A Practical Guide to AI Literacy

Sophie White: Hi everyone. I am so excited for you to check out our latest episode of EDUCAUSE Shop Talk. In this one, we really dive into the practical applications of AI and higher education and how students are using AI and teaching and learning. I think we've been doing a lot of intellectualizing in our community about AI, what it means for humanity, all of these bigger questions, but this one was really important as we think about how can we as higher education practitioners actually support our students as they use AI? And also how do we support faculty as they support students in using AI. So check it out and let us know what you think.

***

Sophie White: Hello everyone, and welcome to EDUCAUSE Shop Talk. I'm Sophie White. I'm a content marketing and program manager with EDUCAUSE, and I'm one of the hosts for today's podcast discussion.

Jenay Robert: And I'm Jenay Robert. I'm a senior researcher at EDUCAUSE, and I'll be your other host.

Sophie White: Great. So we are really excited today to talk about student AI use and how we can guide students in their use of artificial intelligence. Obviously, this is a really important topic and one that's top of mind at a lot of institutions. So I'm thrilled to have three fantastic guests with us today to talk about their experiences working in different spaces on supporting students' use of AI. So I'll introduce them briefly and then we'll jump into the chat. So first off, Dr. Leo S. Lo is the incoming Dean of Libraries and university librarian at the University of Virginia, where he also serves as advisor to the provost for AI literacy and holds a courtesy professorship in education, A past president of the Association of College and Research Libraries, he leads national initiatives focused on developing AI competencies for library professionals. Leo earned his EdD from the University of Pennsylvania and studied AI at the University of Oxford. Thanks, Leo, for being with us.

Leo S. Lo: Hello.

Sophie White: Next up, we have Dr. Jeanne Beatrix Law, professor of English and coordinator of Kennesaw State's graduate certificate in AI and writing technologies. She has published widely in both academic and popular spaces on the intersections of generative AI and rhetorical prompting. Her thoughts on open educational resources and responsible AI have been featured in the Chronicle of Higher Education and by the Council of Writing Program Administrators. She serves as an AAC&U faculty mentor and is a founding member of Open AI's Educator Council. Jeanne, can you say what AAC&U is? Just in case folks don't know that acronym.

Jeanne Beatrix Law: The American Association of Colleges and Universities.

Sophie White: Beautiful. Thank you. Thanks for being with us.

Jeanne Beatrix Law: Of course.

Sophie White: And then lastly, we have Anissa Vega, Associate Vice Provost for Curriculum, Digital Learning and Academic Innovation at Kennesaw State University, where she is also a professor of instructional technology. Her research explores innovative learning models including learning analytics, teaching at scale, and personalized learning. In her administrative work, she leads strategic initiatives that enhance academic excellence, advanced digital pedagogy, and promote equitable learning environments. So thank you Jeanne, Anissa, and Leo for being with us today.

Anissa Vega: Thank you.

Sophie White: So I'm curious, all of you are working on AI in your professions. Did you anticipate this happening five years ago when you thought about your career in higher education? Did you see AI being a major part of it?

Jeanne Beatrix Law: I don't know that I thought that AI was going to be a major part of it, but I felt like technology and digital pedagogy for me definitely would. I mean, I'm a second career academic, so I come out of government and NGO industries. So using digital tools to write is kind of something that I brought with me. And also that's kind of how I teach. I mean, we were doing Twitter essays in the mid two thousands in my classes. We were doing lots of multimodal composing, writing podcasts, video scripts, things like that. So I think for me it was more transitioning to the speed of AI than it was the actual tool itself.

Anissa Vega: And I would say that for me, I certainly anticipated that AI would be part of society at some point during my career. What I didn't anticipate was that it was going to be so accessible and that accessibility really shifted how I, or at least the level of involvement that we would have to have at this time.

Leo S. Lo: Yeah, definitely not five years ago and not to this extent and not how disruptive it has been. I remember when generative AI, when it first came out, all the headlines were about education, higher education, and certain, I certainly did not expect that five years ago.

Jenay Robert: I have to agree with that. Of course, working at EDUCAUSE, I knew that technology played a central role in my career path, but definitely didn't think it would go from something that I thought, huh, that's interesting. I wonder where that will go to almost overnight feeling like, oh, it's going everywhere all at once.

Sophie White: I agree on Jeanne, your background with English and writing is really interesting to me. I got a master's in English literature and I took a couple of classes on, what was it called, digital humanities. And I thought they were really interesting, but I feel like they were kind of part of my regular curriculum with other things that I was studying. And if I could go back years and do that class again in the age of AI, I wonder how different it would be just to think about how we interact as humans in the age of AI and how disruptive it is. So it's fascinating.

Jenay Robert: Sophie I have a feeling someone on this call would let you sit in on one of those types of courses.

Sophie White: That's true. I guess if I ask nicely, maybe I can go to one.

Jenay Robert: Yeah, I wanted to kind of just call out that we got the idea for this episode. This episode is all about how we guide students and we really wanted to get a little bit more into, I think a lot of the conversations about AI right now are kind of high level and they have been for several years, but I've definitely seen a shift, I would say over the last year or so to where people are really getting into the nitty gritty details about, okay, yes, we know it's important to guide students on ethical and appropriate AI use, but what does that look like? And it's almost getting to the point where nobody wants to hear the high level anymore. We just want to know how do you do it? What are the nuts and bolts? And so we had this idea in the back of our minds and then we saw something popped up across my LinkedIn feed with Leo. You had a framework that I'd love for you to talk about that reached its way to some students at another institution and they did a video about it and we thought this was just such an interesting story. I'd love for you to tell that story. And then of course, I just kind of want to get into how all three of you are hands-on guiding students. What does that look like in practice?

Leo S. Lo: Yeah, thank you Jenay. So maybe give some context first, when ChatGPT first came out, it was what? Late November, 2022. It's near finals time and final assignments due time. And all of a sudden we have students coming into our libraries asking our staff librarians to find them journalism books and we couldn't find them and we're like, what's going on? Why is something wrong with our system? Something wrong with us? And then we found out they were using ChatGPT, and those were hallucinated citations basically. They seem very real, but we couldn't find them. Now those could have been teachable moments, but we were not ready. So I wanted to make sure that our staff, our librarians and faculty are ready to help our students when they inevitably, and we have seen by now that it's very easy for them to use that, but they may not use it in a kind of proper way, in a intelligent way to help them actually.

So I started going more in depth into what is this thing, ChatGPT, LLMs. And then I realized that the quality of the output is very related to the quality of the input, what we call prompt now. And so I've come look into how should we think about how to prompt these things? In a way it's very accessible because it is natural language. My mom who is 84 is using it, is loving it because she can just talk to it. But I think there's a little bit more than that if we want to teach our students and then faculty to use it in a efficient way or optimize the use of it and reduce the errors basically. So I came up with a framework called CLEAR framework for prompting basically. So CLEAR stands for concise, logical, explicit, adaptive, reflective. So it's not like one of those formulas out there, which there are a lot and they're very useful.

They say, oh, assign this role and the formula, follow a formula to how to do prompts. You can go to those prompt libraries and copy those. Great, you can use those. But I think mine is a little bit more of a mindset, concise as in being articulate, don't use excessive words and confuse the LLM. Just be able to articulate what you want, logical cover logical structure to it so it makes it clear to the machine basically. And explicit being clear, clear on what exactly you want. These are the context, this is the output format I want in the table format or graph format, all of that being very explicit about what you want. And then adaptive, it's not a one shot prompting every time it comes out where you want to be adaptive to react to it, and then reflective, think about the entire experience and then move on from there and think about the larger picture.

So it's more of a mindset and it's caught on with a lot of library librarians. They put those into lib guides and then more and more people are using it and find that kept useful. So I'm very, very happy to hear that. And what the story the LinkedIn feed was that I think a couple of students who worked at the library at Texas A&M Corpus Christi, they make a very nice video. It's like four minutes long and explaining that concept, a very, very nice way, very easy to understand way for their users. So yeah, I'm just happy that people are finding it useful.

Jenay Robert: We can rattle that link that in the show notes too.

Sophie White: We can. I love that use of just how higher ed can collaborate. I emailed the library, actually it was Texas A&M Corpus Christi, I think their Bell Library and they couldn't be part of the conversation today, but they said how valuable the tool had been to them and that the students were excited to use it. So thanks Leo for creating something so useful for multiple institutions. Anissa and Jean, do you want to talk a bit about what you're doing at Kennesaw to support student use of AI?

Jeanne Beatrix Law: Sure. I can start just from a class level or a program level. So love what Leo did with that framework. It really does kind of speak to the writing process. So when we think about specifically prompting writers for writing or content creation use cases, we have tested. So just take a step back. November, 2022 totally transformed everything for all of us. I mean, I was in a classroom and my students and I were kind of playing around with it. We were writing poems, we were doing all this fun, cool stuff. And then in January, several first year writers came to me and said, this is going to change the way we write and we need to figure out how best to use it cuz we're not using it very well. I'm getting lots of hallucinations. I'm only single prompting it. So we surveyed about 1700 students, first year students at Kennesaw State between 2023 and 2024.

And we came up with and tested and iterated on what we call a rhetorical prompting method specifically for writing. And it is the writing process that is a best practice in our field that has been around since the 1980s. But we explicitly train students and teach students how to engage with whatever large language model they're using to put the different pieces of that writing process in. And then it becomes iterative. So purpose, audience, style, genre, context, and then providing all of the facts. So never trusting a large language model with the facts, but giving it your own facts. And if you do get facts out of it, being able to iterate and use our information literacy model to analyze whether or not that's accurate or a hallucination. And our students have really engaged with it, not just in general education, but also in our upper division and graduate writing courses as well.

Anissa Vega: And Jeanie has started programs at Kennesaw related to AI, bringing those skills to students and helping train them. And in my role, I support her in developing those programs and making sure that they get launched. And so we see AI appearing as a concept, a topic or a skillset that's being offered across all of our colleges now. And those courses tend right now to be optional as electives. But I'm interested to hear as other institutions might be making those sorts of courses required or those skill sets required in their general education programs. I know I'm interested in that. We have not done that at Kennesaw, but I do believe some of our courses are starting to incorporate AI into the instruction itself. So in addition to the curriculum, we're also supporting our faculty in determining what ways they can use AI with their students and how they can ensure academic integrity or how they can create assessments for those skill sets for students as well.

Jenay Robert: Faculty. Oh yeah, go ahead, Leo, Please.

Leo S. Lo: Yeah, so it's just follow up so many questions too. And I want to ask about, first of all, like ChatGPT. I love what you're doing with how teaching students how to use it to write, but humanities is being disrupted the most probably because ChatGPT can write. I wonder how are other faculty in the humanities disciplines taken to that?

Anissa Vega: I had a group of Jeanne's colleagues come to my office in January of 23 asking me to ban and block AI from our campus systems, which is impossible at this point. So that launched new conversations across the campus.

Leo S. Lo: Yeah, we are facing similar challenges because a lot we want to give our faculty the freedom to decide what they want to do, so we cannot mandate or force anybody to do anything. But at the same time, a lot of people are still like, oh, this is completely changing my discipline, changing the way people learn how I teach. I don't know what to do with it. And so definitely a lot of people are very hesitant for sure in humanities. But to Anissa, I just wanted to share a little bit about what we're doing with the curriculum too. We had a one credit test pilot course on AI literacy for students. And we have done that since the spring of 2023 or 4, I don't remember now. But very early on. And now we're trying to expand it into a three credit course and hopefully make it a required general education course in the future. So we're hoping that we can do that at New Mexico, but it'll take some time.

Anissa Vega: That's interesting. I would love to hear what framework or definition you're using for AI literacy. When I try to pull together what that skillset looks like and how we describe that, there's so many different options out there and there's very little research on the effectiveness or the completeness and comprehensiveness of these frameworks. Personally, I like Anthropic's four Ds. It's super simple and practical, but I would love to hear what you're using at your institution.

Leo S. Lo: So I have my own, but I'm not forcing them to use mine, but that is part of their thinking. But I've done some research on it too. And I feel like there are, like you said, there are a lot out there. None have been tested because they're so new, but at the same time, a lot of them are so similar. So in a way just pick one, just use it and then you can iterate by doing that. I feel like it's a lot of the times, and it's very true in higher ed, is that we want to win into something that's perfect to do it, never get to that stage. Maybe we can, as more people do it, we can make it much better than just waiting. And so I'm like, Hey, just pick one. You don't have to use mine. Just pick one and just go with it.

Anissa Vega: Right. Innovation certainly can come in conflict with higher education's culture. For best practice, we always want to put together the best practice, but then how do you innovate if there is no defined best practice?

Sophie White: Yeah, this feels like one of those instances where maybe perfect is the enemy of good. And I like hearing your example of how you're iterating on it, Leo. So I'm looking at your title as advisor to the provost for AI literacy. And I'm curious, how are you also working with leaders at the institution to advise on AI literacy, make sure that these academic integrity models that people are worried about, we're hearing from faculty who are maybe more opposed to AI. How are you taking that stakeholder need into consideration while still allowing this innovation to happen?

Leo S. Lo: So that will be my new role. So I'm a start, but at University of New Mexico, what we've done is through the libraries, we collaborate with the Center for Teaching and Learning and other colleges as well. We've developed some upskilling programs and they are cohort based. So instead of just say, Hey, a workshop, we come and do it one shot, it's like 12 weeks or eight weeks, usually in the summer or in the regular semester. So the first one we did was in the summer of 2023 just for my on college, we paid the subscription of ChatGPT because that way they have a higher quality product and also they can switch out the use our data for training purposes. They have that option and that makes people feel a lot better, even though always privacy is the main issue every time we use it, every time we discuss.

So with a twelve-weeks program, we have first two weeks of introduction, a little bit of training, and then the follow eight weeks, they designed their own individual projects that is directly related to their work for librarians. There's different waiting notes. Some people do cataloging, some people do reference, some people do use it for instruction. And then they'll use ChatGPT to do things for them directly to relate to that work. And they will document it and we'll get together every two weeks is there a community of practice? And they will share prompts, tips, concerns, all of that progress. And then at the end we put together a PowerPoint to share with the rest of the college and the university. So that has proven to be quite well received and people have become more confident in using it. And then we expanded it to the rest of the university with partnership with Center for Teaching Learning.

We have one for teaching faculty, one for people wanted to use AI for literature review. And then we had one for people who wanted to use AI to develop OER open education resources. And the most recent one is this summer we have one for humanities faculty because one of the Russian literature lecturer professor did one the previous summer and she liked it so much. She want to expand it and teach her colleagues basically. And so in a way that becomes a train the trainer kind of model. And we recruit people who are already interested, people already wanted to, I want to learn more instead of trying to convince people who are not interested. So that's our approach in a small group, but hopefully you can scale by this train the trainer model.

Jenay Robert: It's so funny that you asked that question, Sophie, because as we were talking earlier, I was noticing, I mean Anissa had mentioned as well the support for faculty, and it was occurring to me that this conversation about how to support students really naturally drifts into this conversation about how to support faculty because we can't support our students if we're not starting with faculty support. And so I'd like to hear more about what's going on also at Kennesaw with regard to supporting faculty. I know that it's part of what you both are working on. What does that look like in practice? You're muted, Anissa.

Anissa Vega: Sure. Jeanne, you want to go first? Your college is doing quite a bit.

Jeanne Beatrix Law: Yeah, sure. So at a university level, we've done an AI fair, we've done very meta stuff to get meta events, to get folks just to come and listen, just to hear the use cases that some of our more successful or the folks who, the colleagues who have integrated AI into their classes are doing kind of doing the same thing as humanities and social sciences folks. But what I will say is that KSU has done a really good job of providing faculty with multiple opportunities to engage. So there are multiple asynchronous opportunities from things like Udemy, Auburn University has a really great program. They do asynchronously to train Coursera. They're all of these opportunities for faculty to upskill, which is super, super important. It's not just a lecture or a workshop like Leo said, because those tend to not be as successful. But what I have heard from faculty colleagues is providing us funded opportunities.

So when I say Coursera and Udemy and Auburn, Anissa's office and Academic Affairs has helped fund those, right? So faculty can go to Coursera, they can go to the Auburn University Udemy course, and they can take those for free. And that's what's super important. And they're not just lectures. They build products while they're in those courses and peer review each other. And I think that's one thing that faculty like me and other faculty who may be pushing back a little bit are super appreciative of is the multiple modalities and the departure points that KSU has provided to us.

Anissa Vega: And we've had a couple of opportunities to create quite a bit of discussion and debate. We've had panel discussions for the campus as well as a conference where there was quite a bit of discussion that took place as well. We've recently established a AI policy committee and we've been working all summer long on drafting an institutional AI policy for the institution, which also includes efforts to ensure that faculty have access to training as well as students have guidance in their syllabi on what the expectations around AI use is in each class.

Jeanne Beatrix Law: And I think that's one of the things too that I have heard, because before this year for eight years, I was the writing program administrator for first year composition. And KSU is a very large university and we have about 10,000 students who come through first year composition every year. And so there are about a hundred faculty teaching in that program. So I hear a lot from colleagues teaching in gen ed. And what I consistently heard over the past two years was that they really appreciated the syllabus policies that were put together by teaching faculty as well as administrative faculty. And they're kind of on a sliding scale. Like faculty right now can choose their level of engagement for their particular class. And that is something that I don't know if we'd be ever going to be able to scale that going forward because we're such a large research university, but just getting started, it has created a lot of trust coming from faculty to know that they have a choice in those statements on their syllabi.

Sophie White: These are great examples, and I think that really ties Leo to what you were saying about not forcing people to use AI, but giving them the tools to try it out. And we're hearing this a lot across our conversations on Shop Talk about how to start small and build momentum and trust from there. We talked in our last episode with two CIOs about the role of technology leaders on the campus, and they also said similar things with different outcomes of prove your not efficiency, but effectiveness as IT with small projects prove that we can trust you and then we'll escalate to larger and more strategic ones. So I love that in the use of AI, of let faculty try it out, use it in small ways, and then maybe they can work more on this in-depth AI literacy or tell their other maybe a little more resistance faculty members how they can best use it too.

Leo S. Lo: Yeah, a lot of people ask me, how do you convince people who don't want to use it? People are resistant to it. I'm basically saying I'm right now, I'm not spending my energy there, but I usually use the example of the diffusion of innovation, if you've heard of it, it's basically, it's a bell curve that the early the innovators, the early adopters, and then the majority and the laggers, basically they say that if you can get about 18, 20% of people of the population start doing something or using a new piece of technology, it'll create a momentum to spread to the majority. And then eventually the laggers, so I'm like, well, let's get to that 18 20% first and focus on people already interested.

Anissa Vega: Well, and Kennesaw Leo, Jeanne is an opinion leader, so she's doing a great job about helping getting that momentum around our faculty to reach that 20%. So Roger's theory is my favorite theory.

Jeanne Beatrix Law: And I think you can do that with students as well. I think you can reach faculty through students. So we have a significant pushback in my field in writing studies where there is an actual manifesto that has been published nationwide that's a manifesto of refusal to use AI.

And so reaching folks in writing studies and valuing their points of view while trying to reach some kind of common ground proves to be a little bit more complicated. So in our graduate program, one of the things I'm piloting this fall not just to help our grad students, but also our faculty who teach in our grad program who are very anti AI, everything we produce with AI in my prompt engineering class, students are producing a labor and sustainability statement that goes along with that. So they are actually taking MIT's most current data and calculating how much CO2 and water they are using to produce this. And we are using those sustainability statements as conversation starters. And I think when we do things like that, we're addressing some of the concerns that our colleagues have in writing studies around sustainability and labor. When we think about generative AI.

Leo S. Lo: I just want to say that's really fantastic. That is AI literacy, right? You are really teaching students that.

Sophie White: I agree. I feel like we're moving away from it, but I feel like there has been a binary of, oh, it's all good or it's all bad. And I really love this nuanced view of how can we look at the good and the bad sides of it and take that as a whole and what an important critical thinking skill for any students coming out of an institution to take that into the world. So yeah, thanks for sharing that.

Jenay Robert: Sure. Yeah. And on this topic of what even is AI literacy and how do we include people who are pushing back on AI in that conversation, I was actually just talking about that this morning with some colleagues where to me, I think there is no widely accepted definition of AI literacy right now, and that's for us to construct together as a community. But for me, the folks who are pushing back hard and really wanting to emphasize teaching students how to critique the technologies, to me that still is a form of AI literacy. That is AI literacy doesn't necessarily mean that you're adopting all the tools and using them all every day. It means that you are critical of the technologies and you're careful about what you can consume and use and how you do it. So I love that, especially this idea of giving faculty opportunities to sit on panels and talk about the various aspects. I think we can converge in some common ground that's comfortable for all of us. It may not look like strictly one side or the other, it won't, but we need that opportunity for discussion.

Leo S. Lo: That is so true, because I've encountered that as in my role as the president of ACRL. So I set up a task force to develop a set of AI competencies for just specifically for library workers. And during this process, we got not a lot, but very vocal kind of resistance from people who are part of that AI refusal kind of movement. And they would say, oh, environmental impact and all the other things. But what I try to tell them is, you don't have to, like Jenay said, you don't have to use it to adopt everything of the AI, but whether you are for it or against it, if you become more AI literate, your arguments are going to be stronger, even arguments against certain users of AI, and that's needed. I think we need to have that as well. So that's how I try to persuade people. Not always successful by the way, but at least that's the way I see it. So I absolutely agree with you, Jenay.

Anissa Vega: And Leo. I suspect that the more AI literacy that our students have, academic integrity will improve. I have a 15-year-old daughter who I'm teaching how to drive, and she and I were having a conversation about why she doesn't speed, and she says she doesn't speed because she doesn't want to get caught. She doesn't want to get a ticket. And I said, well, that's not the reason that you don't speed. You don't speed because it's dangerous. You can lose your car. There's all these life consequences that could happen. And so when using AI as a student, when you're learning, if you're using it to cheat, you're cheating yourself because my daughter, if I had let her believe that it's about getting caught, then I think she'll be speeding whenever she thinks she's not going to get caught. And so if our students understand that the skills, the foundational skills in their numeracy and literacy and historical analysis and critical thinking, how important that is for them to actually be an effective AI user, then the reason why will help prevent or at least give them the knowledge as to why they should not use AI in irresponsible ways.

Leo S. Lo: That is so true. I love that too. I love that example because in a way, I think traditionally people go to class, they learn and they want to get a good grade. I think that's the outcome almost what people want. But if AI can do a lot of these things, maybe that's what we need to focus on is their motivation and their mindset is that they actually want to learn something because AI can get you an A or a good grade based on traditional kind of conventional measurement. Maybe the next step. That's what keeps me hopeful is that maybe the next step is forget about the grades, just use it to learn. And then we focus on the motivation in some ways. Easier said than done. I don't know how to do it, but

Sophie White: I love that example too. And I had this realization recently that I think because at EDUCAUSE, we're in deep about AI every day. So I feel like I'm generally having a lot of these types of conversations. But I was just on vacation a couple weeks ago and was talking to a cousin, she's a social worker at a children's hospital. I think she'd heard a lot of the headlines about AI will destroy the world, AI will destroy humanity. She hadn't had any case to use it in her work, but we were brainstorming names together for a private practice therapy business that she wants to start. And she was like, oh, I want to add a botanical name, but also the word counseling, but one that's not being used right now. And I was like, oh, that's a great use of name your AI tool. And she was like, oh, well, I've heard AI is bad. And I was like, well, it depends how you use it. So we had this great discussion. We brainstormed a bunch of things in ChatGPT, it did a great job combining and looking online to make sure that these things weren't already registered with other businesses. And then she said, oh, I could probably use this at work to come up with icebreaker prompts for my group therapy sessions. And I was like, that is another great use of it. Let's talk about how to do that. And it was kind of cool to see the wheels spinning. Of course, we talked about the guidelines, make sure you don't use patient information in it, etc. But just to think about how there is not a binary, but there are so many use cases for how it can improve people's lives and the way that they learn and get creative together too. So you inspired me with the story of your daughter Anissa. I think that's a really important point. So after my monologue about AI and my cousin, Leo, I'm curious because you've represented the library perspective too, and we haven't had the opportunity to have a lot of librarians on the podcast. How do you see the role of librarians working with teaching and learning professionals and technology and also faculty at our institutions?

Leo S. Lo: Yeah, I think libraries and librarians can play a really major role in this. And I want to convey that to my colleagues because we are in a great position, Jenay already mentioned information literacy. This is part of that process to use AI. Well, you need to, at least at this point, you really need to know how to verify information, all the facts and all of that. And that's librarian's role in some ways. And search information discovery has become quite different now, if you have used Google recently, it's not just pages of links anymore. It's the AI summary at the very beginning. So AI is doing the search for you. They are giving you "the answer." How do they come up with that? Not sure, but at the same time, if you are going to use it, it's your responsibility to make sure that is accurate.

So that is the librarian's role to help them improve their information literacy in this kind of AI embedded world. I kept moving thinking bigger is that I feel like libraries, librarians, people with librarianship expertise should be part of that AI development to incorporate that into their model to begin with so they can follow those principles to give you, so I keep saying quote unquote the answers because a lot of things don't have answers. They have a lot of opinions and other things, but how do they even have that become more, not necessarily accurate, but that will make more sense that people can follow the logic of how they come up with those answers. And then how do you verify those if you want to. So I think librarians have a big role to play and also a lot of room to upskill to learn about how these things work, the technical part of it, and also the human part. How do we engage with our colleagues across the university, across our community, and then to teach our users? Because if the users are ahead of us, that's a problem. And I think we're facing that in so many different parts of our higher education.

Anissa Vega: Leo, I'd be interested to hear your thoughts on a conversation that I had with our library recently as an institution. We were having conversations about whether or not we needed an AI institutional agent who was an expert about Kennesaw State University. And in that conversation, the question of, well, who's going to curate the knowledge base that that AI will train on and use, and who's going to maintain that knowledge base? And librarians and skill sets associated with librarians seem to fit the best. But I haven't necessarily seen librarians have the opportunity or take the role in managing curation and management of knowledge bases for AI.

Leo S. Lo: Yeah, exactly. I absolutely agree with you, but it also depends on the resources of the library as well. So if you believe that, advocate for them to have the resources to do that, because I'll go to a lot of these big conferences or CNI and other places where I see the most well-resourced libraries are doing amazing things. I look at like Stanford, Yale, they're doing amazing things. They have millions of dollars to support that. But for most libraries, mine included, at UNM, we don't have that kind of resources. So one is to partner with different parts of the university. And also if people, and I think a lot of people agree that all libraries is actually in the perfect role to do a lot of these things, then we need to give them the support to do that, to do those things basically.

Anissa Vega: Interesting.

Jenay Robert: Since we're getting close to the end of our time together, I thought it might be nice to walk away with some concrete advice for listeners or people watching on YouTube or on the website. So maybe as a parting statement, can you give one or two pieces each of you, of concrete advice if someone is listening or watching and they want to improve AI literacy at their institution for their students, and maybe they don't know where to start. I know we had this theme of start small, small actions can have big impacts. I feel like that should just be the theme for Shop Talk in general now. But what are a couple of pieces of advice you would give to someone listening? Just they can log off and go start right now.

Jeanne Beatrix Law: I'm all,

Leo S. Lo: Oh, go ahead.

Jenay Robert: No, you Leo, you please. Please go ahead.

Leo S. Lo: Alright. Yeah, I wonder follow Jeanne already talked about platforms like Coursera and other places where they're very, they love information. I learned a lot from YouTube actually, because I think that YouTube people, they make a living of doing really engaging video to help people to watch the entire thing. Some of them are just so well made and so clearly explained. And I'll highly recommend people going to a platform like YouTube. And don't worry about, oh, it's not like Coursera. It's for not super academic, but you can actually learn a lot of really up-to-date stuff on YouTube. So I get a lot of my information there. And also, if you're not on LinkedIn, LinkedIn is a place that I learn a lot, get a lot of really cool new updates on AI from so those two places.

Jeanne Beatrix Law: For sure. And I might say too, and this is what I say a lot of times when I talk with faculty, is don't be afraid to fail because we fail every day all the time, and we learn from it. Start small, tinker and experiment alongside your students. And the last thing I would say is trust your students. That to me, that teach and trust is important when we think about academic integrity. Because what our data at Kennesaw State has shown is that the majority of students are not using generative AI in first year writing to cheat or to plagiarize that they themselves are getting back to us with these perceptions that they're telling us that they understand these nuanced use cases and they get very defensive when someone falsely accuses them of using generative AI to cheat. So I would say definitely trust students, experiment, tinker. And the last thing, if I could say, use open educational resources, reach out to faculty like the three of us who produce open educational resources that you can then adapt for your own context. That to me has been the most valuable thing is collaboration. And OER.

Anissa Vega: I would suggest reading a book. If you have not read Ethan Mollick's Co-Intelligence, that's a great read. I think that that framing AI as a partner is really important at our institutions. It is not necessarily an adversary or even a replacement of other infrastructure that we have. And then secondly, I would say set up play dates with your friends and your coworkers and play with AI.

Jenay Robert: Love that. In case any of our audience missed that, Jeanne did invite the entire audience to reach out to our three guests on LinkedIn and ask for OER. So please take them up on that offer and then you can all get mad at Jeanne later for that, for the flooding of your inbox. Yeah, I don't know. Sophie, any last thoughts from you?

Sophie White: No, I love the idea of a play date too. I feel like I haven't thought about setting up a play date for myself in a while, but I think that's a really fun idea. It does go back to, I think we've heard some discussion of, we've done a lot of intellectualizing around AI, but actually getting into the weeds and understanding how it works is one of the best ways to learn things. So all of your feedback about use YouTube, it doesn't, everything doesn't have to be peer reviewed, I think is really important. So thank you for making that point for us. And this was a really fun conversation. There's so much we could talk about, and I'm sure if we had this conversation six months from now, there'd be so much new knowledge. So maybe we can circle back to it. But appreciate all of you, the work that you're doing for supporting students, faculty, our institutions, and this world of AI. And thank you for joining the podcast today.