Artificial virtuous agents: from theory to machine implementation

Abstract

Virtue ethics has many times been suggested as a promising recipe for the construction of artificial moral agents due to its emphasis on moral character and learning. However, given the complex nature of the theory, hardly any work has de facto attempted to implement the core tenets of virtue ethics in moral machines. The main goal of this paper is to demonstrate how virtue ethics can be taken all the way from theory to machine implementation. To achieve this goal, we critically explore the possibilities and challenges for virtue ethics from a computational perspective. Drawing on previous conceptual and technical work, we outline a version of artificial virtue based on moral functionalism, connectionist bottom–up learning, and eudaimonic reward. We then describe how core features of the outlined theory can be interpreted in terms of functionality, which in turn informs the design of components necessary for virtuous cognition. Finally, we present a comprehensive framework for the technical development of artificial virtuous agents and discuss how they can be implemented in moral environments.

Artificial virtuous agents in a multi-agent tragedy of the commons

“…It draws on a combination of insights from previous work, in particular the weight analogy of Thornton et al. (2016), the classification method of Guarini (2006), the "dispositional functionalism" explored by Howard and Muntean (2017), and the eudaimonic approach to "android arete" proposed by Coleman (2001) and Stenseke (2021). Essentially, the model is based on the idea that dispositional virtues can be functionally carried out by artificial neural networks that, in the absence of a moral exemplar, learn from experience in light of an eudaimonic reward by means of a phronetic learning system.…”

“…Learning from moral exemplars could also be achieved given that the agent has (i) some way of identifying and adopting suitable exemplars and (ii) some way of learning from them. Besides the solutions offered by Govindarajulu et al (2019) and Berberich and Diepold (2018), Stenseke (2021) has suggested that exemplars can be identified through e-type and e value. Specifically, an agent a takes another agent b as a moral exemplar iff (1) a and b share the same e-type and (2) b has a higher e-value than a.…”

“…(4) The fourth theme is based on the relationship between virtue ethics and connectionism, i.e., the cognitive theory that mental phenomena can be described using artificial neural networks. The emphasis on learning, and the possibility to apprehend context-sensitive and non-symbolic information without general rules, has indeed led many authors to highlight the appeal of unifying virtue ethics with connectionism (Berberich and Diepold 2018;Wallach and Allen 2008;Howard and Muntean 2017;Stenseke 2021). The major reason is that it would provide AVAs with a compelling theoretical framework to account for the development of moral cognition (Churchland 1996;DeMoss 1998;Casebeer 2003), as well as the technological promises of modern machine learning methods (e.g., deep learning and reinforcement learning).…”

“…Virtue ethics has repeatedly been suggested as a promising blueprint for the creation of artificial moral agents (Berberich and Diepold 2018;Coleman 2001;Gamez et al 2020;Howard and Muntean 2017;Wallach and Allen 2008;Mabaso 2020;Sullins 2021;Navon 2021;Stenseke 2021) 4 . Beyond deontological rules and consequentialist utility functions, it presents a path to construe a more comprehensive picture of what it in fact is to have a moral character and be a competent ethical decision maker in general.…”

“…In this paper, we refine and extend the technical details of a conceptual model (Stenseke 2021) and present the first experimental implementation of AMAs that solely focuses on virtue ethics. The experimental results show that our AVAs manage to tackle cooperation problems while exhibiting core features of their theoretical counterpart, including moral character, dispositional virtues, learning from experience, and the pursuit of eudaimonia.…”