The more we can learn from each other the better, says Sara Deveaux, the director of the Kravis Center for Excellence in Teaching.  

That maxim has always been accepted in the teaching world, but with the advent of generative artificial intelligence tools like ChatGPT, the stakes seem higher. No one knows exactly where AI will lead, so trying to figure out the relationship between AI and education needs as many minds as possible. 

“We have to work together on this,” Sara said. “Sharing wisdom is so important.”  

Sharing wisdom is exactly what took place on Monday, December 4, as Loomis Chaffee and the Kravis Center hosted their second AI symposium for independent schools with about 85 people from 35 schools attending. LC held the previous symposium in April.   

“If there is something I have learned in the year-plus this technology has been in the public conversation, it is that there is no shortage of surprising ways you can use the tool and no shortage of things to be aware of,” said Assistant Director of the Kravis Center for Academic Technology Matt Johnson. “It is important we do not silo ourselves at this moment, that in fact we live our mission: best self, common good. I think Loomis Chaffee is trying to be our best self in response to this technology and also recognizes that we play a role in serving the common good of our peer schools.” 

The symposium was a combination of case studies and workshops facilitated by faculty and others from Loomis Chaffee and peer schools. Topics ranged from “Common Ground amid Shifting Norms: Fostering Community Discussion about AI Literacy and Academic Integrity” to “Supporting, Not Supplanting, Student Thought” and “Working with AI Interns.” 

The symposium fits in with the Kravis Center’s commitment to excellence. 

“This is in line with the Kravis Center as a center that reaches out,” Matt said. “We have so many creative and talented and thoughtful teachers that it would be a crime not to share this knowledge within our school community, [and] it would be a missed opportunity not to open our doors and invite people in to learn from what we’re doing. …  We do not have all the answers, and that is partially why we are inviting people in: so we can get better, we can learn from them.” 

Ned Heckman, a science teacher and an associate director of the Kravis Center, presented on using an AI intern. Before his session, he said, “Just like we have human interns that assist us to work more productively and, candidly, take work off our plates that is not all that important, the work of an AI intern is analogous. The idea is to train large-language models and their chat functions to either do work in collaboration with you toward a goal or just delegate a task — data analysis, for example.” 

Ned talked about having an AI tool analyze the previous lesson or homework and suggest openings for the next class. He had to tweak the suggestions, and that is one of the concerns. How much can one trust AI?  

That question also came up at a panel in which five Loomis students talked about AI and took questions from faculty at other schools. The discussion also raised the issue of how to use AI responsibly while maintaining academic integrity. One student described AI as both exciting and scary. This sentiment is not much different from what the rest of the world thinks, and it’s another good reason for convening a symposium such as this one.         

The discussion was thoughtful, and as one faculty member from a peer school remarked on the way out, “Great kids.”