On Wednesday, February 21, at the Tang Institute’s February Lunch & Discussion, institute director Andy Housiaux introduced Dr. Lydia Cao to more than 50 members of the Andover community — a group of faculty, staff, and students who’d gathered to listen to her talk about learning in the age of generative AI [artificial intelligence]. A leading scholar in education and technology, Cao is also a former teacher whose research is partly driven by how educators will be able to use AI in their classrooms.
“Dr. Cao is going to invite us to think about fundamental questions related to schooling and education,” Housiaux said. “What does it mean to learn? What is worth knowing? These are central questions at any time in the life of a school, but especially so for a school like this in a time of generative AI.”
Cao structured her talk around three key metaphors of learning:
- learning as acquisition: to see the mind as a container, and the job of learning to get what’s out there in the world deposited into the mind
- learning as participation: to focus on activities, discourse, and interactions, in order to become part of the community
- learning as knowledge creation: to see knowledge as improvable and progressive
She dug deeply into this last metaphor. “As schools,” she said, “we often invest in the learning business. We are not investing in the knowledge creation business. We think that’s for scientists and people who work in the real world. But this learning metaphor is challenging that perception, by saying ‘How can we transform schools into knowledge creation institutions, so we are not just learning what’s already there in the world, but also helping to improve it?’
“Why knowledge creation is so relevant in the world of AI,” she continued, “why this metaphor is so important right now, is because knowledge acquisition and participation may not meet the needs of knowledge societies anymore. There are pressing challenges in our world that need innovative solutions that require us to constantly create new knowledge and find new ways of doing things….”
Claire Gallou, instructor in French at Andover, appreciated how Cao organized these concepts and put words on ideas that had been floating around in Gallou’s pedagogy without being named, such as the idea of “increasing difference” in order to create a favorable environment for knowledge creation. “Organizing and naming these concepts and values will allow me to use them more intentionally,” Gallou said.
Cao covered a lot of ground in an hour, referencing feedback models, Harvard University’s Graduate School of Education’s Project Zero, ways to partner with AI, plagiarism, motivation as an ecosystem, and much more. Nicholas Zufelt, instructor of mathematics, statistics, and computer science at Andover, said Cao’s talk was a fascinating blend of current research and practical actions that teachers can take. “I loved the generalized and wide-ranging examples that challenged the oversimplification I sometimes see, that ‘generative AI = chatGPT.’ AI is so much more than just text generation.”
He left the gathering with intention. “The most actionable advice for me is twofold,” Zufelt said. “First, to reframe the sticking points of including AI in the classroom using a different metaphor of learning. If something isn’t working, try looking through a different lens at what learning can be, and how does that help me to see a new solution? Second, to try to explore wholly different methodologies of AI, such as AI-generated transcriptions and summaries of meetings with students.”
Near the end of her talk, Cao referred to AI as a kind of mirror that reflects what our society believes and thinks. Gallou was struck by this. “One of my concerns,” Gallou said, “is that if we start using AI to generate content that we can then analyze critically, we end up focusing on AI and its production, as opposed to the real world and people. Cao made me realize that what AI generates is really a reflection of the world and its people, in the data it provides and in the subjectivity of its programmers. Therefore, analyzing AI production is much more about humans and the world than I thought.”
Cao’s hour-long talk succeeded in sparking interest, ideas, and curiosity in the Andover community. And as generative AI continues to evolve and take new shapes, the conversation about it at the Tang Institute will as well.
