In Frontiers in big data
Shannon Vallor has raised the possibility of ethical deskilling as a potential pitfall as AI technology is increasingly being developed for and implemented in military institutions. Bringing the sociological concept of deskilling into the field of virtue ethics, she has questioned if military operators will be able to possess the ethical wherewithal to act as responsible moral agents as they find themselves increasingly removed from the battlefield, their actions ever more mediated by artificial intelligence. The risk, as Vallor sees it, is that if combatants were removed, they would be deprived of the opportunity to develop moral skills crucial for acting as virtuous individuals. This article constitutes a critique of this conception of ethical deskilling and an attempt at a reappraisal of the concept. I argue first that her treatment of moral skills and virtue, as it pertains to professional military ethics, treating military virtue as a sui generis form of ethical cognition, is both normatively problematic as well as implausible from a moral psychological view. I subsequently present an alternative account of ethical deskilling, based on an analysis of military virtues, as a species of moral virtues essentially mediated by institutional and technological structures. According to this view, then, professional virtue is a form of extended cognition, and professional roles and institutional structures are parts of what makes these virtues the virtues that they are, i.e., constitutive parts of the virtues in question. Based on this analysis, I argue that the most likely source of ethical deskilling caused by technological change is not how technology, AI, or otherwise, makes individuals unable to develop appropriate moral-psychological traits but rather how it changes the institution's capacities to act.
Hovd Sigurd N
2022
AI, automation, deskilling, extended cognition, institution, skill, virtue ethics
