The AI Divide in Higher Ed: Identity, Expertise and the Future of Learning With Laura Zarrow, pt. 1

Full Transcript

Cristi Ford
There’s a divide happening in higher education. Leaders are racing towards an AI-powered future, but the longstanding academic systems and culture are struggling to transform at the same pace. In this special two-part episode, we look at the divide created when technological change accelerates faster than people, practices, and identities can adapt.

Speaker 1
Welcome to Teach and Learn a podcast for curious educators brought to you by D2L.

Speaker 2
Each week, we’ll meet some of the sharpest minds in the K to 20 space. Sharpen your pencils. Class is about to begin.

Cristi Ford

Join me in welcoming Laura Zarrow and experienced strategic innovator, host of our Gracie Award-winning radio show called Women at Work, which is available wherever you find your podcast. And the executive director of the Wharton generative AI labs at the University of Pennsylvania who is driving innovative and responsible use at AI. Laura, I am so, so excited to have you joining me.

Laura

I’m thrilled to be here. Thanks for having me.

Cristi

Absolutely. You know, when we met last summer, I was just like, you had so many one-liners and just so many aha moments when we were together in Savannah. I just thought we really need to have you on the podcast. So I’m so excited that we’re finally doing this, and I know that we will have a lot to talk about, but as I was thinking about this conversation with you and some of the, the nuggets or jewels that you gave me from the summer, I’m finding I’m spending a lot of time with presidents and provost who are genuinely excited about the fact that they want to be the AI first university, or they have an AI first vision.

But when I talk with them deeply and talk with them more deeply about their organizational scaffolding, I’m noticing that there’s a gap that’s existing. They wanna transform, but often the capacity building at their institutions is lagging behind. So I’m gonna maybe just jump right in and ask you to maybe think about and give us some insights about what do you think is happening when we see this disconnect from the executive level ambition, and then what’s actually happening with faculty culture and daily practices in classrooms.

Laura
There’s a, a number of different things going on here, and I’ll do my best to try and parse them out, and maybe you can help me. Okay. And we can dig into them together. So one of them is that higher ed, historically is really rooted in the importance of the faculty member that when we think about all the debates around tenure aside, that these are people who have worked to become to contribute unique knowledge to their field.

To become master teachers and to take responsibility for shaping the community. And includes what’s taught, who’s taught, how is it taught. And obviously, universities need to do that in response to the world around them. But remember that with faculty holding that unique role, we can’t look at true change to what happens in the classroom or how the university operates without looking at the importance of the faculty member. And so that’s just right there. So you can make a decision from a leadership position, but you need your faculty to be your partner in that, to have internalized that value, to support it, and to be willing and interested in bringing it to life in practice. Now, even if you have that, you’ve got this visionary leadership, you’ve got, um, a connected faculty that say that direction is where we wanna go. Making that happen isn’t that easy all the time.

Cristi
Right.

Laura
Especially when what we’re looking at is an innovation that may fundamentally change what and how they teach.  And so another thing that we have to understand is that this is a generative AI brings with it the potential of profound change and change to things that have been, um, not just sacred, but time tested, proven practices of pedagogical effectiveness. And so teachers, understandably, teachers, faculty members, they’re gonna be at best cautious, at worst loath to upset the apple cart when they know at the end of the day, they need to effectively teach students.

Cristi
Yeah. I, I, I appreciate the framing here, and I’m just gonna jump to those who are listening with bated breath and just say what I know, um, that we’re not saying with this systems change approach and thinking about faculty and their unit of analysis and the larger ecosystem of higher education, we’re not saying that AI is going to replace the educator, but we are saying that there is a focus there and there needs to be alignment in building capacity, both with that visionary leader and that gra grassroots effort to be able to have that happening in har in harmony. And even when that goes well, it’s, there are still some bumps that happen along the road.

Laura
Yeah. And the way I make that kind of granular or related to I think the classroom experience people can understand is that if we, like, I’m not saying at all that like we still need all these subject matter experts. In fact, it’s never been more important. The question is, and it’s a similar question that we asked ourselves when digital technology, desktop technology first emerged, which is how do we use this to be more effective versions of ourselves? How do we use this to teach more effectively? How do we use this to close gaps in student learning and in an educational process at for centuries has relied on things like lectures, tests, and essays, um, how to build in these resources in ways that are pedagogically sound and responsible. Mm-hmm. Um, and that serve the students well. That’s both the opportunity and the challenge in front of us.

Cristi
Yeah. And I appreciate that offering that you talk about with these earlier technological shifts that desktop to network computing. I mean, are there some lessons learned from that example specifically that feel most relevant as institutions are trying to adopt
Generative AI?

Laura

Yeah. So, um, if you don’t mind, I’ll share a little story from the past. So in an earlier part of my career, I had the great honor of being the associate provost at the University of the Arts in Philadelphia. And at the time, it was the nation’s only university dedicated to the visual and performing arts. It had been around for over a hundred years, and the faculty were true masters of their craft and master teachers. And when the technology emerged, um, similarly, we had this visionary provost who could see clearly that desktop technology needed to be incorporated into how we not just pr ran the business of the school, but yes. Teaching, learning, and creative practice. And, um, the faculty were really hesitant. We had, um, a span that well, yeah. I don’t think it will surprise me.

Laura
Yeah. So at one extreme, there were people who actually were doing what they called electronic media and had been since the fifties or sixties using slide carousels and tape cassettes and, um, going like into the vanguard of how electronic art could be experienced. And then we had people whose life’s work were, was spent protecting ways of working that were ancient and that they had developed decades to mastering. And so they couldn’t, for them, it was hard to see how did this belonged in their art making at all. Yeah. Nevermind that, regardless of discipline, the teachers were part of how they were effective in the classroom was the respect that they got from the students. That came from the fact that they were masters, masters of their craft and master teachers. And it was, and what we didn’t understand, except our provost did, was that it was very frightening for teachers. It’s a challenge to their sense of self, what they do well, how they approach something, because good teaching doesn’t happen overnight. And so, um, what we actually didn’t exercise at the time was to let anybody, any of the faculty who wanted to borrow against payroll and pay it back over two years.

Buy a computer to have it home. We’d help them pick it out. We’d help them set it up. They could have it at home. They could do anything they wanted with it. And we found that when we did that, and we offered some workshops that they could come and learn stuff, that it catalyzed a radical transformation to how they use digital technology. But I think its success was anchored in the fact that we let them do it on their own terms.

Cristi
Mm-hmm.

Laura
We let them struggle in private. We gave them places to support their learning curves and customized it for them. And what we saw was even those who were most, oh, no over minded body, I’m not going there a year later, they’re like, oh, no, I need computers in every class I’m in. I need these tools. We’re using them. So it was a matter of how to meet them where they are to help them come along.

Cristi
I love that example, although I chuckled because I wonder, Laura, if you were still in that role when NFTs hit the market, like, how, how would that transformational moment happen? But I, I think what you offer around faculty, uh, and the fear, if I can just lead us in a different direction here. Sure. Um, as you talk about that faculty expertise and mastery of their craft, you know, what we’re also seeing is that there’s real turbulence around evaluation assessment. You alluded to this a little earlier, and what I’m hearing is that some faculty are just completely overwhelmed, right? That they really haven’t figured out the nuances around how to engage and have, uh, have clear con conversations around evaluation assessment in this era of generative AI. And so many of you know, faculty are still saying AI is a threat as opposed to a co-creator. And we know that when students strive, they strive environments when they’re allowed to be co-constructors of knowledge. I mean, I used the example of Minecraft and why that was such a revolutionary opportunity. And so I guess my question for you is, how do you see AI functioning as a co-instructor or maybe even at minimum a collaborator in the classroom?

Laura
That’s a great question. And actually, um, it’s like I’m, I’m picturing like a diagram in my head of like all the different facets of where AI can come into the teaching enterprise so that we could compartmentalize them and think about them differently, because it’s not just one tool that we apply to everything. Like, as you and I talked about, I see is a creative medium that can be

Cristi

Yes. The paint, Laura, the paint. The paint and not the paintbrush.

Laura

And what we need is to figure out what it is that we’re trying to accomplish. So we use it in a meaningful way. So going back to there’s an evaluation and assessment, which is how do we know if, how do we figure out if we’re getting where we wanna go, and how close are we? But there’s also how do we stimulate the learning process so that hopefully we are ready for that evaluation and assessment. And it’s in its co-creation that when we look at AI as our companion, as a collaborator, as a resource to help us be better at whatever we’re doing, that’s where it holds tremendous promise. We’ve done research that’s demonstrated this, and we can see how differently individuals can perform, invent, and learn and groups when they’re using tools that are designed to be of assistance in a meaningful, responsible way. Now, let me dig into those words a little bit more, and I’ll give you some examples. When, and this relates to actually something we’re building within GAIL, the generative AI labs called Primer, which is where we’re in, we’re developing a multi-agent system that spins up educational simulations.

Cristi
Okay.

Laura
So driven by pedagogical principles that a teacher could customize their curriculum, customize their lesson plans to spin up exercises for students that will help them learn whatever the goals of the class are. It’s not writing essays for them, it’s not designing the test. It’s, um, helping, having another creative resource to come up with new ways to help students learn that can be driven by pedagogical principles, by whatever learning goals you have customized to meet the students where they are. And that’s, um, that’s something we’ve never had the benefit of before. And that’s just one example. You know, you can use it to tutor you, you can use it to coach you. We’ve also learned a lot from research that there are ways to use it to tutor that can hurt, and ways to use it as a tutor that can help. Um, because at the heart of it is, are we preserving the student’s opportunity to struggle with the acquisition of the new skill, and no knowledge that’s necessary to the learning process?

Cristi
That healthy friction.

Laura
Yeah. Yes. That’s healthy, necessary friction. While making, while closing the other gaps that may exist in making mm-hmm. It, um, is accessible for them. You know, we talk about experiential learning all the time.

Cristi
Mm-hmm.

Laura
Not every school district or classroom teacher environment can afford experiential learning or has ways to access it. But with generative AI, we have the capacity to create all kinds of learning experiences that, if we bring pedagogical principles to them, AI ethics we implement them well, can really, we’ve seen, can spark meaningful learning. And you talk about what’s
happening in GAIL, I wonder when you see other institutions who may not have some of the resources at their fingertips, in which, in the ways in which you’ve constructed them in GAIL, what are the biggest maybe opportunities or dangers when we’re asking faculty to redesign their teaching around AI when the system around them has not shifted?

Yeah. So that’s where, you know, the stuff that let’s say is in the public domain, that’s within reach by going online. There are vast, vast resources there. And even, and you don’t need to have, um, elaborate systems as much as you need to have a clear concept of what you’re trying to accomplish, and to work with AI to help you get there. So, for example, um, we have a prompt library that’s available on our wear website, gail wharton.upen.edu. And we developed a whole series of prompts that we very carefully tested that are available to the public. You can take ’em, you can change ’em half at it, but they can help you spin up a tutoring session or spin up a simulation of an experience, um, or get coaching on something. So, simply by constructing well-written prompts or leveraging well-written prompts .

You can work with your average large language model, whether it’s ChatGPT Gemini or Claude, um, to create an experience right there just on your desktop using your access to those models. Um, there, you know, that’s just the, the basic point of entry, develop, there’s a lot of ticks and tricks that we could also talk through for how you use that interface more effectively. So that you can get where you wanna go with it. Um, and then, you know, we have an amazing world that, you know, a great deal about of EdTech, trying to figure out meaningful ways to incorporate AI into educational systems. Um, and then we have, what we see happening actually in business quite often is where, you know, business has the resources and so connected to their bottom line goals of looking at how AI gets plugged into complex systems of production and delivery. And, you know, organizations can spin those things up themselves, but they are very hard to build. They take a great deal of expertise. Um, yes. And they have their own built-in promises and challenges that we can talk about a little bit.

Um, so there may be schools that are developing their own resources. We developed Primer, which we’re releasing in January as an open source system that anybody could download and adapt, build upon, and plug into their systems. So there are an increasing number of resources out there, but don’t, um, don’t, uh, get mistaken that you have to be an expert or you need high levels of technical skill. One of the gifts here is that there’s a low barrier to entry. That’s right. As you bring your subject matter expertise. And actually, this is one of the most powerful ways to use generative AI. Use it to be your critic, use it to be your coach. Use it to be your collaborator.

Cristi
I love that framing there. And I guess I just wanna double down on one area before we move on to another part of the conversation. You know, as I’m listening to you talk about, um, you know, the ease of use, the, the entryway bringing your mastery level expertise. Um, I guess I still am, I, as I’m listening and talking to faculty members around, uh, you know, historic assessment models, you know, we’re really talking about a time and era where evaluation assessment needs to shift. And so I guess, as you’ve been on this research in these areas, what needs to change in assessment models if we’re really trying to measure, not surface learning, right? Deep learning, um, rather than just the compliance. And is that also, um, a, a, a focus on different types of outputs? So not focusing on product, but maybe process. Like I just lo I’d love to know what you think about

Laura
Okay. So what I’m gonna give you what I think about it. ’cause I have to tell you, I’ve not been doing research into the evaluation and testing part of things.

But I wanna pick up on a few things that you said, hopefully it’s valuable, which is that in, we have two different paths of learning that are coming together concurrently. So there’s always this, like the world, we need to understand the world, we have to build our skills. It’s why you and I work in education, you know, is that we know the learning process is enduring and important and valuable. Yes. And it’s always been a challenge for how do you teach everybody everything they need to know in the limited amount of time that you have. Now as we’re trying to continue the learning process and keep it relevant for all of our students in the world that they’re graduating into and serve the world that we live in, by making sure there’s an educated, trained population that can go to work, we also now have this extraordinary new resource. One of the metaphors that, um, has stuck with me that I heard was, it’s like having intelligence as a utility and as humans, we’ve never had that before. Like, we now live in a world where we expect a light switch, we expect electricity, we expect running water. Those things 150 years ago, were not a given

Now. It’s not just that there’s generative AI and it’s this cool thing and it makes stuff, it’s like having all the intelligence there ever was if we can learn how to harness it. And so, figuring out how we work with generative AI to serve the learning process is the other challenge in front of us. And as we incorporate generative AI into teaching and learning, whether it’s the students or the instructors now have to one, learn how to use generative ai. What can it do, what can it do? How am I comfortable using it? So they have to learn, um, the process of how to work with something that, by the way, is changing almost weekly.

Cristi
Right

Laura
Right, right. And that’s its own hard thing. That’s its own hard thing. Yes. And then, given that we’ve never had a utility like this, how do we apply it to that age-old process of learning in meaningful ways?

These are big, juicy, kind of delicious, scary questions. Yeah. They are. And so they are. Well, and then, oh, keep going. But the process thing is more important. I love that you raised the question of process because it, there’s like one, what’s the process that we go through societally? How do we go through it as an organization, but as individuals, how do we develop a productive process for staying up to date with AI and figuring out what we can do with it?

Yeah. And all these things kind of have to happen concurrently.

And it comes back to, you know, how do we, maybe you and I in the work we do, how do we help make environments where all those things can happen concurrently? That’s right. And people, ’cause it, it, we’re asking a lot of people who are already working hard to make challenging things happen.

Cristi

Yeah. Yeah. And as I I love this piece around intelligence as utility, ’cause as a, as a psychologist by training, I’m always thinking about, um, scholarship, evolution, expertise and, and more particularly, identity. Right? So if we’re honest, this AI revolution is really triggering, I’m finding with, with, um, some institutions and, and faculty members an identity crisis in higher education because for centuries, you and I have been rewarded for static expertise for the mastery around the content. We know the per, you know, the permanence, we have the tenure. Um, and so that static expertise has been valued, it’s been rewarded, and now we’re asking faculty to embody this dynamic expertise. And so I guess I’d love to have you give me some of your thoughts around how do you think about generative AI and the way it’s reshaping the very idea of expertise, period.

Laura

Mm-hmm. It’s a really, it’s a great question. So I gotta wrap my head around it so I can dive in in a meaningful way.

So I think there’s a couple of things here. One is that, and they may be two things can be true at once. Are there either size of coin they can be, we’ll see , um, we decide which cliche we’ll attach to it. , um, that we, subject matter expertise has never been more important because we need to be able to discern what matters. We need to know where we’re trying to go, and be able to identify, um, when we’re getting quality information and when we’re not. And it’s with subject matter expertise that we likely have and the experience that we bring to having put it to use for a long time that we can make judgment calls that are necessary to guide ai.

Cristi
Yeah.

Laura
On the other hand, AI provides an extra, as we thought you could learn. Like my husband fixes everything that breaks in the house by watching a YouTube video. , right. , my daughter taught herself to sew really well. Yeah. Watching YouTube. So think about this now. Like that means that one, we can usually find anything we wanna learn on YouTube, but what if we could actually, in the YouTube offering ourselves, that’s what generative AI does for us. So we all have a capacity to learn anything we wanna learn in a different way and to construct any kind of learning experience we want. Do I want somebody to do 20 questions with me as to whether I’m ready to have dinner with in laws? Do I want to have somebody being, um, a practice interview for me? So I’m ready to talk to them . You know, you can use it for anything that you want to.

On the other hand, and going back to identity, um, part of my identity was formed as a designer, and I’ve done a lot of work ss an art director. And so the way that I shape my prompts is often like an art director that, that aspect of my identity helps me come in and say, okay, where are we trying to go? What language should I use to help us get there? What are the parts that are strong? What are the parts that are weak? How do I protect what’s good? How do I improve what’s bad? How do I do that in a reasonable amount of time? And those are all the same traits that we need to use when we’re creating prompts and trying to develop something using generative AI.

Cristi
Yeah. That’s good. And you know what, it reminds me of something that you said on our prep session that stayed with me around art, actually, when you talked about, uh, all art being derivative, and you said, um, you know, what it made me think about was how generative AI exposes accelerates something that has always been true. So I, I appreciate you talking about things being, you know, both things can be true. But I guess I wonder how should educators and creators, um, think about the role of generative AI in the creative process? And I’d also love for you to talk about three C’s. I’d love for you to unpack the three C’s as well.

Laura
Yes. Um, so the three C’s are consent, credit, and compensation. Okay. That’s it. So let’s start there. And, I think that’s a good point of departure that it is essential that we respect and honor the artists who make original contributions and their work, and that they control where and how it’s used. It’s done with their permission. And if somebody’s gonna make money, it includes them. It’s their ip, it was their original work. Now, that’s one part of this, but the other thing is, um, is it meaningful? Does it count when we use generative AI to help us make art?  Like it depends on what we’re making and how we’re making it. But to dig in a little bit to what you were referring to throughout history, every artist looks at the world. They’re in all the art that came before them.

You can think of it as appropriation that has, there’s a lot to appropriation. Yeah. That is distinct. But, um, what it really is, is we as humans, we’re in an ongoing conversation with what we make and how we make it. We learn from the generations that came before we don’t create in a vacuum. Right. And so, um, it, it’s like you can’t walk into a modern art museum and really understand what you’re looking at. If you haven’t, you can’t understand Picasso if you haven’t looked at the last width. Like That’s right. One would not have happened without the other. Baked into how art evolves. . So with that in mind, if you bring curiosity, your own, um, vision and drive for where you wanna go, and you’re working with generative AI to help you generate new ideas, to help you produce work faster, but you stay in the driver’s seat, you’re, you still can be making art. As a matter of fact, um, Ezra Klein did a fantastic podcast interview with, um, Brian Eno, who decades ago was a real, like, he was on the edge of showing the world that you can make meaningful music that’s auto, that’s generated. The question is, who brings the judgment to it? Who decides what to keep and what gets tossed? Who art directs it? And so that’s where expertise comes into play to really shape what the AI is doing.

Cristi
Laura, this is such a great conversation. I think this is a great moment of pause and we’ll pick up this conversation on the next episode of Teach and Learn. We’ll continue this episode with Lara on the next Teach and Learn. Until then, remember to follow us on social media, you can find us on X Instagram, LinkedIn, or Facebook at D2L and subscribe to the D2L YouTube channel. You can also sign up for the Teaching and Learning Studio email list for the latest updates on new episodes like this one, articles and master classes. And if you like what you’ve heard, remember to rate us, review it, share this episode, and remember to subscribe so you never miss what we have in store. You’ve been listening to Teach and Learn, a podcast for curious educators brought to you by D2L.

Speaker
To learn more about our K – 20 and corporate solutions, visit D2L.com. Visit the Teaching and Learning studio for more material for educators by educators, including masterclasses, articles, and interviews.

Speaker
And remember to hit that subscribe button. And please take a moment to rate, review, and share the podcast. Thanks for joining us. Until next time, school’s out.

Resources Discussed in the Episode

• Wharton Generative AI Labs
• Wharton's GAIL Prompt Library
• Wharton's GAIL Primer
• Women @ Work