Tech Talk Tuesdays: Generative AI and Education

Can AI be creative in a way similar to humans? More specifically, can AI improvise the way a human musician can? One way to think about this is to imagine how a machine would learn to improvise. Jazz musicians learn how to improvise by practicing scales, chords, progressions, standards—and listening to the way master improvisors do it. One basic pattern for improvising is called trading fours: One musician plays four bars and another one improvises on what was just played. Good jazz musicians can trade fours while improvising each time it’s their turn, a call-and-response of varying complexity, intensity, and tone. Sometimes it seems like a conversation between two friends trading memories, nudges, jokes, and wisecracks, in other words, human interaction at its most intimate and subtle. A machine could learn the basic patterns of trading fours, but could it learn how to improvise in a way that human listeners would find engaging and entertaining?

Professor Ray Peters, Department of English and Assistant Director, University of Delaware Honors College

Ray Peters coordinates the Honors colloquia, trains and supervises Honors College Writing Fellows, coordinates the Honors College Teaching Fellows, helps students prepare applications for prestige scholarships, and coordinates the Honors College History Project. He is interested in the way the arts and science intersect. In addition to teaching Honors ENGL316, Peer Tutoring and Advanced Composition (the course required to become an Honors College Writing Fellow), he regularly—and happily—teaches three Honors colloquia: The Art of Medicine; Creating Musical Taste; and Quackery, Pseudoscience, and Conspiracy Theories. Currently, he is teaching a course entitled: Creating Musical Taste.

This project serves as a pioneering effort into the broader strategy of how emerging AI, generative AI and large language model technologies will be utilized at UD and in the greater higher education landscape. Harris and Academic Technology Services lead and collaborate with the Working Group – (26 faculty and staff members from units across the University: )— their expertise, passion and deliberate thinking — as an excellent resource to guide our work. This interdisciplinary academic work and cutting-edge research draws on the University’s expertise and resources in data science, computer science, public policy, business, engineering and more. This project will lead to a suite of learning tools that UD students can use to create their own study materials. In partnership with UD faculty, ATS is training an internal AI model to contextualize, summarize and generate materials related to UD course content from instructional video transcripts, learning management system pages and other supporting teaching documents. UDIT Academic Technology Services aims to provide personalized learning experiences to our students through the development of a platform that integrates artificial intelligence with the intellectual property of UD faculty. This ensures that our tool is tailored to our student body and serves as a reflection of UD’s courses and programs. The project marks a pioneering effort in shaping the strategic integration of emerging AI, generative AI, and large language model (LLM) technologies within the University of Delaware and the broader higher education landscape. To build the platform, the project team is utilizing foundation models from Amazon Web Services’ Bedrock, along with other AI models. Leveraging two decades of video content collected from our proprietary lecture capture system, our project team is building a private, internal platform, accessible to students through Canvas Learning Management System.

Erin Ford Sicuranza, Director, Academic Technology Services

Ms. Sicuranza is a higher education leader with a combination of instructional design and technological proficiency, able to bridge communication gaps between technical and non-technical individuals. She has a proven track record of matching effective technologies to instructional goals and business requirements. Her core competencies include instructional design; research, and writing, editing; assessment, evaluation, and survey creation; workshop design and facilitation; visual presentation design and public speaking; web-based, distance learning design. Previously, she was the manager of instructional design for the Lerner College of Business, with an emphasis on supporting online and face-to-face course design for a nationally recognized MBA program. She also designed, developed, and managed online learning modules for external University of Delaware partnerships including Christiana Care Health System and the US Olympic Committee and internally to develop online modules for UD Online, a division of Professional and Continuing Studies, and, to support the online learning management systems (Canvas and Sakai) used to deliver online courses for non-credit and credit courses. In 2009, she co-founded Springboard Careers, Inc., a niche corporation to help professional women overcome obstacles to career re-entry or redirection. She provided career development services such as professional coaching, résumé writing, and job search assistance. The company specialized in networking with area organizations with hiring needs to offer job opportunities to our audience.

As university faculty, our classes are full of students working at various skill levels. In this talk I will share ways of using generative AI to differentiate instruction, both by instructors in their planning and by students for studying purposes.

Dr. Rachel Karchmer-Klein, Associate Professor in the School of Education, and, Coordinator, M.Ed. in Teacher Leadership

Professor Karchmer-Klein teaches undergraduate, graduate, and doctoral level courses in literacy and educational technology. Her research investigates relationships among literacy skills, digital tools, and teacher preparation, with particular emphasis on technology-infused instructional design. Her overall goal is to improve preservice and practicing teachers’ knowledge of how to leverage technological affordances to support students’ literacy learning. Dr. Karchmer-Klein’s work initially focused on K–12 instructional practices and has now expanded to include research on university faculty’s use of technology in online course design and generative artificial intelligence. Her research has been published in literacy and educational technology journals such as Reading Research Quarterly, Journal of Research in Reading, Journal of Research on Technology in Education, and The Reading Teacher. She has published 30 articles/chapters and four books. Her latest, Next-Level Digital Tools and Teaching: Solving Six Major Instructional Challenges, K-12, was published by Teachers College Press last year. Her forthcoming book, AI and Disciplinary Literacy Instruction, will be published by Guilford Press in 2024.