Ying-Tung Lin

Artificial intelligence (AI) systems are increasingly adopted to make decisions in domains such as business, education, health care, and criminal justice. However, such algorithmic decision systems can have prevalent biases against marginalized social groups and undermine social justice. Explainable artificial intelligence (XAI) is a recent development aiming to make an AI system’s decision processes less opaque and to expose its problematic biases. This paper argues against technical XAI, according to which the detection and interpretation of algorithmic bias can be handled more or less independently by technical experts who specialize in XAI methods. Drawing on resources from feminist epistemology, we show why technical XAI is mistaken. Specifically, we demonstrate that the proper detection of algorithmic bias requires relevant interpretive resources, which can only be made available, in practice, by actively involving a diverse group of stakeholders. Finally, we suggest how feminist theories can help shape integrated XAI: an inclusive social-epistemic process that facilitates the amelioration of algorithmic bias.

Imagination and other forms of mental simulation allow us to live beyond the current immediate environment. Imagination that involves an experience of self further enables one to incorporate or utilize the contents of episodic simulation in a way that is of importance to oneself. However, the simulated self can be found in a variety of forms. The present study provides some empirical data to explore the various ways in which the self could be represented in observer-perspective imagination as well as the potential limits on such representations. In observer-perspective imagination, the point of view or perspective is dissociated from the location of one’s simulated body. We have found that while there are different ways to identify with oneself in an observer-perspective imagination, the identification is rarely dissociated from first-person perspective in imagination. Such variety and limits pave the way for understanding how we identify with ourselves in imagination. Our results suggest that the first-person perspective is a strong attractor for identification. The empirical studies and analysis in this paper demonstrate how observer-perspective episodic simulation serves as a special case for research on identification in mental simulation, and similar methods can be applied in the studies of memory and future thinking.

This paper focuses on the potential of “equitech”—AI technology that improves equity. Recently, interventions have been developed to reduce the harm of implicit bias, the automatic form of stereotype or prejudice that contributes to injustice. However, these interventions—some of which are assisted by AI-related technology—have significant limitations, including unintended negative consequences and general inefficacy. To overcome these limitations, we propose a two-dimensional framework to assess current AI-assisted interventions and explore promising new ones. We begin by using the case of human resource recruitment as a focal point to show that existing approaches have exploited only a subset of the available solution space. We then demonstrate how our framework facilitates the discovery of new approaches. The first dimension of this framework helps us systematically consider the analytic information, intervention implementation, and modes of human-machine interaction made available by advancements in AI-related technology. The second dimension enables the identification and incorporation of insights from recent research on implicit bias intervention. We argue that a design strategy that combines complementary interventions can further enhance the effectiveness of interventions by targeting the various interacting cognitive systems that underlie implicit bias. We end with a discussion of how our cognitive interventions framework can have positive downstream effects for structural problems.

Every episodic memory entails a sense of identity, which allows us to mentally travel through time. There is a special way by which the subject who is remembering comes into contact with the self that is embedded in the episodic simulation of memory: we can directly and robustly experience the protagonist in memory as ourselves. This paper explores what constitutes such experience in memory. On the face of it, the issue may seem trivial: of course, we are able to entertain a sense of identity—the experience of our recollection structurally resembles our perception of the original event. However, given the phenomenon of observer memory, in which our visual perspective is decoupled from our embodied dimension, it is unclear whether it is the observing or the embodied one that is identified. This phenomenon is important not only in illustrating the complexity of identification but also in assessing how best to address it. In this paper, the issue is analyzed through concepts introduced from the literature on bodily self-consciousness. The potential approaches to addressing the issue of identification are examined, including the inheritance view, according to which the identification relies on the inheritance of mnemonic content from the original experience. I propose and argue for the self-simulation view, which suggests that what results in the experience of “I am this” in memory is the observing and the embodied dimensions as well as the relation between them, which enable different ways of projecting oneself into an episodic simulation.

Our growing understanding of human mind and cognition and the development of neurotechnology has triggered debate around cognitive enhancement in neuroethics. The dissertation examines the normative issues of memory enhancement, and focuses on two issues: the distinction between memory treatment and enhancement; and how the issue of authenticity concerns memory interventions, including memory treatments and enhancements. rnThe first part consists of a conceptual analysis of the concepts required for normative considerations. First, the representational nature and the function of memory are discussed. Memory is regarded as a special form of self-representation resulting from a constructive processes. Next, the concepts of selfhood, personhood, and identity are examined and a conceptual tool—the autobiographical self-model —is introduced. An ASM is a collection of mental representations of the system’s relations with its past and potential future states. Third, the debate between objectivist and constructivist views of health are considered. I argue for a phenomenological account of health, which is based on the primacy of illness and negative utilitarianism.rnThe second part presents a synthesis of the relevant normative issues based on the conceptual tools developed. I argue that memory enhancement can be distinguished from memory treatment using a demarcation regarding the existence of memory-related suffering. That is, memory enhancements are, under standard circumstances and without any unwilling suffering or potential suffering resulting from the alteration of memory functions, interventions that aim to manipulate memory function based on the self-interests of the individual. I then consider the issue of authenticity, namely whether memory intervention or enhancement endangers “one’s true self”. By analyzing two conceptions of authenticity—authenticity as self-discovery and authenticity as self-creation, I propose that authenticity should be understood in terms of the satisfaction of the functional constraints of an ASM—synchronic coherence, diachronic coherence, and global veridicality. This framework provides clearer criteria for considering the relevant concerns and allows us to examine the moral values of authenticity. rn