Shared research interests bring faculty together with a CBI fellow: Meet CBI faculty hosts Hong Shen and Jodi Forlizzi

Introducing yourself and your research interests

Hong Shen is an assistant research professor in the Human-Computer Interaction Institute and Jodi Forlizzi is a professor of computer science and human and computer interation, as well as the Associate Dean for Diversity, Equity, and Inclusion in the School of Computer Science. Together, our shared research interests focus on understanding and developing tools, methods, and processes to support industry practitioners, users, and community members in responsible AI (RAI).

What about this fellowship program drew your attention?

We are excited about CBI’s focus on understanding the socio-technical aspects of AI systems. Additionally, CBI fosters close collaboration and connections across disciplines, academia, and industry, which closely align with our shared research interests.

Introducing the research topic/problem of our fellow

Qiaosi Wang is interested in integrating her dissertation research on the mutual theory of mind with responsible AI. Specifically, she plans to apply her proposed framework to guide the design of next-generation human-AI interactions, enabling AI systems to responsibly adapt to people’s changing perceptions during interactions. Qiaosi is also interested in bridging the gap between academic discourse and the real-world development of generative AI systems with theory of mind-like capabilities, with a goal to derive concrete design principles for next-generation responsible AI agents.

In your view, why is this topic important/pressing/exciting?

With the rise of generative AI, systems are increasingly demonstrating human-like behaviors through theory of mind-like capabilities. It is important to not only understand people’s perceptions of AI but also to develop tools, methods and processes that enable AI systems to responsibly adapt to these evolving perceptions. Qiaosi’s framework, rooted in the fundamental human capability of theory of mind, provides valuable human-centered insights for designing AI systems that can effectively augment human decision-making, collaboration, and interaction.

Introducing your Postdoc Fellow

Qiaosi earned her Ph.D. from the Georgia Institute of Technology, where she developed the framework of mutual theory of mind for human-AI communication in the context of socialmatching AI agents in online learning. Using a mixed-methods approach, she conducted studies that made significant theoretical, design, and technical contributions to understanding human perceptions of human-like AI agents. Qiaosi is also passionate about advancing responsible AI practices in the industry by collaborating with AI practitioners to address challenges and identify needs in designing responsible, human-centered, user-facing AI applications.

What about your postdoc fellow makes them the right person to tackle this topic?

Qiaosi’s work is interdisciplinary in nature, situated at the intersection of cognitive science,human-AI interaction, and responsible AI. With her background in psychology, HCI, and informatics, she’s uniquely positioned to bridge theoretical insights and practical applications, fostering a deeper understanding of human-AI dynamics in the age of generative AI. Readers who are interested in reading more about Qiaosi’s application of mutual theory of mind in responsible AI can take a look at her paper from the 2021 Conference on Human Factors in Computing Systems.