AI Reshapes Learning: ODU Scholars Discuss Generative Technology's Impact

"Striking the Balance: AI Automation and Human Interaction in Teaching and Learning," a symposium hosted on September 8 by ODU President Brian O. Hemphill, Ph.D., brought together Old Dominion University subject experts to discuss how breakthroughs with generative AI could shape the future of education and create a new generation of learners who've developed skills working with these models.  

Panelists for the symposium included: 

• Helen Crompton, executive director of the Research Institute for Digital Innovation in Learning, Division of Digital Learning; professor of instructional technology; and director of the Virtual Reality Lab 

• Khan Iftekharuddin, professor, Batten College of Engineering and Technology, and director of the Vision Lab 

• Chrysoula Malogianni, associate vice president for digital innovation, Division of Digital Learning 

• Jian Wu, assistant professor, Department of Computer Science 

They gathered in front of a large online and in-person audience at the Ted Constant Convention Center to discuss the problems and solutions offered by generative AI.  

Crompton referred to AI as a "magical resource" that can provide a partner for a student and inform critical thinking and problem solving. She believes AI follows past innovations like the Gutenberg printing press and the internet and search engines as inflection points in technology that humanity harnessed to enhance education. As an example, she pointed to how AI can advance tutoring and create individualized plans for each student.  

"If you get a question right (through tutoring), maybe the AI can change the next question to be a little more advanced," Crompton said. "It allows the teacher to do a better job  meeting a student's needs."  

Added Malogianni, "We can have AI models play the role of Plato or Einstein and provide experiences that were unimaginable before." She also cited the potential for a more inclusive and equitable environment that uses AI to provide closed captioning and visualizations to help learners with specific needs. Data analysis of a student population could reveal trends or needs that a human could overlook.  

Iftekharuddin noted AI could deliver personalized AI experiments to children with autism. He also argued that such collaborative applications can expand access despite remote barriers.  

For grading and assessment, AI tools can help by reducing faculty workload, adding consistency and detecting plagiarism or cheating, the panel discussed. However, the panel cautioned that these systems aren't perfect and shouldn't be wholly relied upon without human oversight. Transparency around how these tools evaluate student work is essential.  

The panel also pointed to the necessary regulation and implementation of AI. Institutions play a vital role in preparing faculty and students to ethically use AI through new policies, building competencies around AI literacy and understanding security/bias issues and education on AI's capabilities. They advocated for developing thoughtful policies and governance while continuing to educate campus communities on AI as this technology advances.  

“ChatGPT doesn’t tell you where to look,” Crompton said. “It tells you the answer. We need to think about what this is going to mean. It’s exciting but we’re going to need to navigate it in the correct way.”