PhD Defence Kristy Claassen | Being Human | Evaluating Artificial Intelligence Through The Lens of Ubuntu

Being Human | Evaluating Artificial Intelligence Through The Lens of Ubuntu

Kristy Claassen is a PhD student in the department Philosophy. (Co)Promotors are prof.dr. C. Aydin and dr. J.S. Hermann from the Faculty of Behavioural, Management and Social Sciences (BMS), University of Twente and prof.dr.ir. P.P.C.C. Verbeek University of Amsterdam and prof.dr. W.A. IJsselsteijn form the Technische University of Eindhoven.

In this dissertation, I explore what it means to be human in the age of Artificial Intelligence (AI) by examining AI through the lens of Ubuntu. Being Human: Evaluating Artificial Intelligence Through the Lens of Ubuntu argues that the quest to make AI ethical cannot be separated from the question of what it means to be human together. The dissertation’s title is reflected in its conclusion: being (fully) human, in the Ubuntu sense, is a communal, relational process, and AI should be evaluated by how it supports or hinders that process. The work’s primary argument is that applying Ubuntu as an evaluative lens reveals hidden ethical dimensions of AI and uncovers conceptual disruptions. By engaging with AI through a heuristic of Ubuntu, this dissertation broadens the theoretical discourse by showing that we perceive and experience technology (postphenomenologically) as socially embedded human beings. It also employs an Interpretative Phenomenological Approach (IPA) as a lens to identify which values are prioritised by the users of AI, how these values are articulated and negotiated, and which of these values are being disrupted.

          In order to arrive at these insights, this dissertation is divided into two parts. Part I (Chapters 1–3) focuses on the theoretical foundations of the study. Chapter 1 discusses the ontological, normative and critical approaches to the essentially contested concept of Ubuntu. Chapter 2 employs a narrative literature review to provide a situated snapshot of current scholarship that engages Ubuntu in relation to the philosophy and ethics of technology. It shows that this body of literature remains concentrated in a limited number of application domains, that questions concerning how humans relate to technologies are often confined to specific domains, and that significant conceptual and methodological gaps remain that require further attention. Key normative and ontological foundations were revisited to formulate an Ubuntu heuristic that guides the dissertation. Chapter 3 brings Ubuntu into conversation with postphenomenology. It argues that the social dimension of human–technology relations is underemphasised in contemporary postphenomenological debates. Drawing in particular on the phenomenological work of Alfred Schütz, the chapter shows that human beings are socially embedded and that experiences and perceptions of technology do not occur in isolation from others. Ubuntu is presented as exemplifying this emphasis on social embeddedness, thereby serving as a conceptual bridge between Ubuntu philosophy and postphenomenology.

Part II of this dissertation shifts from the theoretical to the empirical and builds upon these empirical findings to develop new insights. Building upon the previous findings, Chapter 4 integrates empirical research to examine how AI is experienced and evaluated within Ubuntu-practising contexts. By implementing the method of AI Value-Mapping, participants identified which AI-related values they considered central, which they experienced as disrupted, and which they perceived as missing from current discourse. Values such as privacy and human dignity emerge as central, while values like creativity and integrity are underexplored or missing from the mainstream discourse. This chapter, therefore, serves as a lens through which the values that warrant further conceptual analysis are identified.

          The remaining chapters explore these values in depth. In Chapter 5, privacy is analysed as a value-concept subject to intercultural conceptual disruption, demonstrating how dominant frameworks fail to account for relational understandings of privacy rooted in duties and communal well-being. It argues for a broadening of current theories of conceptual disruption that acknowledges that disruption does not take place in a vacuum. It further argues for a theory of Intercultural Conceptual Disruption that calls for the decentering or decolonisation of particular minority world value-concepts. Chapter 6 investigates the concept of Human Dignity.  Ubuntu-practising participants in the focus group emphasised that human dignity underlies all other values and positioned it as the most important value in the Value-mapping exercise. This chapter explores the meaning of Human Dignity in an Ubuntu context and argues that generative AI poses a risk for attaining personhood, the basis for Human Dignity. Finally, Chapter 7. Missing Values: Creativity, Integrity and Diversity is the final and most exploratory chapter of this dissertation. These were values that participants in the empirical research identified as underexplored or missing. This chapter thus explores these values in more depth and calls for further exploration into these topics.

          Overall, the dissertation contributes theoretically, methodologically, and normatively to the philosophy and ethics of AI. It advances Ubuntu scholarship by bringing it into conversation with postphenomenology, engages with contemporary themes in AI ethics and demonstrates the value of integrating empirical methods into philosophical methods. By doing so, it argues for an Ubuntu-informed approach to AI ethics that takes seriously the question of what it means to be human in a technologically mediated world.