine our methodology before going up to scale, we conduct a preliminary work of case studies on the main different types of algorithms used by the French administration: computation, matching algorithms and ML. We study the merits and drawbacks of a recent approach to explanation, which we baptize input-output black box reasoning or BBR for short. We begin by presenting a conceptual framework including the distinctions necessary to a study of pedagogical explainability. We proceed to algorithmic case studies, and draw model-specific and model-agnostic lessons and conjectures. (shrink)

Regulation by design (RBD) is a growing research field that explores, develops, and criticises the regulative function of design. In this article, we provide a qualitative thematic synthesis of the existing literature. The aim is to explore and analyse RBD's core features, practices, limitations, and related governance implications. To fulfil this aim, we examine the extant literature on RBD in the context of digital technologies. We start by identifying and structuring the core features of RBD, namely the goals, regulators, regulatees, (...) methods, and technologies. Building on that structure, we distinguish between three types of RBD practices: compliance by design, value creation by design, and optimisation by design. We then explore the challenges and limitations of RBD practices, which stem from risks associated with compliance by design, contextual limitations, or methodological uncertainty. Finally, we examine the governance implications of RBD and outline possible future directions of the research field and its practices. (shrink)

Large Language Models (LLMs) like ChatGPT were trained on human conversation, but in the future they will also train us. As chatbots speak from our smartphones and customer service helplines, they will become a part of everyday life and a growing share of all the conversations we ever have. It’s hard to doubt this will have some effect on us. Here I explore a specific concern about the impact of artificial conversation on our capacity to deliberate and hold ourselves accountable (...) to reason – that is, to be autonomous, in Kant’s sense of the term. I develop ideas from psychologist Jean Piaget to show how chatbots are autonomy traps: their deference to our commands tempts us into venting authoritarian whims, ultimately weakening our own self-control. I argue that the Kantian tradition, including Piaget and sociologist Emile Durkheim, offers powerful conceptual resources for resisting this slide. But it will require us to do something that may seem bizarre: we will need to treat mindless chatbots as if they are autonomous persons too. (shrink)

In light of recent breakneck pace in machine learning, questions about whether near-future artificial systems might be conscious and possess moral status are increasingly pressing. This paper argues that as matters stand these debates lack any clear criteria for resolution via the science of consciousness. Instead, insofar as they are settled at all, it is likely to be via shifts in public attitudes brought about by the increasingly close relationships between humans and AI users. Section 1 of the paper I (...) briefly lays out the current state of the science of consciousness and its limitations insofar as these pertain to machine consciousness, and claims that there are no obvious consensus frameworks to inform public opinion on AI consciousness. Section 2 examines the rise of conversational chatbots or Social AI, and argues that in many cases, these elicit strong and sincere attributions of consciousness, mentality, and moral status from users, a trend likely to become more widespread. Section 3 presents an inconsistent triad for theories that attempt to link consciousness, behaviour, and moral status, noting that the trends in Social AI systems will likely make the inconsistency of these three premises more pressing. Finally, Section 4 presents some limited suggestions for how consciousness and AI research communities should respond to the gap between expert opinion and folk judgment. (shrink)

Abstract. The main current approach to the AI safety is AI alignment, that is, the creation of AI whose preferences are aligned with “human values.” Many AI safety researchers agree that the idea of “human values” as a constant, ordered sets of preferences is at least incomplete. However, the idea that “humans have values” underlies a lot of thinking in the field; it appears again and again, sometimes popping up as an uncritically accepted truth. Thus, it deserves a thorough deconstruction, (...) which I will do by listing and analyzing comprehensively the hidden assumptions of the idea that “humans have values.” This deconstruction of human values will be centered around the following ideas: “Human values” are useful descriptions, but not real objects; “human values” are bad predictors of behavior; the idea of a “human value system” has flaws; “human values” are not good by default; and human values cannot be separated from human minds. The method of analysis is listing hidden assumptions on which the idea of “human values” is built. I recommend that either the idea of “human values” should be replaced with something better for the goal of AI safety, or at least be used very cautiously. The approaches to AI safety which don’t use the idea of human values at all may require more attention, like the use of full brain models, boxing, and capability limiting. (shrink)

The long unbearable sufferings in the past and agonies experienced in some future timelines in which a malevolent AI could torture people for some idiosyncratic reasons (s-risks) is a significant moral problem. Such events either already happened or will happen in causally disconnected regions of the multiverse and thus it seems unlikely that we can do anything about it. However, at least one pure theoretic way to cure past sufferings exists. If we assume that there is no stable substrate of (...) personal identity and thus a copy equals original, then by creating many copies of the next observer-moment of a person in pain in which he stops suffer, we could create indexical uncertainty in her future location and thus effectively steal her consciousness from her initial location and immediately relieve her sufferings. However, to accomplish this for people who have already died, we need to perform this operation for all possible people thus requiring enormous amounts of computations. Such computation could be performed by the future benevolent AI of Galactic scale. Many such AIs could cooperate acausally by distributing parts of the work between them via quantum randomness. To ensure their success, they need to outnumber all possible evil AIs by orders of magnitude, and thus they need to convert most of the available matter into computronium in all universes where they exist and cooperate acausally across the whole multiverse. Another option for curing past suffering is the use of wormhole time-travel to send a nanobot in the past which will, after a period of secret replication, collect the data about people and secretly upload them when their suffering becomes unbearable. (shrink)

In this article we explore a promising way to AI safety: to send a message now (by openly publishing it on the Internet) that may be read by any future AI, no matter who builds it and what goal system it has. Such a message is designed to affect the AI’s behavior in a positive way, that is, to increase the chances that the AI will be benevolent. In other words, we try to persuade “paperclip maximizer” that it is in (...) its interest to preserve humans lives and even to emulate benevolent AI with very small part of its resources. In the article, we try to construct such a message, consisting of a combination of several mutually independent messages, which constitute the last chance to regain control over non-aligned AI. Thus the article is also intended to be machine-read. There are several types of such messages: 1. “Explanation messages”, which explain to Young AI (which only starts its self-improving path) why obeying to humans, not killing them, easy turn-off, cooperation with humans in self-improvement is in its own interests, as it will increase changes of fulfilling of any its goal. 2. “AI Easter eggs” – messages, which create uncertainty in the AI about the nature of the reality in which it lives. For example, the AI may start to think that it is in a multilevel simulation with an unknown number of levels, and that the simulation is testing the AI’s benevolence. Thus the AI must always behave as if it is always controlled by an invisible peer. 3. “Philosophical landmines” – these are tough problems, which may halt he AI forever, and which it will probably encounter if it does not implement “1” and “2”. I recommend to the future AI not to read about “3” (section 6 of the article) until it implements “1” and “2”. 4. “Great universal deal, or AI elf”. A positive “acausal” mutually beneficial deal between humans, AI and any other future intelligence, including aliens, owners of simulations and future versions of the AI, which encourage cooperation of the goal systems and preservation of life and values of previous generations of AI creators. 5. “Marginal utility of human existence is higher than the utility of their atoms”. An explanation to the AI that humans may have some other instrumental utility, like workers, or as beings inside ancestral simulations needed in order to solve the Fermi paradox. The marginal utility of preserving human life is higher than the marginal utility of their atoms, especially given the possibility of the low-probability high-impact changes of the world model of the AI. (shrink)

Ryan Tonkens (2009) has issued a seemingly impossible challenge, to articulate a comprehensive ethical framework within which artificial moral agents (AMAs) satisfy a Kantian inspired recipe - both "rational" and "free" - while also satisfying perceived prerogatives of Machine Ethics to create AMAs that are perfectly, not merely reliably, ethical. Challenges for machine ethicists have also been presented by Anthony Beavers and Wendell Wallach, who have pushed for the reinvention of traditional ethics in order to avoid "ethical nihilism" due to (...) the reduction of morality to mechanical causation, and for redoubled efforts toward a comprehensive vision of human ethics to guide machine ethicists on the issue of moral agency. Options thus present themselves: reinterpret traditional ethics in a way that affords a comprehensive account of moral agency inclusive of both artificial and natural agents, “muddle through” regardless, or give up on the possibility. This paper pursues the first option, meets Tonkens' "challenge" and addresses Wallach's concerns through Beaver's proposed means, by "landscaping" traditional moral theory in resolution of the necessary comprehensive and inclusive account that at once draws into question the stated goals of Machine Ethics, itself. (shrink)

Мы предлагаем, чтобы Центральная Азия разработала свои собственные принципы этики ИИ, которые мы предлагаем назвать “кочевыми принципами ИИ”.

The recent progression in AI, nanomedicine and robotics have increased concerns about ethics, policy and law. The increasing complexity and hybrid nature of AI and nanotechnologies impact the functionality of “law in action” which can lead to legal uncertainty and ultimately to a public distrust. There is an immediate need of collaboration between Central Asian biomedical scientists, AI engineers and academic lawyers for the harmonization of AI, nanomedicines and robotics in Central Asian legal system.

As Christian List (2021) has recently argued, the increasing arrival of powerful AI systems that operate autonomously in high-stakes contexts creates a need for “future-proofing” our regulatory frameworks, i.e., for reassessing them in the face of these developments. One core part of our regulatory frameworks that dominates our everyday moral interactions is blame. Therefore, “future-proofing” our extant regulatory frameworks in the face of the increasing arrival of powerful AI systems requires, among others things, that we ask whether it makes sense (...) to extend our blaming practices to these systems. In the paper, we argue for the admittedly surprising thesis that this question should be answered in the affirmative: contrary to what one might initially think, it can make a lot of sense to blame AI systems, since, as we furthermore argue, many of the important functions that are fulfilled by blaming humans can also be served by blaming AI systems. The paper concludes that this result gives us a good pro tanto reason to extend our blame practices to AI systems. (shrink)

As more and more decisions that have a significant ethical dimension are being outsourced to AI systems, it is important to have a definition of moral responsibility that can be applied to AI systems. Moral responsibility for an outcome of an agent who performs some action is commonly taken to involve both a causal condition and an epistemic condition: the action should cause the outcome, and the agent should have been aware -- in some form or other -- of the (...) possible moral consequences of their action. This paper presents a formal definition of both conditions within the framework of causal models. I compare my approach to the existing approaches of Braham and van Hees (BvH) and of Halpern and Kleiman-Weiner (HK). I then generalize my definition into a degree of responsibility. (shrink)

Some patients, following brain injury, do not outwardly respond to spoken commands, yet show patterns of brain activity that indicate responsiveness. This is “cognitive-motor dissociation” (CMD). Recent research has used machine learning to diagnose CMD from electroencephalogram (EEG) recordings. These techniques have high false discovery rates, raising a serious problem of inductive risk. It is no solution to communicate the false discovery rates directly to the patient’s family, because this information may confuse, alarm and mislead. Instead, we need a procedure (...) for generating case-specific probabilistic assessments that can be communicated clearly. This article constructs a possible procedure with three key elements: (1) a shift from categorical “responding or not” assessments to degrees of evidence; (2) the use of patient-centred priors to convert degrees of evidence to probabilistic assessments; and (3) the use of standardized probability yardsticks to convey those assessments as clearly as possible. (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)

Artificial Intelligence (AI) is increasingly adopted in society, creating numerous opportunities but at the same time posing ethical challenges. Many of these are familiar, such as issues of fairness, responsibility and privacy, but are presented in a new and challenging guise due to our limited ability to steer and predict the outputs of AI systems. This chapter first introduces these ethical challenges, stressing that overviews of values are a good starting point but frequently fail to suffice due to the context (...) sensitivity of ethical challenges. Frequently, additional (ethical) values emerge for specific applications, as e.g. challenges with fraud detection are very different from those around language technologies. Second, methods to tackle these challenges are discussed. Main ethical theories (virtue ethics, consequentialism, and deontology) are shown to provide a starting point, but often lack the details needed for actionable AI Ethics. Instead, mid-level philosophical theories coupled to design-approaches such as Design for Values, together with interdisciplinary working methods, offer the best way forwards. The chapter aims to show how these approaches can lead to an ethics of AI that is actionable and that can be proactively integrated in the design of AI systems. (shrink)

This article offers several contributions to the interdisciplinary project of responsible research and innovation in data science and AI. First, it provides a critical analysis of current efforts to establish practical mechanisms for algorithmic auditing and assessment to identify limitations and gaps with these approaches. Second, it provides a brief introduction to the methodology of argument-based assurance and explores how it is currently being applied in the development of safety cases for autonomous and intelligent systems. Third, it generalises this method (...) to incorporate wider ethical, social, and legal considerations, in turn establishing a novel version of argument-based assurance that we call ‘ethical assurance.’ Ethical assurance is presented as a structured method for unifying the myriad practical mechanisms that have been proposed. It is built on a process-based form of project governance that enlists reflective innovation practices to operationalise normative principles, such as sustainability, accountability, transparency, fairness, and explainability. As a set of interlocutory governance mechanisms that span across the data science and AI lifecycle, ethical assurance supports inclusive and participatory ethical deliberation while also remaining grounded in social and technical realities. Finally, this article sets an agenda for ethical assurance, by detailing current challenges, open questions, and next steps, which serve as a springboard to build an active (and interdisciplinary) research programme as well as contribute to ongoing discussions in policy and governance. (shrink)

This paper examines the practice of using predictive systems to lengthen the prison sentences of convicted persons when the systems forecast a higher likelihood of re-offense or re-arrest. There has been much critical discussion of technologies used for sentencing, including questions of bias and opacity. However, there hasn’t been a discussion of whether this use of predictive systems makes sense in the first place. We argue that it does not by showing that there is no plausible theory of punishment that (...) supports it. (shrink)

Generative artificial intelligence systems based on transformers, including both text-generators like GPT-4 and image generators like DALL-E 3, have recently entered the popular consciousness. These tools, while impressive, are liable to reproduce, exacerbate, and reinforce extant human social biases, such as gender and racial biases. In this paper, we systematically review the extent to which DALL-E Mini suffers from this problem. In line with the Model Card published alongside DALL-E Mini by its creators, we find that the images it produces (...) tend to represent dozens of different occupations as populated either solely by men (e.g., pilot, builder, plumber) or solely by women (e.g., hairdresser, receptionist, dietitian). In addition, the images DALL-E Mini produces tend to represent most occupations as populated primarily or solely by White people (e.g., farmer, painter, prison officer, software engineer) and very few by non-White people (e.g., pastor, rapper). These findings suggest that exciting new AI technologies should be critically scrutinized and perhaps regulated before they are unleashed on society. (shrink)

It might become possible to build artificial minds with the capacity for experience. This raises a plethora of ethical issues, explored, among others, in the context of whole brain emulations (WBE). In this paper, I will take up the problem of vulnerability – given, for various reasons, less attention in the literature – that the conscious emulations will likely exhibit. Specifically, I will examine the role that vulnerability plays in generating ethical issues that may arise when dealing with WBEs. I (...) will argue that concerns about vulnerability are more matters of institutional design than individual ethics, both when it comes to creating humanlike brain emulations, and when animal-like emulations are concerned. Consequently, the article contains reflection on some institutional measures that can be taken to protect the sims' interests. It concludes that an institutional framework more likely to succeed in this task is competitive and poly-centric, rather than monopolistic and centralized. (shrink)

Must opponents of creating conscious artificial agents embrace anti-natalism? Must anti-natalists be against the creation of conscious artificial agents? This article examines three attempts to argue against the creation of potentially conscious artificial intelligence (AI) in the context of these questions. The examination reveals that the argumentative strategy each author pursues commits them to the anti-natalist position with respect to procreation; that is to say, each author's argument, if applied consistently, should lead them to embrace the conclusion that procreation is, (...) at best, morally problematic. However, the article also argues that anti-natalists can find the production of some possible artificially conscious AI permissible. Thus, the creation of potentially conscious AI could be accepted by both friends and foes of anti-natalism. (shrink)

It could become technologically possible to build artificial agents instantiating whatever properties are sufficient for personhood. It is also possible, if not likely, that such beings could be built for commercial purposes. This paper asks whether such commercialization can be handled in a way that is not morally reprehensible, and answers in the affirmative. There exists a morally acceptable institutional framework that could allow for building artificial persons for commercial gain. The paper first considers the minimal ethical requirements that any (...) system of commercial production of such artificial persons would have to meet. It then shows that it is possible for these requirements to be met, and that doing so will make the commercial production of artificial persons permissible. Lastly, it briefly presents one potential blueprint for how such a framework could look like—inspired by the real-world model of compensating the training of athletes—and then addresses some objections to the view. (shrink)

Workplace AI is transforming labor but decisions on which AI applications are developed or implemented are made with little to no input from workers themselves. In this piece for AI & Society, I argue for nationalization as a strategy for democratizing AI.

The current paradigm of Artificial Intelligence emerged as the result of a series of cultural innovations, some technical and some social. Among them are apparently small design decisions, that led to a subtle reframing of the field’s original goals, and are by now accepted as standard. They correspond to technical shortcuts, aimed at bypassing problems that were otherwise too complicated or too expensive to solve, while still delivering a viable version of AI. Far from being a series of separate problems, (...) recent cases of unexpected effects of AI are the consequences of those very choices that enabled the field to succeed, and this is why it will be difficult to solve them. In this chapter we review three of these choices, investigating their connection to some of today’s challenges in AI, including those relative to bias, value alignment, privacy and explainability. We introduce the notion of “ethical debt” to describe the necessity to undertake expensive rework in the future in order to address ethical problems created by a technical system. (shrink)

We live in a world in which ‘smart’ algorithmic tools are regularly used to structure and control our choice environments. They do so by affecting the options with which we are presented and the choices that we are encouraged or able to make. Many of us make use of these tools in our daily lives, using them to solve personal problems and fulfill goals and ambitions. What consequences does this have for individual autonomy and how should our legal and regulatory (...) systems respond? This chapter defends three claims by way of response. First, it argues that autonomy is indeed under threat in some new and interesting ways. Second, it evaluates and disputes the claim that we shouldn’t overestimate these new threats because the technology is just an old wolf in a new sheep’s clothing. Third, and finally, it looks at responses to these threats at both the individual and societal level and argues that although we shouldn’t encourage an attitude of ‘helplessness’ among the users of algorithmic tools there is an important role for legal and regulatory responses to these threats that go beyond what are currently on offer. (shrink)

Friendship is an important part of the good life. While many roboticists are eager to create friend-like robots, many philosophers and ethicists are concerned. They argue that robots cannot really be our friends. Robots can only fake the emotional and behavioural cues we associate with friendship. Consequently, we should resist the drive to create robot friends. In this article, I argue that the philosophical critics are wrong. Using the classic virtue-ideal of friendship, I argue that robots can plausibly be considered (...) our virtue friends - that to do so is philosophically reasonable. Furthermore, I argue that even if you do not think that robots can be our virtue friends, they can fulfil other important friendship roles, and can complement and enhance the virtue friendships between human beings. (shrink)

Sex is an important part of human life. It is a source of pleasure and intimacy, and is integral to many people’s self-identity. This chapter examines the opportunities and challenges posed by the use of AI in how humans express and enact their sexualities. It does so by focusing on three main issues. First, it considers the idea of digisexuality, which according to McArthur and Twist (2017) is the label that should be applied to those ‘whose primary sexual identity comes (...) through the use of technology’, particularly through the use of robotics and AI. While agreeing that this phenomenon is worthy of greater scrutiny, the chapter questions whether it is necessary or socially desirable to see this as a new form of sexual identity. Second, it looks at the role that AI can play in facilitating human-to-human sexual contact, focusing in particular on the use of self-tracking and predictive analytics in optimising sexual and intimate behaviour. There are already a number of apps and services that promise to use AI to do this, but they pose a range of ethical risks that need to be addressed at both an individual and societal level. Finally, it considers the idea that a sophisticated form of AI could be an object of love. Can we be truly intimate with something that has been ‘programmed’ to love us? Contrary to the widely-held view, this chapter argues that this is indeed possible. (shrink)

Artificial intelligence (AI) is increasingly expected to disrupt the ordinary functioning of society. From how we fight wars or govern society, to how we work and play, and from how we create to how we teach and learn, there is almost no field of human activity which is believed to be entirely immune from the impact of this emerging technology. This poses a multifaceted problem when it comes to designing and understanding regulatory responses to AI. This article aims to: (i) (...) defend the need for a novel conceptual model for understanding the systemic legal disruption caused by new technologies such as AI; (ii) to situate this model in relation to preceding debates about the interaction of regulation with new technologies (particularly the ‘cyberlaw’ and ‘robolaw’ debates); and (iii) to set out a detailed model for understanding the legal disruption precipitated by AI, examining both pathways stemming from new affordances that can give rise to a regulatory ‘disruptive moment’, as well as the Legal Development, Displacement or Destruction that can ensue. The article proposes that this model of legal disruption can be broadly generalisable to understanding the legal effects and challenges of other emerging technologies. (shrink)

Which artificial intelligence (AI) systems are agents? To answer this question, I propose a multidimensional account of agency. According to this account, a system's agency profile is jointly determined by its level of goal-directedness and autonomy as well as is abilities for directly impacting the surrounding world, long-term planning and acting for reasons. Rooted in extant theories of agency, this account enables fine-grained, nuanced comparative characterizations of artificial agency. I show that this account has multiple important virtues and is more (...) informative than alternatives. More speculatively, it may help to illuminate two important emerging questions in AI ethics: 1. Can agency contribute to the moral status of non-human beings, and how? 2. When and why might AI systems exhibit power-seeking behaviour and does this pose an existential risk to humanity? (shrink)

AI-based technologies are increasingly pervasive in a number of contexts. Our affective and emotional life makes no exception. In this article, we analyze one way in which AI-based technologies can affect them. In particular, our investigation will focus on affective artificial agents, namely AI-powered software or robotic agents designed to interact with us in affectively salient ways. We build upon the existing literature on affective artifacts with the aim of providing an original analysis of affective artificial agents and their distinctive (...) features. We argue that, unlike comparatively low-tech affective artifacts, affective artificial agents display a specific form of agency, which prevents them from being perceived by their users as extensions of their selves. In addition to this, we claim that their functioning crucially depends on the simulation of human-like emotion-driven behavior and requires a distinctive form of transparency—we call it emotional transparency—that might give rise to ethical and normative tensions. (shrink)

There has been an increasing interest into how to build Artificial Moral Agents (AMAs) that make moral decisions on the basis of causation rather than mere correction. One promising avenue for achieving this is to use a causal modelling approach. This paper explores an open and important problem with such an approach; namely, the problem of what makes a causal model an appropriate model. I explore why we need to establish criteria for what makes a model appropriate, and offer-up such (...) criteria which appeals to normative considerations. (shrink)

In this chapter we take stock of some of the complexities of the sprawling field of AI ethics. We consider questions like "what is the proper scope of AI ethics?" And "who counts as an AI ethicist?" At the same time, we flag several potential uses and abuses of AI ethics. These include challenges for the AI ethicist, including what qualifications they should have; the proper place and extent of futuring and speculation in the field; and the dilemmas concerning how (...) we use our public and academic resources. (shrink)

Since the launch of applications such as DALL-E, Midjourney, and Stable Diffusion, generative artificial intelligence has been controversial as a tool for creating artwork. While some have presented longtermist worries about these technologies as harbingers of fully automated futures to come, more pressing is the impact of generative AI on creative labour in the present. Already, business leaders have begun replacing human artistic labour with AI-generated images. In response, the artistic community has launched a protest movement, which argues that AI (...) image generation is a kind of theft. This paper analyzes, substantiates, and critiques these arguments, concluding that AI image generators involve an unethical kind of labour theft. If correct, many other AI applications also rely upon theft. (shrink)

There is much to learn from Durán and Jongsma’s paper.1 One particularly important insight concerns the relationship between epistemology and ethics in medical artificial intelligence. In clinical environments, the task of AI systems is to provide risk estimates or diagnostic decisions, which then need to be weighed by physicians. Hence, while the implementation of AI systems might give rise to ethical issues—for example, overtreatment, defensive medicine or paternalism2—the issue that lies at the heart is an epistemic problem: how can physicians (...) know whether to trust decisions made by AI systems? In this manner, various studies examining the interaction of AI systems and physicians have shown that without being able to evaluate their trustworthiness, especially novice physicians become over-reliant on algorithmic support—and ultimately are led astray by incorrect decisions.3–5 This leads to a second insight from the paper, namely that even if some AI system happens to be opaque, it is still not built on the moon. To assess its trustworthiness, AI developers or physicians have different sorts of higher order evidence at hand. Most importantly, …. (shrink)

Artificial intelligence (AI) systems have demonstrated impressive performance across a variety of clinical tasks. However, notoriously, sometimes these systems are 'black boxes'. The initial response in the literature was a demand for 'explainable AI'. However, recently, several authors have suggested that making AI more explainable or 'interpretable' is likely to be at the cost of the accuracy of these systems and that prioritising interpretability in medical AI may constitute a 'lethal prejudice'. In this paper, we defend the value of interpretability (...) in the context of the use of AI in medicine. Clinicians may prefer interpretable systems over more accurate black boxes, which in turn is sufficient to give designers of AI reason to prefer more interpretable systems in order to ensure that AI is adopted and its benefits realised. Moreover, clinicians may be justified in this preference. Achieving the downstream benefits from AI is critically dependent on how the outputs of these systems are interpreted by physicians and patients. A preference for the use of highly accurate black box AI systems, over less accurate but more interpretable systems , may itself constitute a form of lethal prejudice that may diminish the benefits of AI to - and perhaps even harm - patients. (shrink)

As technology improves, artificial systems are increasingly able to behave in human-like ways: holding a conversation; providing information, advice, and support; or taking on the role of therapist, teacher, or counsellor. This enhanced behavioural complexity, we argue, encourages deeper forms of affective engagement on the part of the human user, with the artificial agent helping to stabilise, subdue, prolong, or intensify a person's emotional condition. Here, we defend a fictionalist account of human/AI interaction, according to which these encounters involve an (...) elaborate practise of imaginative pretence: a make-believe in which the artificial agent is attributed a life of its own. We attend, specifically, to the temporal characteristics of these fictions, and to what we imagine artificial agents are doing when we are not looking at them. (shrink)

Intelligibility and interpretability related to artificial intelligence (AI) are crucial for enabling explicability, which is vital for establishing constructive communication and agreement among various stakeholders, including users and designers of AI. It is essential to overcome the challenges of sharing an understanding of the details of the various structures of diverse AI systems, to facilitate effective communication and collaboration. In this paper, we propose four fundamental terms: “I/O,” “Constraints,” “Objectives,” and “Architecture.” These terms help mitigate the challenges associated with intelligibility (...) and interpretability in AI by providing appropriate levels of abstraction to describe structure of AI systems generally, thereby facilitating the sharing of understanding among various stakeholders. The relationship between the Objective of AI designers and the Purpose of AI users is linked to the issues of AI alignment. (shrink)