← ASU+GSV Summit
ASU+GSV Summit
Higher Education in the AI Age with AAC&U's Dr. C. Edward Watson | GSVtv Podcast
00:17:43
Loading summary
Transcript56 lines
- [00:00]
A
Foreign. This is David Harris with ASU gsv and I'm delighted to have Eddie Watson with us today. Eddie is the Vice President for Digital Innovation at AACNU and also best selling author of Teaching with AI. And Eddie has a long history of educational technology in the space prior to AAC and U, who's the Director of Teaching and Learning at the University of Georgia. And Eddie, you've also been an incredible advocate for OER too and driving that nationally. And I've always, always appreciated that you've
- [00:40]
B
been a great partner in that work over the years too, David.
- [00:43]
A
Great. So what we want to do today is really kind of jump into, you know, the work that you've done writing, teaching with AI. Everything around AI seems to be accelerating at an increasing rate. And if you look at higher education, they have very predictable cycles of adaptation and adoption. Are those working now? Is something larger needed? I'd love to hear your perspective on that.
- [01:11]
B
I would say there is nothing about the current moment that feels like past cycles in higher education. I mean, there are a number of different pressures that are on our plates as a result of, well, initially generative AI. We haven't necessarily solved all of those challenges for higher education. And of course agentic AI now brings a new set of challenges. But you know, where the conversation was three years ago focused largely on academic integrity. And so figuring out how to handle those challenges which really touches every discipline, you know, so it's, it's a grand challenge in terms of classroom based instruction for higher education. But then on top of that, as time passed maybe 18 months ago, it was very clear that the world of work was rapidly adopting generative AI. And if at least part of the purpose of higher education is prepare students for life beyond graduation, which certainly includes the world of work, well then maybe that suggests that AI is now essential learning. Developing students competency and fluency with AI. Is that among the learning outcomes that we would ascribe to all institutions such as critical thinking and the like? Increasingly so it seems like it likely is. Well then it's not just an academic integrity challenge. It's now a set of challenges around curriculum reform and change in teaching practice and also ensuring that changes you might make in your curriculum to promote AI literacy actually don't also diminish student achievement of the pre existing learning outcomes such as critical thinking or discipline specific outcomes. So you have an array of challenges that touch on almost, I don't want to say all vectors, but certainly multiple vectors associated with teaching, learning, curriculum and assessment. And then of course now we have new challenges that have emerged as more and more students are rapidly discerning what agentic AI might be able to do in the context of class settings.
- [03:07]
A
And so for people who may not be aware of agentic AI, they hear about generative AI all the time. If there was one or two items that you'd love them to know about the role of agentic AI in education, what would that be?
- [03:24]
B
I don't know if I would describe it as the role like, I think that we're still discerning what the role might be, but what it might do that could complicate current instructional methodologies. So generative AI, you know, it's a very much sort of call and response kind of technology. You, you ask it to write a letter and it provides it for you. Ask it to give me a seven day plan, you know, for a trip to Italy. You know, it can do things like that. Agentic AI is actually, you give it a task and it can go and perform that task. So I could give it a photo of my, my, my car and say, you know, write a description for, of this car, make a website that would collect bids on the car and then distribute that website along with a sales pitch for the car on social media. So we could go and do all of those things and then in three weeks return like the top bidder, for instance. So this is an example of the complexity of the kinds of tasks you can now ask that form of AI to do. But within the context of higher ed, as a student, I could log into my learning management system and then ask an agent to complete everything that's due.
- [04:39]
A
And it will do that.
- [04:40]
B
It would do a discussion board post, it would write a paper and submit it. You know, all of those kinds of things. Take a quiz.
- [04:46]
A
I'm sure that the major LMS providers, Blackboard D2L et cetera, they must have safeguards against this, don't they?
- [04:54]
B
Oh, look at that look on your face. So the, the truth is, as I understand it at the current moment, there are significant challenges for the LMS providers to be able to actually do those kinds of safeguards because they can't tell if it's an agent or not. Within the system, doing activities within a course, there are things that might signal that it's an agent. Like if everything was turned in 75 seconds before it was due, always, well, that might tell someone or something that it looks like it's not a human doing that work. But the AI, generative AI, agentic AI providers are resistant to putting any kind of signal that would tell another system that this is an agent versus a human. So there's some real significant challenges right now. LMS providers feel that they can't necessarily tell if there's an agent there. So then how would a, a university be able to take.
- [05:52]
A
So what it sounds like to me, absence of using technological intervention for signals, having a communication with your students around expectations and standards and saying that you're monitoring this seems like something you would advocate for.
- [06:06]
B
Oh, absolutely. That's maybe the only practice that we have at the current moment. We might see as time passes, that we have additional opportunities or, you know, some technology signaling might be put in place that would enable a faculty member or an institution to tell if there was indeed an agent taking that class. But currently there's very little that provides that's provided in terms of those domains. So really, you know, faculty having conversations with their students about what is allowed and what isn't allowed would be central.
- [06:38]
A
So, you know, with the second edition of Teaching with AI, actually, I wouldn't call it a second edition. I would almost call it all new. It must be like 30 or 40% longer than the first edition.
- [06:47]
B
Right.
- [06:49]
A
But you still have a big focus on pedagogy, responsibility, with AI providing a great overview of the different models. I'm wondering, as you were writing on that and working on it, what assumptions did you go in when you worked on the new book? And where were the assumptions right. And where were the assumptions questioned?
- [07:13]
B
Well, I think one thing that we were very careful with both editions was not to talk about specific technologies because we know that that landscape has change. You know, companies are going to buy other companies. You know, Bard no longer exists. Google rebranded, you know, from the first edition to the second edition. So we chose to focus very specifically on practices rather than specific technologies. And I think that that approach has made it have greater longevity than say, some of the other publications that are out there just because we aren't focusing on a technology. But it's more about implications, policy, academic integrity approaches, as well as pedagogical practice that's, that's enabled by AI. So I think that those are some of the assumptions that sort of underscored our approach that did prove to be successful. But, you know, at any point that you might write a book on AI, as soon as you, you know, quit out of word for the last time, and then the next day you get up and it's like, oh, well, that's, that's something else that we should have. So, you know, there's a point where you have to draw a line in the sand and. And wait for the third edition to be able to address some of those things. So I think it's just sort of like having comfort with saying, okay, we're done with this for this moment, and time will pass, and in 18 months, we'll have a new edition out or whatever.
- [08:30]
A
But, you know, I think that gets into some of the leadership challenges for, you know, provosts and presidents, when we're looking at AI on the learning side, is what do you think, putting budget aside for a second, what do you think that they can provide to their faculty in regards to AI? What type of frameworks would you recommend? Because it's changing so rapidly, you can't keep up with the changes. So what do you need? How do you manage that as a leader?
- [08:56]
B
Well, one thing that I heard literally 15 minutes ago from a colleague who's a provost, he said that they. He is receiving input from faculty that, you know, thanks for giving us space to explore and think, but we need a little bit more guidance now.
- [09:14]
A
Right.
- [09:14]
B
You know, we need more direction from you regarding what we do, what we should do next. So I think that it's sort of that notion of shared governance within higher education, that there really is a partnership and there are roles for, you know, both the faculty in the classroom as well, those that are leading institutions, and that, you know, these are active spaces that both should be having conversations as they kind of move forward within this context around AI. I mean, certainly provost and presidents are the ones that have access to budget strings. And you're right that that's something that is a key concern, and that's certainly their domain. But there is guidance out there regarding helping faculty. I mean, there's, you know, at least. Well, there's probably more. About nine different AI frameworks, literacy frameworks, and maybe the most recent one was from the US Department of Labor. There's an AI literacy framework that there's
- [10:10]
A
nothing
- [10:12]
B
remarkable or surprising in that model. In fact, it kind of summarizes maybe the common elements across many of the other models. But I think putting these models in front of the faculty and saying, okay, well, we. We're now at a point where AI literacy is something that we need to teach. How do we define that for ourselves? And so looking at these models is really helpful. I think also as part of that process is engaging recent graduates from your institution. So someone has been on the out in the real world working for the past year, and, you know, if they. How are they experiencing AI is There expectations. What are, what are the specific elements that they wish that they had known as they began that job? But of course, being in contact with employers themselves is a helpful strategy too.
- [10:58]
A
But so this is really an all hands on deck moment. It's not delegating it to three people to make recommendations is a consistent engagement with faculty and department chairs to get it, to get that conversation going around AI and getting those strategies implemented into the classroom.
- [11:16]
B
I would say yes and no.
- [11:17]
A
Okay.
- [11:18]
B
Like, I don't think that on every college campus, every faculty member needs to be teaching AI, but someone should on every college campus, most likely, this is something that should be somewhere within the curriculum so that students can have some exposure and preparation before they graduate. Now, with that said, I mean, it's a bit of a curriculum mapping activity, right? Like, if it, if it was something that you saw as an institutional learning outcome, we would look over the curriculum. Okay, maybe it makes sense here in this capstone course. Maybe as a, as a, as an entry point into our institution, we want a level set for students, you know, but not everyone is teaching it. But someone should. But with, with that said, that's the AI as a learning outcome.
- [11:59]
A
Right.
- [11:59]
B
The Academic Integrity Challenge is something that's touching every classroom. Right. So understanding what the tools are capable of doing, what the tools, how the tools might handle your specific assignments, I think that is something that needs to engage all faculty just to sort of understand what are the ramifications within the context of your own course and curriculum. But there's another challenge, maybe an additional challenge, maybe it's layered on top of. The first challenge, is that there's a subgroup of faculty that would be engaging with the challenges around AI literacy and actually teaching students to use AI well ethically and responsibly.
- [12:36]
A
So you launched an amazingly successful AI institute at aacnu. Can you talk a little bit about how that has changed as AI has changed and what you see looking into next year?
- [12:49]
B
Yes. So we've had, we're just, we just completed our second cohort of the Institute on AI Pedagogy and the Curriculum. And so the institute runs seven months from September to April. And it's comprised of campuses that send teams to focus on all of the challenges around AI. And the first year, it was very much focused on the academic side of the house. You know, as the title might suggest, AI Pedagogy and the Curriculum. But we had, in that first year for us as an institute for running institutes for over 40 years on lots of different topics, we had our largest cohort with 129 campuses.
- [13:25]
A
Wow.
- [13:26]
B
And then this second year that just concluded earlier in April, we had 191 campuses participate. So we've had. We've had 330 campus teams participate in the Institute thus far. And what we learned during that first year is that confining the conversation really to academics limited the broader discussion that campuses were feeling that they needed to have. So we broadened the mission of the Institute, even though it has the same name. We added elements around how AI might be leveraged for administrative functions across campus, so career services, IRB and research practices. So we had tracks within the most recent institute that focused on those domains as well, just because that's where institutions. I mean, every campus team has to have at least one senior leader.
- [14:26]
A
How many participants did you have this year?
- [14:28]
B
Around 1300.
- [14:30]
A
Wow, that's incredible. That tells you something about the hunger to educate and to come up with strategies that work effectively on campus.
- [14:39]
B
Absolutely. I mean, the Institute having that much time, we spend like the first month, and it's all about action planning.
- [14:45]
A
Okay.
- [14:45]
B
And then we pivot from action planning to helping campuses implement those plans.
- [14:51]
A
And so if you look into the year ahead, will you be running it next year? And how do people find out about the Institute?
- [14:57]
B
So, yeah, we are running it again next year. The call for applicants for the Institute is open now until the end of May 2026.
- [15:06]
A
Okay.
- [15:06]
B
And we expect another large class of participating campuses. In fact, one of the things that we saw in the evaluations this year was campuses really valued the opportunity not just to have the mentorship and the structure, but also to be able to level and hear from other campuses. So, you know, we have campuses clustered into small groups with mentors and that are specifically with that have some common attributes like sam Carnegie classification, or three institutions from the same city, or some reason why they would have interest in hearing from one another. And so we have provide lots of opportunities for them to be able to share across the.
- [15:48]
A
And if they go to AAC and org, they can apply there.
- [15:52]
B
Aacu.org yes, we've got the call is out there now. We also have a new conference on AI and higher education, which has that broad mission for the academic side of the house. It's going to be in Atlanta, Georgia on October 28th through 30th. So we kind of almost see it as a companion conversation to the ASU GSV Summit, you know, placing it at that sort of point six months down the road. But the call for proposals for those who would like to present at the conference is open until April 22nd.
- [16:24]
A
Right.
- [16:25]
B
And of course, registration's open now.
- [16:27]
A
Yeah, I bet you'll get a huge demand to that. So, one final question for you, Eddie. What five words that sum up your vision or aspirations for the future of teaching? Not putting you on the spot, though.
- [16:41]
B
Five words. They don't have to be a sentence, right?
- [16:44]
A
No, they don't have to be five words. Well, yeah, try and put them together. You are a great author.
- [16:49]
B
Oh, boy.
- [16:49]
A
Okay.
- [16:50]
B
Well, I mean. Oh, five words, David, you know me. I'm. My kids say I'm full of words so that you know five words. But I will say that, you know, one of the. The. The foci is really on student learning and student success. I mean, those. Those are central to what the. The mission of an institution. While there's these new opportunities and challenge. Challenges remembering why we're there, what. What we're. What our missions are, what our identity is, and then aligning our activity. So five words. Student learning, student success, Promise, commitment, and love.
- [17:29]
A
Oh, there you go. I was going to say Centeredness. Center on the students.
- [17:33]
B
Okay.
- [17:34]
A
Okay. This was terrific. Thank you so much for joining us today.
- [17:37]
B
Thank you, David.