Joshua L.M. Brand

Si certaines œuvres balzaciennes trahissent l’influence de l’esthétisme romantique, qu’il soit intime ou romanesque, ainsi dans Le Lys dans la vallée(1835), Balzac a aussi saisi les limites des romantiques et s’est tourné vers le roman réaliste. Devenant finalement le père du mouvement réaliste français, il a su créer une écriture renommée pour l’observation détaillée des milieux sociaux. Dans cet article, je me penche sur le roman moins connu de Balzac, Une fille d’Ève(1838), qui représente un angle du réalisme en abordant la problématique de la tension entre les femmes et les hommes. Mon analyse vise à faire ressortir le fait que cette tension qui a inspiré Balzac provient sans doute de la méthode Jacotot liée au roman Les Aventures de Télémaque de Fénelon.Le présent article propose d’explorer cette relation philosophique en partant des références et des pratiques intertextuelles du roman de Balzac, deux caractéristiques essentielles de l’écriture balzacienne.

Through consequentialist reasoning, Sass (2025) offers an insightful response that challenges my argument for human decision-making involving duty-right relationships, arguing that algorithmic decision-making may do a better job at optimizing outcomes beyond efficiency and accuracy, namely human-centric values like fairness and compassion. He also suggests that the humanistic motivation of my account can be supported by human-centred communication, rather than human discretion over decision processes. I respond that this misconstrues the importance of human decision-making, i.e., ensuring a shared humanity between decision-makers and those affected when constrained by duties and correlative rights understood as the moral law. This is not to bring about more fair or compassionate decisions, but is primordially important insofar as this shared humanity is a necessary condition for these moral relationships. I show this by restating my argument as transcendental. I conclude with practical implications, where algorithmically-assisted decision-making remains morally sound, yet human-centred communication in the absence of human discretion over decision processes is relegated to tasks that do not involve directed duties and their correlative rights.

This paper presents the argument that there ought to be a categorical ban on autonomous weapons systems (AWS) in warfare. First, I provide a foundational argument that international humanitarian law (jus in bello) is deontological. Following the argument shared by Peter Asaro and Robert Sparrow, I then argue that AWS lack the ability to properly acknowledge its target and consequently, breaches jus in bello principles. I, however, go further than Asaro and Sparrow by emphasizing the necessity of reciprocity for deontological law. Because AWS lack a constitutive symmetry with human combatants, humans and AWS cannot coexist in warfare if they are to respect the existing international principles. After addressing foreseeable objections, including arguments for reducing deaths and the prohibition of other weapons, I conclude that a categorical ban of AWS remains a reasonable consideration. The benefit of this paper is that it avoids complex and hypothetical considerations of future developments of AWS capabilities. It also shows that if the moral underpinnings of jus in bello principles are respected, then categorically banning AWS from warfare is already an accepted position.

This paper begins by exploring the phenomenon known as algorithm aversion, where users distrust AI and prefer human advice or decision-making even when they are aware of the algorithm’s superior performance. Current literature generally frames it as a misguided bias that harms decision accuracy and speed, likening it to a form of neo-Luddism. This view, however, overlooks the fact that the two groups (supporters and sceptics of algorithmic decisions) are speaking different moral languages: the supporters are outcome-orientated, arguing for accuracy and performance, while the sceptics offer a Kantian-position, that asks us to challenge the very precept that AI systems can be a decision-maker, or obligation-bearer, for decisions constrained by rights. To consider their position on the human-AI moral relationship, I take advantage of Korsgaard’s work on interspecies moral relationships, concluding that to be an obligation-bearer toward human right-holders, there needs to be a reciprocal reasons-giving relationship which AI in principle cannot fulfil. To meet objections that argue AI could eventually replicate the necessary moral agency requirements, I show that reciprocal relationships also call for constitutive symmetry, highlighting the importance of not only matching rationality, but also the vulnerability inherent in the human condition. With this account, algorithm sceptics are not misguided and have something morally important to say. This does not suggest eliminating AI entirely from decision processes. AI-assisted decision-making can be defended as long as a robust human-centred approach where genuine human control is upheld.