By Louis DiPietro
On


When’s the last time you cooked a roast in a microwave?

In the late 1970s and early 1980s, the microwave was considered a possible replacement to ovens. Eventually – after untold quantities of doomed beef, rubbery eggs, and soggy bacon – we came to understand what microwaves can and can’t do.

That may be where we are today with generative artificial intelligence (AI) tools in the teaching and learning environment, according to Steve Jackson, vice provost for Academic Innovation at Cornell and professor of information science.

“I use my microwave every day. It hasn’t gone away, but I still have an oven,” Jackson told the more than 200 attendees at the panel, “AI + Education: Teaching and Learning in the Age of AI,” held Saturday, Oct. 25, in the newly dedicated Computing and Information Science Building during the 75th Trustee-Council Annual Meeting. “I think that's where we're going to land [with AI], and our effort to work through this experimentally, through research and in an evidence-based way, is how we’ll get there. I think in five years, we'll be using it more and less, but we’ll be using it in the right places.” 

David Mimno, professor and chair of the Department of Information Science, and Jadon Geathers, a doctoral student in the field of information science, joined Jackson on the panel, which was hosted by Cornell Bowers and moderated by Claire Cardie, the John C. Ford Professor of Engineering in the Departments of Computer Science and Information Science and the Cornell Bowers associate dean for education. 

A color photo showing 3 men sitting on a stools, on a stage.

Luke Stewart

Held as part of the 75th Trustee-Council Annual Meeting weekend, the AI + Education panel featured (from left) Steve Jackson, vice provost for Academic Innovation at Cornell and professor of information science, David Mimno, professor and chair of the Department of Information Science, and Jadon Geathers, a doctoral student in the field of information science.

Together, the trio painted a more complete picture of the state of AI in education at Cornell. Jackson outlined AI tools currently in use across the university, shared results of a recent Cornell-led survey of faculty and students about their use of AI – the largest single institutional survey in the U.S. – and offered examples of the ways that faculty are “leaning in or leaning out” with AI in the classroom. Mimno offered a view into how faculty are approaching student use of AI and the ways students are learning coding in the age of AI, and Geathers showed how AI impacts research by demoing ChitterChatter, his AI-powered tool that reimagines foreign language learning. 

AI is not all that new, and its development is grounded in federal research, panelists said. The field dates to the 1950s and advanced through the 20th century – from the age of knowledge representation in the 1980s and 1990s, Deep Blue, and IBM Watson, up to today’s generative AI like ChatGPT.

“This didn't jump out of any one individual's brain, unformed or unsupported,” Jackson said, pointing to a slide charting AI’s development over the last 70 years. “This is a story of federal research. No federal research, no AI.” 

Mimno characterized AI as the third part of a “triple whammy” that today’s students are experiencing, after growing up with the ubiquity of smartphones and learning through academic disruptions from coronavirus.  

“No matter how motivated students are outside of the classroom, if there’s an easy button, someone’s going to press it,” Mimno said. 

Adapting to AI requires recognizing what our actual educational goals are, Mimno said. Quoting a doctoral student, he suggested that in programming, students are not just learning the syntax of a computer language, but an entirely new way to think. 

Geathers’ ChitterChatter offers learners a low-stakes platform to work on a new language. AI, in this case, is helping improve human-to-human interactions, he said.  

“We all want to learn a language, but when it comes time to engage with people, we're faced with barriers – anxiety, discomfort, things that prevent us from being able to engage naturally,” Geathers said. “Sometimes we hear the idea that AI is preventing students from communicating and thinking fully, but in this case, AI is actually helping to reduce many of the barriers that prevent students from wanting to engage in the first place.”
 

Louis DiPietro is a writer for the Cornell Ann S. Bowers College of Computing and Information Science.