The first half of this presentation provides an update on my recent work on computer agents in learning and assessment environments. This includes a version of AutoTutor that helps adults learn comprehension skills, electronics, and various subject matters. These systems have three-party conversations, called trialogues, where two agents (such as a tutor and a peer) interact with the adult. This work is compatible with the intelligent tutoring system architecture of the Generalized Intelligent Framework for Tutoring (GIFT, gifttutoring.org). The second half of the presentation provides observations, reflections, and questions about the use of Generative AI (large language models, ChatGPT) in the development of learning and assessment applications with conversational agents. I will comment on some current projects of other researchers that are exploring how Gen-AI can enhance the development, quality, and scope of conversation-based learning and assessment.
Biosketch
Art Graesser is a professor in the Department of Psychology and the Institute of Intelligent Systems at the University of Memphis, as well as an Honorary Research Fellow at University of Oxford. He received his Ph.D. in psychology from the University of California at San Diego. His research interests question asking and answering, tutoring, text comprehension, inference generation, conversation, reading, problem solving, memory, emotions, artificial intelligence, computational linguistics, and human-computer interaction. He served as editor of the journal Discourse Processes (1996–2005) and Journal of Educational Psychology (2009-2014), as well as presidents of 4 societies, including Society for Text and Discourse (2007-2010), the International Society for Artificial Intelligence in Education (2007-2009), and the Federation of Associations in the Behavioral and Brain Sciences (2012-13). He and his colleagues have developed and tested software in learning, language, and discourse technologies, including those that hold a conversation in natural language and interact with multimedia (such as AutoTutor) and those that analyze text on multiple levels of language and discourse (Coh-Metrix and Question Understanding Aid — QUAID). He has served on five panels with the National Academy of Sciences and four OECD expert panels on problem solving, namely PIAAC 2011 Problem Solving in Technology Rich Environments, PISA 2012 Complex Problem Solving, PISA 2015 Collaborative Problem Solving (chair), and PIAAC Adaptive Problem Solving 2021.
