James Lester, North Carolina State University, shares more about his NSF AI Institute for Engaged Learning: EngageAI (#2112635).
Team: James Lester (Principal Investigator), Gautam Biswas (Co-Principal Investigator), Jeremy Roschelle (Co-Principal Investigator), Cindy Hmelo-Silver (Co-Principal Investigator), Mohit Bansal (Co-Principal Investigator)
Driven by a vision in which AI supports and extends the intelligence of teachers and learners, the NSF AI Institute for Engaged Learning is designing, developing, and investigating AI-driven narrative-centered learning environments that create engaging story-based, collaborative problem-solving experiences.
What is the big idea of your project?
The AI Institute for Engaged Learning emerged from a recognition of the significant potential of interactive narrative to support learning. There’s this well-known phenomenon that story plays a prominent role in human cognition, and that it thus plays a prominent role in human learning. Story is known to be deeply engaging, an effect we know from multiple media types that utilize stories: we know it from film, we know it from drama, we know it from literature, and we know it from games. There’s this real opportunity to leverage the power of story to create learning experiences that are engaging, hence the name “engage” in the AI Institute. Of course, we want students to learn, but we also want them to become so engaged in the learning that they become better learners, and story offers a very promising vehicle for creating engagement.
Recent developments in AI have introduced the opportunity to create incredibly powerful learning technologies, which was really unimaginable twenty years ago, and it was just barely imaginable five years ago. We now have systems that can dynamically create experiences that are story-based problem-solving episodes. We’re working in STEM education for K-12, and doing so in a way that’s highly customized for learners. We’re looking at what it means to tailor story-based learning experiences for both individual learners and also groups of learners working collaboratively, and investigating how these technologies can be deployed in classrooms at scale. There are advances in natural language processing, advances in machine learning, and advances in computer vision that are foundational technologies. The Institute has core AI specialists in each of these areas, and together with our education researchers, they’re driving forward research in three different dimensions that interact in really interesting ways: interactive narratives, embodied conversational agents, and multimodal learning analytics.
Tell me a little bit about your partnerships.
Educator and Student Partnerships
The Engage AI Institute has a dual focus on formal and informal learning contexts. Our main school partnerships are in North Carolina, Indiana, and Tennessee, which are significantly extended by the phenomenal networks of schools that Digital Promise brings. We also have museum and afterschool partners. For example, we’ve been working for many years with the North Carolina Museum of Natural Sciences in Raleigh. They’re our main partner for museum-based work within the Institute. We also work with Boys and Girls Clubs.
One of the differences that we’ve found over the years working with museums is that it provides a fun design challenge. The visit times that you have in a museum are at the opposite end of the spectrum from what you have in a classroom. In a classroom, you might work with students and teachers for weeks on end, whereas in a museum, a visit time of five minutes would be considered very long. If you manage to have an exhibit that engages learners for ten minutes, that would be off-the-charts fantastic. This time difference means designing for fundamentally different kinds of story-based learning experiences.
I think the most important thing to know about partnerships like these is that they are fully based on relationships, and relationships developed over a long period of time. It’s important to make a good impression when you’re working with teachers and students and administrators early on so you might be fortunate enough to get to work with them again. It’s far easier to keep a current partner than it is to recruit a new partner.
Engage AI Partnerships
One thing that I think is really interesting about the NSF AI Institutes, especially those that are in education, is the amazing interdisciplinary nature of the teams that make the work happen. I’ll just mention three categories of people we have, each in itself very diverse. We have core AI people, and this ranges all the way from machine learning and computer vision to natural language processing, including understanding language and language generation. Then we have our AI and education folks – I’m one of those – who are people that are right in between education and AI. The third category has some exceptionally strong folks in education across the learning sciences, including those who study embodied learning.
We’re bringing together all three of these groups to re-envision what education can be like. The AI Institute is very much being driven by this shared vision from people with incredibly different backgrounds, which not only makes for delightful projects and interactions, but also moves the field forward in ways that are often unexpected.
What impacts do you hope to see from this project?
The main objective driving the Institute is creating a whole new generation of learning technologies that can support interactive narrative-based learning experiences in both formal and informal learning environments and to deliver those learning experiences at scale. Imagine what it would be like for a middle school student in three years to be able to learn any unit of science in the classroom and do so in a way that is driven by an engaging interactive, immersive story, and doing this through deeply collaborative, inquiry-based learning. Then imagine doing this all over the country. That’s what we’re aiming for.
