Erica Kleinman

Research Vision

Computational systems, ranging from artificially intelligent assistants to data visualizations, become more common in the lives of end-users in the forms of tools such as AI coaches for esport games, data trackers in health applications, and voice activated assistants in their homes and phones. As these computationally assistive systems become more common in day to day use, it becomes increasingly important that we understand how end users interact with these systems and how their lives are impacted by that interaction. This is important to both create systems that meet end users’ needs more efficiently and also to remain ethical in the advancement of research and development of these tools.

I hope to make an impact in this emerging and highly relevant area. Through my research, I explore the interplay between computational assistants and learning. In educational contexts, intelligent systems are seeing increased use to create more personalized and accessible learning opportunities. Learning is not limited to academic contexts, however. Learning can also apply to acquisition of knowledge of how to perform a task or routine. Further, such data-driven, intelligent tools are seeing continued use as support beyond the learning phase, helping users pursue tasks and maintain optimal performance. For example, in health and fitness applications, data visualizations are employed to help users make informed decisions about their routines.

Especially in high-impact domains, understanding the impact these systems have on users, and how we can design them to meet their needs without doing harm, is critical to success. I chose to explore these questions in the context of games not only to make the benefits of complex gameplay more accessible to more people, but also because games, especially esports, provide an opportunity to study how users interact with computational tools and how they are impacted by that interaction in dynamic, unpredictable, high-stakes contexts. Findings collected from such a scenario may transfer to similar situations, such as disaster relief, and provide valuable and precious insights into the development of AI tools to aid users. My work has also identified opportunities to improve the “gamification” of non-game systems and potentially improve their functionality.