Joshua Hatherley

Postmortem avatars (PMAs) — AI systems that simulate a deceased person by being fine-tuned on data they generated or that was generated about them — have attracted growing scholarly attention, yet their potential role in clinical settings remains largely unexplored. This paper examines the ethics of deploying PMAs as therapeutic tools in grief therapy. Drawing on the dual-process model of grief, the theory of continuing bonds, and the philosophical framework of fictionalism, we propose two potential therapeutic applications: incorporating PMAs into established imaginal exercises such as the empty chair exercise, and treating the process of PMA creation as an art-therapeutic exercise in its own right. We consider five ethical objections to these applications and argue that none constitute knock-down arguments against therapeutic use, particularly given the risk-mitigating role of the clinical context. We conclude by identifying outstanding governance challenges and calling for empirical research, without which neither the promise nor the dangers of therapeutic PMAs can be adequately assessed.

Federated learning (FL) is a machine learning (ML) approach that allows multiple devices or institutions to collaboratively train an ML model without sharing their local data with a third-party. It has recently received significant attention as a promising way to overcome longstanding ethical obstacles to training medical ML models with patient health data. This paper examines the promise of FL in healthcare from an ethical perspective. It argues that medical FL generates a new variety of opacity – federation opacity, wherein stakeholders cannot access, analyze, or curate the data on which a model has been trained – which (a) presents distinctive ethical challenges concerning institutional fairness and accountability in medical ML; and (b) makes FL models especially vulnerable to data poisoning attacks. It then identifies several key claims about the expected benefits of FL in healthcare and argues that they may be either exaggerated, misleading, or incomplete – often due to the problem of federation opacity.

Mechanistic interpretability (MI) aims to explain how neural networks work by uncovering their underlying mechanisms. As the field grows in influence, it is increasingly important to examine not just models themselves, but the assumptions, concepts and explanatory strategies implicit in MI research. We argue that mechanistic interpretability needs philosophy as an ongoing partner in clarifying its concepts, refining its methods, and navigating the epistemic and ethical complexities of interpreting AI systems. There is significant unrealised potential for progress in MI to be gained through deeper engagement with philosophers and philosophical frameworks. Taking three open problems from the MI literature as examples, this paper illustrates the value philosophy can add to MI research, and outlines a path toward deeper interdisciplinary dialogue.

Federated learning (FL) is a machine learning approach that allows multiple devices or institutions to collaboratively train a model without sharing their local data with a third-party. FL is considered a promising way to address patient privacy concerns in medical artificial intelligence. The ethical risks of medical FL systems themselves, however, have thus far been underexamined. This paper aims to address this gap. We argue that medical FL presents a new variety of opacity -- federation opacity -- that, in turn, generates a distinctive *double* black box problem in healthcare AI. We highlight several instances in which the anticipated benefits of medical FL may be exaggerated, and conclude by highlighting key challenges that must be overcome to make FL ethically feasible in medicine.

Machine learning (ML) systems are vulnerable to performance decline over time due to dataset shift. To address this problem, experts often suggest that ML systems should be regularly updated to ensure ongoing performance stability. Some scholarly literature has begun to address the epistemic and ethical challenges associated with different updating methodologies. Thus far, however, little attention has been paid to the impact of model updating on the ML-assisted decision-making process itself. This article aims to address this gap. It argues that model updating introduces a new sub-type of opacity into ML-assisted decision-making—update opacity—that occurs when users cannot understand how or why an update has changed the reasoning or behaviour of an ML system. This type of opacity presents a variety of distinctive epistemic and safety concerns that available solutions to the black box problem in ML are largely ill-equipped to address. A variety of alternative strategies may be developed or pursued to address the problem of update opacity more directly, including bi-factual explanations, dynamic model reporting, and update compatibility. However, each of these strategies presents its own risks or carries significant limitations. Further research will be needed to address the epistemic and safety concerns associated with model updating and update opacity going forward.

Since the early days of the Explainable AI movement, post-hoc explanations have been praised for their potential to improve user understanding, promote trust, and reduce patient safety risks in black box medical AI systems. Recently, however, critics have argued that the benefits of post-hoc explanations are greatly exaggerated since they merely approximate, rather than replicate, the actual reasoning processes that black box systems take to arrive at their outputs. In this article, we aim to defend the value of post-hoc explanations against this recent critique. We argue that even if post-hoc explanations do not replicate the exact reasoning processes of black box systems, they can still improve users’ functional understanding of black box systems, increase the accuracy of clinician-AI teams, and assist clinicians in justifying their AI-informed decisions. While post-hoc explanations are not a “silver bullet” solution to the black box problem in medical AI, we conclude that they remain a useful strategy for addressing the black box problem in medical AI.

It is commonly accepted that clinicians are ethically obligated to disclose their use of medical machine learning systems to patients, and that failure to do so would amount to a moral fault for which clinicians ought to be held accountable. Call this ‘the disclosure thesis.’ Four main arguments have been, or could be, given to support the disclosure thesis in the ethics literature: the risk-based argument, the rights-based argument, the materiality argument and the autonomy argument. In this article, I argue that each of these four arguments are unconvincing, and therefore, that the disclosure thesis ought to be rejected. I suggest that mandating disclosure may also even risk harming patients by providing stakeholders with a way to avoid accountability for harm that results from improper applications or uses of these systems.

Recently, a growing number of experts in artificial intelligence (AI) and medicine have be-gun to suggest that the use of AI systems, particularly machine learning (ML) systems, is likely to humanise the practice of medicine by substantially improving the quality of clinician-patient relationships. In this thesis, however, I argue that medical ML systems are more likely to negatively impact these relationships than to improve them. In particular, I argue that the use of medical ML systems is likely to comprise the quality of trust, care, empathy, understanding, and communication between clinicians and patients.

A novel advantage of the use of machine learning (ML) systems in medicine is their potential to continue learning from new data after implementation in clinical practice. To date, considerations of the ethical questions raised by the design and use of adaptive machine learning systems in medicine have, for the most part, been confined to discussion of the so-called “update problem,” which concerns how regulators should approach systems whose performance and parameters continue to change even after they have received regulatory approval. In this paper, we draw attention to a prior ethical question: whether the continuous learning that will occur in such systems after their initial deployment should be classified, and regulated, as medical research? We argue that there is a strong prima facie case that the use of continuous learning in medical ML systems should be categorized, and regulated, as research and that individuals whose treatment involves such systems should be treated as research subjects.

Objectives: Machine learning (ML) has the potential to facilitate “continual learning” in medicine, in which an ML system continues to evolve in response to exposure to new data over time, even after being deployed in a clinical setting. In this article, we provide a tutorial on the range of ethical issues raised by the use of such “adaptive” ML systems in medicine that have, thus far, been neglected in the literature. Target audience: The target audiences for this tutorial are the developers of ML AI systems, healthcare regulators, the broader medical informatics community, and practicing clinicians. Scope: Discussions of adaptive ML systems to date have overlooked the distinction between 2 sorts of variance that such systems may exhibit—diachronic evolution (change over time) and synchronic variation (difference between cotemporaneous instantiations of the algorithm at different sites)—and underestimated the significance of the latter. We highlight the challenges that diachronic evolution and synchronic variation present for the quality of patient care, informed consent, and equity, and discuss the complex ethical trade-offs involved in the design of such systems.

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.

What does Artificial Intelligence (AI) have to contribute to health care? And what should we be looking out for if we are worried about its risks? In this paper we offer a survey, and initial evaluation, of hopes and fears about the applications of artificial intelligence in medicine. AI clearly has enormous potential as a research tool, in genomics and public health especially, as well as a diagnostic aid. It’s also highly likely to impact on the organisational and business practices of healthcare systems in ways that are perhaps under-appreciated. Enthusiasts for AI have held out the prospect that it will free physicians up to spend more time attending to what really matters to them and their patients. We will argue that this claim depends upon implausible assumptions about the institutional and economic imperatives operating in contemporary healthcare settings. We will also highlight important concerns about privacy, surveillance, and bias in big data, as well as the risks of over trust in machines, the challenges of transparency, the deskilling of healthcare practitioners, the way AI reframes healthcare, and the implications of AI for the distribution of power in healthcare institutions. We will suggest that two questions, in particular, are deserving of further attention from philosophers and bioethicists. What does care look like when one is dealing with data as much as people? And, what weight should we give to the advice of machines in our own deliberations about medical decisions?

In a recent Response published in the Journal of Medical Ethics,1 Buturovic provides two criticisms of my argument in ‘Is the exclusion of psychiatric patients from access to physician-assisted suicide discriminatory?’2 First, Buturovic argues that my argument effectively ‘erases the distinction between healthy adults and patients (whether somatic or psychiatric) essentially implying that PAS [physician-assisted suicide] should be available to all, for all reasons or, ultimately no reason’ (Buturovic,1 pg. 1). Second, Buturovic argues that opening the doors to medical assistance in dying (MAID) for psychiatric patients could have a number of undesirable implications. In particular, Buturovic highlights the potential negative implications for relations of trust in medicine—psychiatry in particular—along with potential effects on the rate of organ donation. I would here like to respond to these two criticisms. In short, my response to Buturovic’s first argument is that the slope is not nearly as slippery as Buturovic suggests. The reason for this is that the plausibility of Buturovic’s argument rests on a significant misinterpretation of my argument, along with an important equivocation in her own. Buturovic argues that, under the three criteria that I propose for the provision of MAID—sufficient decision-making capacity, demonstrated treatment resistance, and a lack of substantially negative implications for existing standards of psychiatric treatment and research—the provision of MAID for trivial reasons, even no reason at all, is justifiable. The main problem with this argument is that I propose no such positive criteria. My argument is that none of the three arguments addressed in my previous paper are sufficient to justify the exclusion of any and all psychiatric patients from access to MAID. I do not claim, in other words, that any individual …

Artificial intelligence (AI) is expected to revolutionise the practice of medicine. Recent advancements in the field of deep learning have demonstrated success in variety of clinical tasks: detecting diabetic retinopathy from images, predicting hospital readmissions, aiding in the discovery of new drugs, etc. AI’s progress in medicine, however, has led to concerns regarding the potential effects of this technology on relationships of trust in clinical practice. In this paper, I will argue that there is merit to these concerns, since AI systems can be relied on, and are capable of reliability, but cannot be trusted, and are not capable of trustworthiness. Insofar as patients are required to rely on AI systems for their medical decision-making, there is potential for this to produce a deficit of trust in relationships in clinical practice.

In Deep Medicine, Eric Topol argues that the development of artificial intelligence (AI) for healthcare will lead to a dramatic shift in the culture and practice of medicine. Topol claims that, rather than replacing physicians, AI could function alongside of them in order to allow them to devote more of their time to face-to-face patient care. Unfortunately, these high hopes for AI-enhanced medicine fail to appreciate a number of factors that, we believe, suggest a radically different picture for the future of healthcare. Far from facilitating a return to “the golden age of doctoring”, the role of economic and institutional considerations in determining how medical AI will be used mean that it is likely to further erode therapeutic relationships and threaten professional and patient satisfaction.

Advocates of physician-assisted suicide (PAS) often argue that, although the provision of PAS is morally permissible for persons with terminal, somatic illnesses, it is impermissible for patients suffering from psychiatric conditions. This claim is justified on the basis that psychiatric illnesses have certain morally relevant characteristics and/or implications that distinguish them from their somatic counterparts. In this paper, I address three arguments of this sort. First, that psychiatric conditions compromise a person’s decision-making capacity. Second, that we cannot have sufficient certainty that a person’s psychiatric condition is untreatable. Third, that the institutionalisation of PAS for mental illnesses presents morally unacceptable risks. I argue that, if we accept that PAS is permissible for patients with somatic conditions, then none of these three arguments are strong enough to demonstrate that the exclusion of psychiatric patients from access to PAS is justifiable.