AI "Companionship"

Abstract

The use of AI-powered chatbots for “companionship” is on the rise across demographics, and the multi-billion dollar industry is actively cultivating such use cases despite widespread concern about societal impacts. But what actually constitutes “AI companionship,” and what might go wrong? We present the state of the field on “AI companionship” and unveil relationship monetization mechanisms, such as dark patterns, observed in five of the most widely used AI companion mobile applications.

Short Bio

Maribeth Rauh is a postdoctoral researcher in TCD’s AI Accountability Lab whose research focuses on the evaluation of AI systems, with particular attention to how they exacerbate structural inequality. Her current work is studying AI 'companions' for deceptive and harmful behaviour. Trained as a computer scientist and software engineer, she previously worked as a Research Engineer at Google DeepMind evaluating the risks posed by large language models. Abeba Birhane is the founder and PI of TCD’s AI Accountability Lab (AIAL). She is an assistant professor of AI at the School of Computer Science and Statistics (SCSS), with a particular research focus on AI evaluation and accountability– work for which she was featured in Wired UK and named TIME’s TIME100 Most Influential People in AI list in 2023. Dr. Birhane also served on the United Nations Secretary-General’s AI Advisory Body and currently serves at the AI Advisory Council in Ireland.

Venue

Large Conference Room, O’Reilly Institute