We propose a formal framework for interpretable machine learning. Combining elements from statistical learning, causal interventionism, and decision theory, we design an idealised explanation game in which players collaborate to find the best explanation for a given algorithmic prediction. Through an iterative procedure of questions and answers, the players establish a three-dimensional Pareto frontier that describes the optimal trade-offs between explanatory accuracy, simplicity, and relevance. Multiple rounds are played at different levels of abstraction, allowing the players to explore overlapping causal (...) patterns of variable granularity and scope. We characterise the conditions under which such a game is almost surely guaranteed to converge on a optimal explanation surface in polynomial time, and highlight obstacles that will tend to prevent the players from advancing beyond certain explanatory thresholds. The game serves a descriptive and a normative function, establishing a conceptual space in which to analyse and compare existing proposals, as well as design new and improved solutions. (shrink)

The last 5 years have seen a series of remarkable achievements in deep-neural-network-based artificial intelligence research, and some modellers have argued that their performance compares favourably to human cognition. Critics, however, have argued that processing in deep neural networks is unlike human cognition for four reasons: they are (i) data-hungry, (ii) brittle, and (iii) inscrutable black boxes that merely (iv) reward-hack rather than learn real solutions to problems. This article rebuts these criticisms by exposing comparative bias within them, in the (...) process extracting some more general lessons that may also be useful for future debates. (shrink)

AI research is growing rapidly raising various ethical issues related to safety, risks, and other effects widely discussed in the literature. We believe that in order to adequately address those issues and engage in a productive normative discussion it is necessary to examine key concepts and categories. One such category is anthropomorphism. It is a well-known fact that AI’s functionalities and innovations are often anthropomorphized (i.e., described and conceived as characterized by human traits). The general public’s anthropomorphic attitudes and some (...) of their ethical consequences (particularly in the context of social robots and their interaction with humans) have been widely discussed in the literature. However, how anthropomorphism permeates AI research itself (i.e., in the very language of computer scientists, designers, and programmers), and what the epistemological and ethical consequences of this might be have received less attention. In this paper we explore this issue. We first set the methodological/theoretical stage, making a distinction between a normative and a conceptual approach to the issues. Next, after a brief analysis of anthropomorphism and its manifestations in the public, we explore its presence within AI research with a particular focus on brain-inspired AI. Finally, on the basis of our analysis, we identify some potential epistemological and ethical consequences of the use of anthropomorphic language and discourse within the AI research community, thus reinforcing the need of complementing the practical with a conceptual analysis. (shrink)

Machine learning-based AI algorithms lack transparency. In this article, I offer an interpretation of AI’s explainability problem and highlight its ethical saliency. I try to make the case for the legal enforcement of a strong explainability requirement: human organizations which decide to automate decision-making should be legally obliged to demonstrate the capacity to explain and justify the algorithmic decisions that have an impact on the wellbeing, rights, and opportunities of those affected by the decisions. This legal duty can be derived (...) from the demands of Rawlsian public reason. In the second part of the paper, I try to show that the argument from the limitations of human cognition fails to get AI off the hook of public reason. Against a growing trend in AI ethics, my main argument is that the analogy between human minds and artificial neural networks fails because it suffers from an atomistic bias which makes it blind to the social and institutional dimension of human reasoning processes. I suggest that developing interpretive AI algorithms is not the only possible answer to the explainability problem; social and institutional answers are also available and in many cases more trustworthy than techno-scientific ones. (shrink)

We propose a formal framework for interpretable machine learning. Combining elements from statistical learning, causal interventionism, and decision theory, we design an idealisedexplanation gamein which players collaborate to find the best explanation(s) for a given algorithmic prediction. Through an iterative procedure of questions and answers, the players establish a three-dimensional Pareto frontier that describes the optimal trade-offs between explanatory accuracy, simplicity, and relevance. Multiple rounds are played at different levels of abstraction, allowing the players to explore overlapping causal patterns of (...) variable granularity and scope. We characterise the conditions under which such a game is almost surely guaranteed to converge on a (conditionally) optimal explanation surface in polynomial time, and highlight obstacles that will tend to prevent the players from advancing beyond certain explanatory thresholds. The game serves a descriptive and a normative function, establishing a conceptual space in which to analyse and compare existing proposals, as well as design new and improved solutions. (shrink)

While philosophers hold that it is patently absurd to blame robots or hold them morally responsible [1], a series of recent empirical studies suggest that people do ascribe blame to AI systems and robots in certain contexts [2]. This is disconcerting: Blame might be shifted from the owners, users or designers of AI systems to the systems themselves, leading to the diminished accountability of the responsible human agents [3]. In this paper, we explore one of the potential underlying reasons for (...) robot blame, namely the folk's willingness to ascribe inculpating mental states or "mens rea" to robots. In a vignette-based experiment (N=513), we presented participants with a situation in which an agent knowingly runs the risk of bringing about substantial harm. We manipulated agent type (human v. group agent v. AI-driven robot) and outcome (neutral v. bad), and measured both moral judgment (wrongness of the action and blameworthiness of the agent) and mental states attributed to the agent (recklessness and the desire to inflict harm). We found that (i) judgments of wrongness and blame were relatively similar across agent types, possibly because (ii) attributions of mental states were, as suspected, similar across agent types. This raised the question - also explored in the experiment - whether people attribute knowledge and desire to robots in a merely metaphorical way (e.g., the robot "knew" rather than really knew). However, (iii), according to our data people were unwilling to downgrade to mens rea in a merely metaphorical sense when given the chance. Finally, (iv), we report a surprising and novel finding, which we call the inverse outcome effect on robot blame: People were less willing to blame artificial agents for bad outcomes than for neutral outcomes. This suggests that they are implicitly aware of the dangers of overattributing blame to robots when harm comes to pass, such as inappropriately letting the responsible human agent off the moral hook. (shrink)

Through the lens of feminist speech act theory, this paper argues that artificial intelligence romance systems objectify and subordinate nonvirtual women. AI romance systems treat their users as consumers, offering them relational invulnerability and control over their (usually feminized) digital romantic partner. This paper argues that, though the output of AI chatbots may not generally constitute speech, the framework offered by an AI romance system communicates an unjust perspective on intimate relationships. Through normalizing controlling one’s intimate partner, these systems operate (...) as speakers that rank women as non-player characters and license their oppression, which is unjust even in the absence of empirical harms. Understanding AI romance systems as speakers has implications for policymakers. First, this account helps close the so-called responsibility gap in the operation of an AI system. Second, it places these systems under the purview of institutions’ policies regarding hate speech, suggesting one avenue for arguing that these systems may violate extant policies. Finally, it provides support for the pursuit of empirical research into AI romance systems’ effects on users. (shrink)

According to a widespread view, people often anthropomorphize machines such as certain robots and computer and AI systems by erroneously attributing mental states to them. On this view, people almost irresistibly believe, even if only subconsciously, that machines with certain human-like features really have phenomenal or subjective experiences like sadness, happiness, desire, pain, joy, and distress, even though they lack such feelings. This paper questions this view by critiquing common arguments used to support it and by suggesting an alternative explanation. (...) Even if people’s behavior and language regarding human-like machines suggests they believe those machines really have mental states, it is possible that they do not believe that at all. The paper also briefly discusses potential implications of regarding such anthropomorphism as a popular myth. The exercise illuminates the difficult concept of anthropomorphism, helping to clarify possible human relations with or toward machines that increasingly resemble humans and animals. (shrink)

In this paper, we demonstrate how the language and reasonings that academics, developers, consumers, marketers, and journalists deploy to accept or reject AI as authentic intelligence has far-reaching bearing on how we understand our human intelligence and condition. The discourse on AI is part of what we call the “authenticity negotiation process” through which AI’s “intelligence” is given a particular meaning and value. This has implications for scientific theory, research directions, ethical guidelines, design principles, funding, media attention, and the way (...) people relate to and act upon AI. It also has great impact on humanity’s self-image and the way we negotiate what it means to be human, existentially, culturally, politically, and legally. We use a discourse analysis of academic papers, AI education programs, and online discussions to demonstrate how AI itself, as well as the products, services, and decisions delivered by AI systems are negotiated as authentic or inauthentic intelligence. In this negotiation process, AI stakeholders indirectly define and essentialize what being human(like) means. The main argument we will develop is that this process of indirectly defining and essentializing humans results in an elimination of the space for humans to be indeterminate. By eliminating this space and, hence, denying indeterminacy, the existential condition of the human being is jeopardized. Rather than re-creating humanity in AI, the AI discourse is re-defining what it means to be human and how humanity is valued and should be treated. (shrink)

Can AI solve the ethical, moral, and political dilemmas of warfare? How is artificial intelligence (AI)-enabled warfare changing the way we think about the ethical-political dilemmas and practice of war? This article explores the key elements of the ethical, moral, and political dilemmas of human-machine interactions in modern digitized warfare. It provides a counterpoint to the argument that AI “rational” efficiency can simultaneously offer a viable solution to human psychological and biological fallibility in combat while retaining “meaningful” human control over (...) the war machine. This Panglossian assumption neglects the psychological features of human-machine interactions, the pace at which future AI-enabled conflict will be fought, and the complex and chaotic nature of modern war. The article expounds key psychological insights of human-machine interactions to elucidate how AI shapes our capacity to think about future warfare's political and ethical dilemmas. It argues that through the psychological process of human-machine integration, AI will not merely force-multiply existing advanced weaponry but will become de facto strategic actors in warfare – the “AI commander problem.”. (shrink)

Choice architecture describes the environment in which choices are presented to decision-makers. In recent years, public and private actors have looked at choice architecture with great interest as they seek to influence human behaviour. These actors are typically called choice architects. Increasingly, however, this role of architecting choice is not performed by a human choice architect, but an algorithm or artificial intelligence, powered by a stream of Big Data and infused with an objective it has been programmed to maximise. We (...) call this entity the autonomous choice architect. In this paper, we present an account of why artificial intelligence can fulfil the role of a choice architect and why this creates problems of transparency, responsibility and accountability for nudges. We argue that choice architects, be them autonomous computational systems or human-beings, at a most basic level select, from a range of designs, the design which is most likely to maximise a pre-determined objective. We then proceed to argue that, given the growing demand for targeted, personalised choice architecture and for faster, dynamic reconfigurations of choice architecture, as well as the ever-expanding pool of data from which feedback can be drawn, the role of the human choice architect is increasingly obscured behind algorithmic, artificially intelligent systems. We provide a discussion of the implications of autonomous choice architects, focusing on the importance of the humans who programme these systems, ultimately arguing that despite technological advances, the responsibility of choice architecture and influence remains firmly one human beings must bear. (shrink)

Assistive systems based on Artificial Intelligence (AI) are bound to reshape decision-making in all areas of society. One of the most intricate challenges arising from their implementation in high-stakes environments such as medicine concerns their frequently unsatisfying levels of explainability, especially in the guise of the so-called black-box problem: highly successful models based on deep learning seem to be inherently opaque, resisting comprehensive explanations. This may explain why some scholars claim that research should focus on rendering AI systems understandable, rather (...) than explainable. Yet, there is a grave lack of agreement concerning these terms in much of the literature on AI. We argue that the seminal distinction made by the philosopher and physician Karl Jaspers between different types of explaining and understanding in psychopathology can be used to promote greater conceptual clarity in the context of Machine Learning (ML). Following Jaspers, we claim that explaining and understanding constitute multi-faceted epistemic approaches that should not be seen as mutually exclusive, but rather as complementary ones as in and of themselves they are necessarily limited. Drawing on the famous example of Watson for Oncology we highlight how Jaspers’ methodology translates to the case of medical AI. Classical considerations from the philosophy of psychiatry can therefore inform a debate at the centre of current AI ethics, which in turn may be crucial for a successful implementation of ethically and legally sound AI in medicine. (shrink)

In this paper, I draw on Hannah Arendt’s notion of ‘banality of evil’ to argue that as long as AI systems are designed to follow codes of ethics or particular normative ethical theories chosen by us and programmed in them, they are Eichmanns destined to commit evil. Since intelligence alone is not sufficient for ethical decision making, rather than strive to program AI to determine the right ethical decision based on some ethical theory or criteria, AI should be concerned with (...) avoiding making the wrong decisions, and this requires hardwiring the thinking activity as a prerequisite for decision making. (shrink)

Explanations are conceived to ensure the trustworthiness of AI systems. Yet, relying solemnly on algorithmic solutions, as provided by explainable artificial intelligence (XAI), might fall short to account for sociotechnical risks jeopardizing their factuality and informativeness. To mitigate these risks, we delve into the complex landscape of ethical risks surrounding XAI systems and their generated explanations. By employing a literature review combined with rigorous thematic analysis, we uncover a diverse array of technical risks tied to the robustness, fairness, and evaluation (...) of XAI systems. Furthermore, we address a broader range of contextual risks jeopardizing their security, accountability, reception alongside other cognitive, social, and ethical concerns of explanations. We advance a multi-layered risk assessment framework, where each layer advances strategies for practical intervention, management, and documentation of XAI systems within organizations. Recognizing the theoretical nature of the framework advanced, we discuss it in a conceptual case study. For the XAI community, our multifaceted investigation represents a path to practically address XAI risks while enriching our understanding of the ethical ramifications of incorporating XAI in decision-making processes. (shrink)

This article argues that large language models (LLMs) should be interpreted as a form of gods. In a theological sense, a god is an immortal being that exists beyond time and space. This is clearly nothing like LLMs. In an anthropological sense, however, a god is rather defined as the personified authority of a group through time—a conceptual tool that molds a collective of ancestors into a unified agent or voice. This is exactly what LLMs are. They are products of (...) vast volumes of data, literally traces of past human (speech) acts, synthesized into a single agency that is (falsely) experienced by users as extra-human. This reconceptualization, I argue, opens up new avenues of critique of LLMs by allowing the mobilization of theoretical resources from centuries of religious critique. For illustration, I draw on the Marxian religious philosophy of Martin Hägglund. From this perspective, the danger of LLMs emerge not only as bias or unpredictability, but as a temptation to abdicate our spiritual and ultimately democratic freedom in favor of what I call a tyranny of the past. (shrink)

Three decades ago, William Ramsey, Steven Stich & Joseph Garon put forward an argument in favor of the following conditional: if connectionist models that implement parallelly distributed processing represent faithfully human cognitive processing, eliminativism about propositional attitudes is true. The corollary of their argument (if it proves to be sound) is that there is no place for folk psychology in contemporary cognitive science. This understanding of connectionism as a hypothesis about cognitive architecture compatible with eliminativism is also endorsed by Paul (...) Churchland, a radical opponent of folk psychology and a prominent supporter of eliminative materialism. I aim to examine whether current connectionist models based on long-short term memory (LSTM) neural networks can back up these arguments in favor of eliminativism. Nonetheless, I will rather put my faith in the eliminativism of the limited domain. This position amounts to the claim that even though connectionist cognitive science has no need whatsoever for folk psychology qua theory, this does not entail the illegitimacy of folk psychology per se in other scientific domains, most notably in humanities, but only on condition that one regards folk psychology as mere heuristics. (shrink)

Despite the reality that self-learning artificial intelligence systems (SLAIS) are gaining in sophistication, humanity’s focus regarding SLAIS-human interactions are unnervingly centred upon transnational commercial sectors and, most generally, around issues of intellectual property law. But as SLAIS gain greater environmental interaction capabilities in digital spaces, or the ability to self-author code to drive their development as algorithmic models, a concern arises as to whether a system that displays a “deceptive” level of human-like engagement with users in our physical world ought (...) to be uniquely protected. Although many voices in the legal and technology realms have continued to argue against unique protections for digital entities, the fact at hand is that SLAIS design is becoming increasingly anthropomorphic so as to make these systems more capable of interacting with a wide range of (potentially) vulnerable populations—generally as a means to enhance these populations’ overall well-being. To frame this concern in a different way, the specific question at hand is whether a human’s “ownership” of such an advanced SLAIS is legal, considering that it (or they) may possess intelligence on par with a human or a convincing-enough display of such behaviour. Given that “ownership” over entities with (seemingly) intelligent behaviours consistent with human populations has been effectively banned by the international community, an examination into this subject and its implications is wholly necessary given humanity’s quest to exist solely in digital environments through whatever means possible. (shrink)

The thesis introduces and criticizes the discourse on technology, with a specific reference to the concept of AI. The discourse on AI is particularly saturated with reified metaphors which drive connotations and delimit understandings of technology in society. To better analyse the discourse on AI, the thesis proposes the concept of “Expected AI”, a composite signifier filled with historical and sociocultural connotations, and numerous referent objects. Relying on cultural semiotics, science and technology studies, and a diverse selection of heuristic concepts, (...) the thesis delves beneath the surface of AI discourse and demonstrates the hidden political, social, cultural, and ecological dangers of AI. The entanglement of the discourse(s) with (science) fiction, folklore, myth, and religion impacts how AI is perceived and received, as well as the expectations to AI-enabled technologies now and in the future. The thesis also proposes a more ethical and comprehensive ontological model for AI systems. The model describes AI systems as complex figurations, considering their socio-material organisation, global economic-material becoming, and impact on the environment, social institutions, and the semiosphere. The dissertation argues that AI should be understood not just as an object or sociotechnical system, but as its entire product chain encompassing people and cultures, as well as the used resources and impact (both material-ecological and semiotic) on a planetary scale. (shrink)