Frenemies & That's OK: Useful Tensions & Possibilities Between Libraries and AI

Getting to talk with librarians about AI!

Last week, I got to talk to one of my favorite groups of folks—librarians! I was invited to be a speaker for the University of Rhode Island’s Graduate School of Library and Information Science Annual Gathering. It was to be a 20-25 minute talk with some time for Q&A.

If you know me at all, you know that I love the library and librarians. They’re the first people I look to make friends with when I move somewhere or join a new institution.

As always, I had a Resource Link where you can find the slides and follow along.

Frenemies & That's OK

So yeah, I'm here to talk about AI and libraries.

Here’s the rundown of this 20–25 minute talk, and then we’ll jump into questions. One of the things I find most important about talking about and engaging with AI is that it needs to be conversational. One of the things I’m going to emphasize throughout is: go and try it. It can be scary. There are lots of concerns to think through. I know on my drive over here today, I had 20 different thoughts about AI.

But why me? I’m not a librarian. I don’t play one on TV or even TikTok, for that matter. I have been thinking about libraries for most of my life, since I was a kid when I may have been kicked out of a library or two. Mostly because I deserved it. It was not the place to perfect the art of chair surfing. Yet, I kept coming back.

I went to libraries in part because they were places where I could explore. The idea of moving from one place to another, exploring fiction, non-fiction; it’s where I found my love of audiobooks. I kid you not, I’ve given talks and written academic articles about audiobooks because I am that level of a nerd.

I’m such a big fan of libraries, I even brought one of my librarians here today. I say that jokingly. I met Robin when I did what I always do: I find the library. I make friends. They're always smart, thoughtful, caring people, always engaging, always giving me things to think about.

The library on the right is the one I grew up in and the one I got kicked out of. It’s where I found all those passions. For me, libraries have been spaces of exploration. And as the internet emerged, they made me think more about issues like access, copyright, piracy, cultural ownership, and power.

I’ve thought about this as it ended up being a focus of my dissertation. Even though my doctorate is in higher education, my focus was largely on libraries and access to knowledge. It was about academic piracy and what happens when we commercialize knowledge so much that scholars, from community colleges up through Ivy League schools, feel it’s necessary to use platforms like Sci-Hub and LibGen just to do their research.

So the idea of what libraries offer us, what they can do, and how they serve as collective repositories of cultural knowledge has been incredibly important to me. Especially now, when so much of that knowledge is being privatized.

The Tension of AI and Libraries

With the rise of AI, a new and interesting tension is emerging. As I said, on my drive over here I had 20 different thoughts. I’ve been in a lot of conversations about the role of libraries in the age of AI. I won't cover them all. I promise you’ll get home at a reasonable hour but there's a lot of opportunity, responsibility, and a real need for communication with your communities.

We need to talk about where AI fits in our lives and where it doesn’t.

Obviously, there are real concerns. We’re talking about disruptions to knowledge, authority, and reliability. The fact that AI can generate text that sounds eerily or strikingly human introduces challenges around misinformation and disinformation. It shifts what it means to be information literate. What does it mean in a world where so much information can now be manufactured in ways that were simply not possible five years ago?

And then there are the ethical and legal concerns bound up in these tools such as challenges to our current concepts of copyright. I find myself on both sides of this. I’m concerned for our artist friends, but I also have a deep critique of how copyright currently operates. Again, I just spent the last five years talking to scholars who said, “Because of copyright, I have to steal research literature in order to do my work.”

There’s a lot to say about what these tools offer. But there are also important considerations for librarians and communities across different institutions. You’ve been on the forefront of knowledge-based technological transitions. I don’t have to tell you that but I don’t think we always realize it, either.

We’ve spent the last five months or five years, or even 15 years trying to figure out each new wave of chaos coming our way. Taking on AI feels like yet another existential, daily challenge. But if there’s a place that can figure it out; if there’s a place where it’s critical to do that work on the community level, it’s the library.

One of the most powerful things I’ve seen happening in libraries is their pivotal role in supporting workforce transitions. We already know this shift is underway. AI is going to be part of that story; both as a challenge to employment opportunities and also, how it helps people level up and access opportunities.

And I’ve seen it. I’ve seen students do it. I’ve seen adults do it.

Problematic as it is, GenAI can often be really empowering to people. When you have a tool that can talk to you the way you need it to, explain things in a way that makes sense to you, give you the time of day that often is hard to get from humans; that’s transformative.

I think about my partner. She’s brilliant, but she’s never been a big writer. She’s more of a verbal thinker. When it comes time to look at a screen and write, her brain just freezes. It’s not because she isn’t intelligent or trying; she has an advanced degree and can absolutely do it. But it’s a modality of diminishing returns for her. But generative AI has helped her build a business, solve problems, and serve as a dialogic partner. It’s helped her figure things out, and that has been phenomenal to see how people can leverage GenAI’s affordances to make their work shine.

I think there are a lot of people for whom a tool like this opens doors they haven’t been able to access before.

One of my favorite examples is helping people write cover letters. I’m biased, but I think the cover letter is the most ridiculous form of writing there is. And I say that as someone who writes good ones. I know because of my job interview success rate. But they’re rhetorically obnoxious. You’re often performing these feats of persuasion that have very little to do with the actual job you’re applying for.

Using AI to help people communicate their values and abilities to an organization—earnestly, not fraudulently—can be transformational. When people use it to clarify who they are, what they care about, and how they fit with an opportunity, that’s powerful.

Academic Libraries

Now, I want to pivot briefly. I know not everyone here is an academic librarian, but much of my work has been in that space. I’d be doing a disservice not to acknowledge the issues and opportunities AI presents in higher education.

Librarians are navigating shifting definitions of academic integrity and what to make out of AI outputs. We have shifting grounds of knowledge, but that is familiar territory too. In the past, every new technology, like the move to online journals, was viewed with suspicion. Online publications weren’t “real” until they were printed. So now, we’re having to rethink what these ideas mean in this AI era.

We’re also dealing with tool evaluation and tool communication. It’s been just over two and a half years since ChatGPT came out, and it feels like every day there's a new update, a new tool, a new affordance. It's exhausting. And that’s in addition to the fact that libraries, both public and academic, are perpetually understaffed.

Then there's the question of literacies. What are the new literacies we need to teach? How do we rethink media literacy, information literacy, and now, AI literacy? And of course, data and privacy concerns are a huge piece of that.

But it’s not all challenges; there are some significant possibilities. One is enhanced research support. A tool like Google’s NotebookLM changes the game for how people can organize and validate research, more effectively than something like ChatGPT, in some cases.

Accessibility is another area of transformation. We’re seeing AI actively at work in live-captioning what I’m saying right now. These tools are making the world more legible for many people, not just those with visual or hearing disabilities, but also those with cognitive differences, or those who prefer different modalities of communication. That’s powerful. AI can also support collection development and metadata processing.

Public Libraries

In public libraries, however, we’ll also need to keep navigating the equity and digital divide: who gets access to which tools?

We haven’t solved the digital divide; we just act like we have. There are still millions of people in the U.S. without real internet access. And even among those who are online, access to quality AI tools is uneven. Paid versions offer significantly more capabilities such as unlimited queries or advanced features. Free tools are great, but limited, and often require users to figure out more on their own.

This raises questions: When does AI become ubiquitous? When do we treat it as a utility? We haven’t done that yet even with the internet, though other countries have. But if AI becomes necessary infrastructure, we’ll need to start having that conversation.

Then there’s the ongoing battle with misinformation. Been there, done that but it’s not going away. We also have to grapple with privacy, surveillance, and deciding when and where those technologies are appropriate.

Sustainability is another concern. These tools consume massive amounts of energy. But to be fair, even pre-AI digital infrastructure had a significant environmental toll. We just haven’t collectively decided how to address it in general so AI is just one more piece of that concern.

Despite all this, the opportunities remain immense. AI can support learning in communities, allowing people to do more, understand more, and find new, creative ways to engage.

One example I came across recently involved storytelling: using AI with children to co-create stories in real-time, or to generate character visuals and narrations from student-created content. That’s the kind of fun and creative possibility I come to see libraries already doing with other technologies.

Multilingual support is another major win. As concerned as I am about access, the ability to instantly translate or transcribe into different languages, mediocre though it can be at times, is game-changing. Accessibility has often been left behind because it’s perceived or actually is more work. But when AI can automatically generate transcripts or captions, that lowers the barrier significantly and changes who can be part of the conversation.

Libraries also stand to benefit from operational efficiencies. AI can help with repetitive, tedious, or time-consuming tasks. That can free up time and energy for more impactful, human-centered work.

Prompts To Try

(All of these can be found in the slide deck and prompt library here.)

I’d like to shift into some practical suggestions. I always try to leave people with something they can take home, something tangible to try. I'm going to recommend a few prompts. All the slides and prompts are in the resource I shared. It’ll be on the last slide. It’s a 50-page prompt library, covered under a Creative Commons license.

Here’s one example: You’ve got a patron who keeps maxing out your recommendations; they’re in every week asking for something new. It’s nice to have an enthusiastic patron but also, it gets exhausting. Instead of digging through years of Library Journal issues, you can throw a few smart prompts into a tool and quickly generate a great set of recommendations.

This example was originally a full list of ten, but I’m just sharing three, because, well, space and time.

What do you think about these recommendations? You might consider grabbing one of them or seeing which one you already have in your system. You might be like, “Oh, you definitely want to check this out; here’s one that builds on what you’re looking for.”

Let’s look at planning. We're all inundated with building out reports, documents, and the like. Using something like generative AI to start building a draft can be helpful. One prompt example: drafting a new five-year strategic plan that aligns with community priorities and addresses emerging trends. It can help create a framework that includes measurable goals and assessment mechanisms. You can use AI to build the template, but also to bring in other documents such examples of mission statements from other organizations, or existing city or state programs you want to align with.

The important part is to always treat it as a draft; that’s the most useful way of dealing with the outputs. I’m not going to take whatever it produces and just hand it in somewhere. But it gets me started faster. I’ve been a writer for 20 years, and the scariest thing in my life is still that white screen. No text. Just sitting there staring at the empty canvas. So having a jump start, especially when it’s a topic I know a lot about but don’t want to begin from scratch; that’s quite valuable.

It can also help with your library’s storytelling, taking data and helping you figure out how best to present it. How do I pitch this? How do I make this resonate with a particular audience? You can feed in budget considerations or other context, and then say: help me craft compelling stories around this. Basically, you’re remixing what’s already there, seeing what it produces, and deciding what you want to keep and what to discard.

Now, policy. And I know, who enjoys policy? Okay, right, we’ve got a few. That’s fine. You can enjoy policy. But for a lot of us, it’s hard to get into the weeds. Using AI here as a starting tool can be build upon other documentation that you have or want to build from. It can help you generate the legalese or performative language you need, and then you, or someone on your team, can come in later to massage, refine, and qualify. It’s much easier than starting on page one.

Say there’s a new concern or issue, something that requires a policy update. You can ask AI: how do we update this effectively?

Usage Tips

Those are we some prompt ideas to explore. Here, I want to offer some basic usage tips; things we should all be doing and sharing with patrons to help them level up how they’re using these tools.

First tip: ask for multiple answers. When I use AI, I often ask it for three different versions. I may like one best, or I may like parts of all three. Then I’ll say: combine this from version one, this from version two, and this from version three and give me a new version. It saves you from picking the first response and helps you shape something more personalized.

One of the challenges in understanding AI is that it performs something that looks like thinking. You input a question, it gives you a response, and it feels human. We know it’s not but we trick ourselves into believing it is. How many people have named their car? Many of us! We anthropomorphize everything, so of course we’ll do that with AI.

But we need to remember: it’s not human. That said, because we feel like it’s human, we often fail to ask it questions that might push us into new thinking. We assume it thinks like us and it doesn’t. That’s kind of a meta point, but think of it this way: if you were talking to another species entirely, you wouldn’t know what their theory of mind is. With AI, it doesn’t have a theory of mind, but it does construct and make sense of the world in a way that’s different from ours.

To access that difference, we have to learn to think differently with it.

Here’s an example: If someone asked me, “What did you do wrong in that last presentation?” I’d probably say, “Nothing! It was perfect.” Right? Or I might be overly self-critical, but I might still miss the mark. That’s why we tell writers: step away from your work for a few days and come back with fresh eyes.

With AI, you can do something unusual: after it gives you a response, you can say, “Now tell me what was wrong with what you just wrote.” And it will. It’ll critique its own output. That’s wild and we can’t really do that with ourselves in real time. That’s what I mean by saying its “theory of mind” is different, and that opens up possibilities we often overlook.

One of the most transformative ways to use AI isn’t to get it to give you an answer; it’s to have it interview you.

Let’s say you have a project or a piece of writing you’re struggling with. Ask AI to interview you about it. Say: “I want you to interview me about subject X.” Maybe it’s a grant proposal, a research project, a community initiative. Tell it who your audience is and what you're trying to do.

Then, let it guide you through the process with questions. It becomes a dialogic tool, not just a content generator. It helps you reflect, organize your thinking, and synthesize your ideas. In the higher ed world, this is one of the most exciting uses we’re seeing: AI as a reflection partner, a simulator for intellectual exploration.

Another thing I’ve encouraged and even run workshops on is using AI to support people who are seeking employment. For example: take your resume, take the job application, give it to AI, and say, “Interview me as if you’re the hiring manager.”

This helps people prepare and engage. It encourages them to reflect on what they actually know and how they want to present it. Again, it moves us away from that concern about “cognitive laziness” and into a space where the tool is actively helping.

One of the most transformative things for me personally again, I’ve been writing since high school (not always well!) and professionally for about 20 years now, since 2004 or 2005 is how AI helps me get past that moment of staring at a blank screen.

One of my favorite workflows now is to flip the script. I go to AI and say something like, “Hey, I’m trying to figure out XYZ. I’m giving a talk on this topic. Interview me about my ideas, what I’m thinking about, how I’m putting it all together.”

I use the paid version of ChatGPT, which has voice functionality. I’ll open it on my phone, turn on the voice setting, and have a back-and-forth conversation with it. What’s really cool is that I can do this while I’m walking, driving, doing dishes, whatever. Then, when I get back to my computer, everything I said has been transcribed into text. I now have large chunks of the talk or piece I’m writing already drafted.

I also use this approach with other people. We’ll get into a Zoom room, turn on transcription, exchange our ideas around a shared project, and then use that transcript. AI can clean it up, and suddenly, we have usable blocks of text. It’s a powerful example of how this enhances a workflow not by replacing thinking, but by freeing us from the screen and helping us get into the flow of our ideas.

And here’s the last point. If you leave here with only one thing, especially if you haven’t already tried this, this is the tip to take home and share with patrons:

If you don’t know how to ask something, or you’re not getting good results from AI, use this prompt:

“Revise the following prompt to maximize the output of a large language model: ”

Then, after the colon, put in what you want, for example, “I want a strategic plan for the library.”

The AI will respond with an improved version of your prompt. You take that revised prompt, paste it into a new chat, and run it.

What you’ve done is ask the AI: “Hey, you know how to talk to AI better than I do. Help me phrase this in a way that makes sense to you.”

This does two things:

1. It gives you a better prompt more quickly, which leads to better results.

2. It helps you understand how the model works, how it reads your inputs, how it structures responses, and how it interprets intent.

It’s a simple technique, but it delivers strong returns. I use it regularly, especially when I’m working through something complex and I want to ensure I’m getting the strongest, most targeted responses possible.

So those are my tips and I think I’m at time. Thank you so much!

---

The Update Space

Upcoming Sightings & Shenanigans

• EDUCAUSE Online Program: Teaching with AI. Facilitating sessions: April 7–18, 2025, April 28–May 9, 2025, June 23–July 3, 2025

• Panelist with Nicole Allen & Robert Awkward, Ph.D. at the Open Ed in 2030: Exploring the Possibilities and Challenges Ahead of Us by Bridgewater State University on Monday, May 12th, 2025.

• NERCOMP: Thought Partner Program: Navigating a Career in Higher Education. Monday June 2, 9, and 16, 2025 from 3pm-4pm (ET).

Recently Recorded Panels, Talks, & Conversations

• “Growing Orchids Amid Dandelions” in Inside Higher Ed, co-authored with JT Torres & Deborah Kronenberg.

• Bristol Community College Professional Day. My talk on “DestAIbilizing or EnAIbling?“ is available to watch.

• OE Week Live! March 5 Open Exchange on AI with Jonathan Poritz (Independent Consultant in Open Education), Amy Collier and Tom Woodward (Middlebury College), Alegria Ribadeneira (Colorado State University - Pueblo) & Liza Long (College of Western Idaho)

• Reclaim Hosting TV: Technology & Society: Generative AI with Autumm Caines

• 2024 Open Education Conference Recording (recently posted from October 2024): Openness As Attitude, Vulnerability as Practice: Finding Our Way With GenAI Maha Bali & Anna Mills

AI Policy Resources

• AI Syllabi Policy Repository: 165+ policies (always looking for more- submit your AI syllabus policy here)

• AI Institutional Policy Repository: 17 policies (always looking for more- submit your AI syllabus policy here)

---

AI+Edu=Simplified by Lance Eaton is licensed under Attribution-ShareAlike 4.0 International