Morgan Luck

In a recent paper, Sparrow and Flenady (AI Soc 40:5285–5296, 2025) argue that the use of GenAI in tertiary education should be resisted. Sparrow and Flenady give many reasons for such resistance, but they say that ‘[t]he most important reason to resist the use of AI in universities is that its outputs are fundamentally bullshit…’. In this paper, we take issue with this claim. We argue that (a) GenAI cannot bullshit, and (b) bullshit is often compatible with the goals of tertiary education. The upshot is that we need to look at other considerations besides whether the outputs of GenAI are bullshit to determine whether and how to integrate GenAI into university education.

Sometimes actions that are (morally) impermissible in the actual world also seem impermissible in virtual worlds. For example, rape is impermissible in the actual world, and it also seems impermissible to direct a videogame character to rape another in a virtual world. But sometimes actions that are impermissible in the actual world seem permissible in virtual worlds. For example, stealing is impermissible in the actual world, but it may seem permissible to direct your video game character to steal another character’s possessions in a virtual world. This asymmetry has complicated the task of finding the line between permissible and impermissible gameplay. So, how should we play video games while this line remains unclear? Is there a way of playing games that ensures our actions are permissible? In this paper we consider a promising moral principle that might guarantee permissible gameplay.

Extremism often carries “an unmissable and inescapable negative connotation” (Cassam 2021, p.34) and has become “an increasingly convenient insult” (Berger 2018, p.2). But what exactly is wrong with being an extremist? One answer is that the wrongness of being an extremist comes, at least in part, from the extremeness of their beliefs. I assess two arguments, offered by Clarke (2019) and Coady (2024), which reject this view and go further by suggesting that being an extremist is never, in itself, wrong. Drawing on two contrasting accounts of what it means for a belief to be extreme (the edge and outlier accounts) I reconstruct Clarke and Coady’s arguments and explore their implications. I conclude that, although their defence of extremism overreaches, there are nonetheless reasons to challenge the idea that holding extreme beliefs is wrong.

This paper is about the workplace surveillance of a particular class of employees: professionals. Professionals have professional obligations. We identify four different ways in which employers' use of workplace surveillance can make it difficult for professionals to fulfil professional obligations. We argue that when employers proceed in these ways they violate the principle of unhindered professionalism, which states that employers ought not to significantly hinder the ability of their professional employees to fulfil professional obligations. We finish up by identifying several constraints on morally permissible uses of workplace surveillance of professionals that follow from the principle of unhindered professionalism.

Artificial moral patients (or AMPs) are those things successfully made to resemble moral patients, but are not. They are artificial both in the sense that they are made by us (artefacts), and that they are not a real instance of what they are made to resemble (artifice). ChatGPT, sex dolls, social robots, and non-player characters are all examples of AMPs. As these technologies start to resemble humans with greater accuracy the question as to how we should treat them becomes increasingly important. We consider whether work on the Gamer's Dilemma, a puzzle concerning virtual wrongdoings, may provide a useful framework for addressing this question.

One response to the existential threat posed by a super-intelligent AI is to design it to be friendly to us. Some have argued that even if this were possible, the resulting AI would treat us as we do our pets. Sparrow (AI & Soc. https://doi.org/10.1007/s00146-023-01698-x, 2023) argues that this would be a bad outcome, for such an AI would dominate us—resulting in our freedom being diminished (Pettit in Just freedom: A moral compass for a complex world. WW Norton & Company, 2014). In this paper, I consider whether this would be such a bad outcome.

Coghlan and Cox (Between death and suffering: Resolving the gamer’s dilemma. Ethics and Information Technology) offer a new resolution to the Gamer’s Dilemma (Luck, The Gamer’s Dilemma. Ethics and Information Technology). They argue that, while it is fitting for a person committing virtual child molestation to feel self-repugnance, it is not fitting for a person committing virtual murder to feel the same, and the fittingness of this feeling indicates each act’s moral permissibility. The aim of this paper is to determine whether this resolution – the repugnant resolution – successfully resolves the Gamer’s Dilemma. We argue that it does not.

The Gamer’s Dilemma (Luck, 2009a) is a paradox concerning the moral permissibility of two types of acts performed within computer games. Some attempt to resolve the dilemma by finding a relevant difference between these two acts (Bartel, 2012; Patridge, 2013; Young, 2016; Nader, 2020; Kjeldgaard-Christiansen, 2020; and Milne & Ivankovic, 2021), or to dissolve the dilemma by arguing that the permissibility of these acts is not as they seem (Ali, 2015; Ramirez, 2020). More recently some have attempted to resist the dilemma by undermining the intuitions supporting it (Montefiore & Formosa, 2022; Formosa et al., 2023). The purpose of this paper is to consider what might follow from such resistance.

One response to the existential threat posed by a super-intelligent AI is to design it to be friendly to us. Some have argued that even if this were possible, the resulting AI would treat us as we do our pets. Sparrow (AI & Soc. https://doi.org/10.1007/s00146-023-01698-x, 2023) argues that this would be a bad outcome, for such an AI would dominate us—resulting in our freedom being diminished (Pettit in Just freedom: A moral compass for a complex world. WW Norton & Company, 2014). In this paper, I consider whether this would be such a bad outcome.

Montefiore and Formosa (Ethics Inf Technol 24:31, 2022) provide a useful way of narrowing the Gamer’s Dilemma to cases where virtual murder seems morally permissible, but not virtual child molestation. They then resist the dilemma by theorising that the intuitions supporting it are not moral. In this paper, I consider this theory to determine whether the dilemma has been successfully resisted. I offer reason to think that, when considering certain variations of the dilemma, Montefiore and Formosa’s theory may not be the most likely theory available to us.

Purpose This paper aims to evaluate a potential resolution to the gamer’s dilemma that arises from Gary Young’s metaethical theory of constructive ecumenical expressivism (CEE). Design/methodology/approach In this paper, the gamer’s dilemma is reformulated as a paradox and the potential resolution is evaluated in light of this new formulation. Findings The author argues that this resolution does resolve the dilemma, but CEE itself has limited appeal. Originality/value This paper contributes to the growing scholarship dedicated to resolving the gamer’s dilemma.

abstract Intuitively, all killings are equally wrong, no matter how old one's victim. In this paper we defend this claim — The Equal Wrongness of Killings Thesis — against a challenge presented by Kasper Lippert-Rasmussen. Lippert-Rasmussen shows The Equal Wrongness of Killings Thesis to be incompatible with two further theses: The Unequal Wrongness of Renderings Unconscious Thesis and The Equivalence Thesis. Lippert-Rasmussen argues that, of the three, The Equal Wrongness of Killings Thesis is the least defensible. He suggests that the most convincing considerations apparently in favour of the Equal Wrongness thesis may be satisfied just as well if we adopt an alternative principle, a 'Prioritarian View' about the wrongness of killing. We argue that The Prioritarian View does not resolve the trilemma: it too is inconsistent with the other two theses. Instead, we argue, the most plausible resolution of the trilemma involves a rejection, rather, of The Unequal Wrongness of Renderings Unconscious Thesis. In its place, we offer an attractive principle that is compatible with both The Equal Wrongness of Killings Thesis as well as The Equivalence Thesis.

This is a good book. It is good because: (a) it outlines well the central arguments of the debate (that is, the arguments relating to what a miracle is, whether they are possible, whether we can have evidence of their occurrence, and what would follow from such evidence were we to have it); (b) it furthers the debate; and (c) it is a clearly written. If you are a philosopher religion whose research area is miracles, the book is a must-read. If you are philosopher of religion whose research area is not miracles, then, as the book connects with core topics in this field, it will prove a valuable read. If you are merely a philosopher, there are far worse things you could be reading.Rather than provide a summary of each chapter of the book (Larmer provides such a summary on pages 3 and 4 of his introduction [which one can access freely online] and I cannot offer better), I will outline some of the book’s arguments. The type of arguments I focus on are not explicitly formulated by Larmer, t ..

Most people agree that murder is wrong. Yet, within computer games virtual murder scarcely raises an eyebrow. In one respect this is hardly surprising, as no one is actually murdered within a computer game. A virtual murder, some might argue, is no more unethical than taking a pawn in a game of chess. However, if no actual children are abused in acts of virtual paedophilia (life-like simulations of the actual practice), does that mean we should disregard these acts with the same abandon we do virtual murder? In this paper I shall outline several arguments which attempt to permit virtual murder, whilst prohibiting virtual paedophilia

In a recent paper in Religious Studies, Clarke criticizes Mumford's definition of a miracle as it fails to recognize a supernatural agent capable of intent. Clarke believes that in order for an event to qualify as a miracle a supernatural agent must intend it. It is my aim to dismiss this qualification and demonstrate how Mumford's intent-neutral definition is less problematic. I will do this by examining each of the three cases against Mumford's definition and give reason to reject Clarke's criticism and his own definition of a miracle

In his book The Concept of Miracle and his paper ‘For the Possibility of Miracles’ Swinburne claims that there are no logical difficulties in supposing that there could be strong historical evidence for the occurrence of miracles. This claim is based on three assertions; two of which I demonstrate are only true contingently. In this paper I identify several logical difficulties regarding the possibility of attaining historical evidence for the occurrence of miracles. On the strength of these logical difficulties I hope to demonstrate that there is sufficient reason to doubt Swinburne’s central claim.

In this paper I shall assess Clarke’s assertion that all definitions of miracles that purport to satisfy the criterion of religious inclusiveness should substitute the term ‘supernatural’ for ‘non-natural’. In addition, I shall attempt to strengthen Clarke’s conception of the supernatural by offering an analysis of what it means for something to be ‘above’ nature. Lastly, I shall offer a new argument as to why Clarke’s intention-based definition of miracles is necessarily less religiously inclusive than Mumford’s causation-based definition.