How Special Librarians Can Put Generative AI to Work

Hello, everyone, and thank you for joining us for today’s webinar with Lauren Hayes. My name is Bradley and I’ll be your moderator for this webinar titled How Special Librarians Can Put Generative AI to Work.

Before we start, I would like to provide some information about our company and introduce today’s presenter. Lucidea is a software developing company specialized in museum and archival collections management solutions, as well as knowledge management and library automation systems. Our brands include Sydney, Presto, Argus, ArchivEra, Eloquent, and CuadraSTAR.

Now I would like to take a moment to introduce today’s presenter, Lauren Hayes. Lauren is an assistant professor of instructional technology at the University of Central Missouri. Previously, she worked as an instructional and research librarian at a private college in the Kansas City metro area. Prior to working in higher education, she was employed by the National Archives and Records Administration and worked as an intern at the Harry s Truman Presidential Library and Museum. Her professional interests include the scholarship of teaching and learning, information literacy, digital literacy, educational technology, and academic development. Take it away, Lauren.

Thank you, Bradley. Thank you for that introduction, and thank you everyone who is listening in and joining in on this webinar. I am excited to be here with you all and to talk about generative AI in special libraries. This is a topic that I am sure has, probably been discussed a lot amongst your colleagues and amongst peers at other organizations, so I hope that we can this is a useful presentation for you and just you can think about some practical ways to use generative AI in your context.

So, I want to start out with a definition of generative AI. At this point, I’m not sure that it’s necessarily needed, but I like to make sure that everybody’s on the same page, just when we’re talking about generative AI, what it is and what it is not. And so this definition comes from Wikipedia, but it is generative artificial intelligence, the subset of artificial intelligence that uses generative models to produce text, images, videos, or other forms of data.

And these models learn the underlying patterns and structures of their training data and use them to produce new data based on the input, which often comes in the form of a natural language prompt. So in other words, when you type in a prompt into the generative AI you’re using, it, the model has all that, like, the training data, which is the large language model, which you’ve likely heard of the LLMs, and then it creates data and an output, based on the data that it’s trained on and the question or your input that you asked.

And so there are a lot of variations when it comes to artificial intelligence broadly.

There are, you know, this Niro AI, which has been around for quite some time, everything from, you know, my Roomba vacuum to, you know, Siri, Alexa. You know, they’re designed to perform a specific task, even, facial recognition. When I look at my phone, it opens just because it recognizes my face. So generative AI is certainly a big leap beyond that.

And, because it can be it’s so almost human like in what it can do. So that’s really where a lot of the conversation is around, and that’s where we’re going to focus our time here today.

So here are some examples of generative AI tools.

I kind of organized them into some text based outputs and image or video outputs.

But that said, the, you know, paid versions of ChatGPT now can create the images, for you and do pretty well. I was just using it last week to create some images and was and asking it and refined it in a lot of various ways, and it was doing a really nice job of creating images. So a lot of the tools that the primary ones, the ChatGPT, Gemini, Copilot, Claude, they’re moving into also being able to be image, not quite as much video output yet, but at least image generators along with the textual, content that they can create. That said, there are some specific generative AI image and video creators as well that if you’re looking for kind of a variety of different tools that you just wanna be familiar with what’s out there, I encourage you to explore and look at some of those as well.

As you are looking at these different tools and, getting a feel for what is unique about each of them, one thing that I think you will notice is that they each especially with those text based ones, they each have their own voice. The outputs, there are some kind of telltale signs with what might come from chat g p t versus what might come from Claude.

Personally, I find Claude to be a little bit more academic in voice.

ChatGPT tends to be a little bit more for the average reader, and there can be reasons I wanna use ChatGPT or reasons I wanna use Claude, for example.

And, so you’ll just think about that as you kind of explore them. I think you’ll become very familiar with the kind of the different ways and the different outputs that they give you.

And then, of course, within these, there are the free versions and there’s also paid versions of them. The paid versions often do a lot more than the free version is going to be able to. And if you have a free paid version whether, personally or through the organization that you work with, let’s say you, you know, have used the Google Suite, and therefore had Gemini built in or if you use the Microsoft suite and had Copilot built in, you would be able to do a lot more with that kind of enterprise, paid license than you would do would just be able to do with the free version.

So one specific tool that I want to talk about today and kind of think about some examples for it in the special library context is notebook l m from Google.

And this is a really amazing tool in that you can add in PDFs, websites, YouTube videos, audio files, slide decks, all this different content that you can put in, and then the notebook l m will look at those specific items that you put together and create outputs.

We can create some really interesting podcasts.

It can summarize documents. So, you know, all the documents look for themes across all of those documents.

And, one way, that somebody that I would I work with or know was telling me that they were using it the other day is that they work for, an organization where the kind of CEO is trying to build something new, and this person was trying to explain the CEO’s specific leadership style. And so they put in PDFs, blog posts that the CEO had written, a few videos that had been recorded and kind of various documents and content that the CEO had created over many years, uploaded them all into NotebookLM, and then asked it to create a podcast discussing this individual’s leadership style.

And, it popped out a podcast with two individuals, AI generated individuals, discussing this person’s leadership style. And the person I know listened to it and felt that it was a very accurate representation of how this CEO works.

And so I thought it was just fascinating to think about all the different ways that we can use it to think about kind of knowledge management in our organization.

So when you were thinking about trying to manage information, sometimes you just wanna get a lot of information together in one spot, look for themes, look for, you know, what is happening in a particular area, what might not be as prevalent in another area, so that there is would where might the notebook LM recommend adding content, etcetera. And so that could be a really interesting way of thinking about it.

So I encourage you to maybe have explore notebook l m if you haven’t, because it is going I think it can do a lot of really interesting things.

It can also some more some kind of specific things it can do.

You can create literature reviews based on the specific, like, uploaded documents.

It can summarize longer documents you want.

It can ask clarifying questions.

It will create citations for you.

Of course, everything needs to be double checked and that human oversight is really important, but this can be a way of, you know, getting some information as a starting place. It could, you know, help draft some initial blog posts if you’re wanting to create a blog post, you know, off of some internal documentation, for example.

You want to explain how to use a particular database within the library or get access or help people understand how to get access to particular information from within the library, and it could help, you would upload maybe the training documents and it could make those in a more user friendly format for you.

It can also, you know, create bullet pointed summaries of meeting notes. It will kind of extract action items from particular planning documents or additional meeting notes, and it can kind of compare multiple vendor proposals, all sorts of different things. Of course, with any of these use cases, you need to be aware of copyright, issues or, you know, privacy violations. So having some conversations with, like, the general counsel of your organization or whoever that person is that you would need to have a conversation with HR to make sure that if you’re using it within, you know, your organization that you’re using it correctly and within the bounds of what is ethical for your organization.

So we’ve talked to kind of about ChatGPT, Gemini, Cloud, Copilot, NotebookLM.

These are kind of big overarching tools that are really good for standard things that we want to do. They do very specific things as well, but they are made to do a lot of things.

There are specific research AI examples. Now these this this list is not exhaustive, and it does come from a blog post that was published on the Lucideas blog, a few months ago, so you can read more about these there. But there are specific AI research tools that can help individuals, you know, researchers find journal articles, find, you know, patents.

I, you know, I like the idea of the connected papers where it’s exploring academic papers in a visual graph just to really kind of see where the connections are. You know, for those of us in the library world, we think a lot about, you know, citation chaining and understanding the con the scholarly conversations that occur within academic literature, what information was cited, what’s going to, you know, cite the article that I’m looking at in in the future.

And so having that visual graph can be really nice with connected papers.

A lot of people really like perplexity, but I encourage you to explore these individual kind of research AI tools because there’s quite a few out there, and, they’re gaining a lot of traction within, you know, some circles. But at the same time, there are the broader tools that are gaining a lot of, you know, traction and space in broader broader communities.

So, again, those chat GPTs, the clods.

And so what you want to think and compare, do I want a specific, you know, research AI, or can ChatGPT do what I need it to do? And ChatGPT has a lot more deep research tools now that it didn’t used to have. Originally, a lot of these tools were hallucinating. They were making up sources that didn’t exist.

Most of that has now been corrected. The tools are now, finding accurate information. They can pull information directly from the web that we just published today, and there isn’t as much concern about that. And especially with those deep research tools that Chatt GPT and others have in their wheelhouse now, you’re able to really do a lot of good research.

But I encourage you to explore and see what works best for your organization and for your use in the library.

So we’ve I’ve given some good examples, hopefully, of already how special librarians, how you can be using generative AI, but I do want to give some more specific examples here in the next few slides moving forward.

So as you’re thinking about the different AI tools and what they can do, one thing that I think gets it talked about a lot in, the conversation is just how is generative AI going to transform what we do.

A lot of times it gets discussed as a transformative technology, and there’s a good reason for that. Right? There it does it’s going it is likely to change a lot of how we work, and a lot of what we are able to do and get done in the future.

So kind of just some high level overview of how it could transform services. It can help with automation of routine tasks, streamline content creation, and create new services.

So we are going to kind of look at these in a bit more detail, but, I what I want you to think about with this idea of transforming services is not that AI should take the place of people or take the place of human oversight Because I think that is part of the conversation as well that, there’s some concern, right, about what, AI can do. And, ultimately, that’s really that conversation is beyond the scope of this presentation. But what I hope, I am able to share with my work is that AI can be a tool to assist with what we do. And if we use and I was just reading a blog post earlier today from Ethan Moloch, who is a professor at the University of Pennsylvania and writes a lot about AI, and he was saying that, you know, AI can make us lazy if we use it in lazy ways, but it can also really help us and help us be more productive if we use it in the right ways.

So that’s where I think we need to be thinking about AI is how can we use it in the right ways to make us more productive, help us gain new insights, but not losing, you know, who we are and what we are as, as librarians, as people.

So just be thinking about that as we have some of these more detailed discussions about know, automating of routine tasks, streamlining content creation, and creating new services.

So when you’re thinking about that list of what else it can do, this is specifically from Gemini, and it says, okay. It can help you write emails, refine work, and streamline tasks. So that can be some of that, like automating and processes.

So let me give an example of how I find myself often using, either Claude or ChatGPT at this point. So I do a lot of writing in my work, and I have and I spend a lot of time thinking about wordsmithing sentences and how to best to make sure that I am conveying the idea as clearly as I possibly can in the way that I want to convey that idea.

And there’s a lot of value in thinking about that and kind of working with the words and making sure that they are put together in a way that is clear and concise and accurate.

But I’ve also found that sometimes I struggle with that. And even when I was writing my dissertation, I worked with an editor, to ensure that my ideas were well thought out. And I have found that it’s really hard sometimes to do that on, you know, when I’m writing a lot of emails every day or I’m, you know, doing some of those tasks that are not quite as serious as writing my dissertation or writing another paper at this point.

And so I have found that if I’m writing something and I want to just kind of words with it a little bit, I can ask Claude or ChatGPT, give me five other ways I could say this, especially if I’m not happy with what I just wrote. And it will do that, and it doesn’t get tired. Like, that’s the really nice thing. You could ask it for five ways.

You could ask it for ten ways. You could ask it for fifty ways. You might not wanna read all fifty, examples that it gave you, but it would give you all of those. And so that can just be one way to use it to help you get something done a little bit faster, but also help you create a you know, meet a goal.

So, hopefully, that’s helpful. I want so I want you to think about that in your context. So where are those things that you sometimes feel, like, take up a lot more time than maybe they need to, and there’s still important tasks. It’s important that I respond to email. It’s important that I’m communicating clearly, but that, an AI tool might just give you a little bit of a support for what you’re needing to do and do do it well.

So something else to think about is with optimizing workflows.

So it can, you know, create it can automate content generation with summarizing data or research, drafting and creating draft documents, policy, operating procedures, reports.

And so this is where I want you, and if you’re having conversations in your organizations too to think about how the way that your organization wants to use AI.

So you could either have the AI tool, like, draft an initial policy or operating procedures, for example. Sometimes people like something to respond to, and then you can see what AI drafted. Then you say, oh, well, no. I don’t like this. I don’t like this, and you could make those changes.

Alternatively, you could do the initial drafting and then get feedback from others and then have AI take that feedback and make the changes to the document.

So, again, there are some different ways to approach the use of AI in your work, whether it is helping you get started with something, it’s helping you refine something, it’s helping you get unstuck.

Think about what that means in your context and where you would want to use AI and where it’s appropriate for what you’re wanting to do.

Then for, like, summarizing data and research.

So this is also going to vary in how you use it based on ethical principles. Right? So is the data that you have subject to IRB approval, institutional review board? And did the institutional review board give you the, you know, give you the permission to be able to use AI to work with those themes? If they don’t if you don’t have permission, then you should definitely not upload the data into an AI system.

But if it’s not subject to an ethical review board, and it’s not you know, you’re not sharing company secrets or anything like that or, you know, any sort of competitive intelligence, there might be ways where you do want to upload data and have it summarize, for you. Alternatively, again, you could do some initial summarizing of the data and then have it check your work or see if it finds something different than you did.

So that could be a way even if you’re thinking about, like, if you’re doing conducting some literature reviews for someone at your organization, you could, you know, do what you normally do for literature reviews and then just have the AI tool say, is there anything else that it finds that maybe you didn’t find? And you could upload, you know, the citations and say, I found all of these. Is there anything else that you would recommend, you being the AI system would recommend for this topic? And you can see what it comes up with, and that can also just kinda help you get an idea of, is this a useful tool for what I need to do in my work, or is it not? And what I do recommend is you’re you’re kind of testing those out is to go back and try those same thing the same thing every few months because the AI tools are jumping ahead so quickly, and they are changing so fast that, what you do today is likely to look very different than the output that you might get in a couple months.

But maybe the jump and maybe the change isn’t, a big enough shift even in two months for you to want to use it. Try again in another two to three months and see if that change, because I mentioned earlier with hallucinations that originally, when generative AI was first kind of starting to take the world by storm, there was a lot of hallucinations. It wasn’t giving accurate content that has changed so much. You know, before, it also you know, you could say, well, if you give it a new topic, it won’t be able to work with it very well because it doesn’t have access to the the live Internet. Well, that’s not true anymore either. It does pull information directly from today. So you just have to keep using it and keep understanding where it’s getting better and where it’s improving to know in what ways it can best really help your workflow.

Something else to think about is, you know, metadata generation and cataloging.

So there are tools, that can you know, create some of this some metadata for you, cataloging records. It can help you classify documents.

This can be really useful if you have a big, you know, a backlog of, you know, scanned documents or physical documents that need cataloging.

Of course, it’s still really important for that human oversight for these things. So you want to, again, think about where can the AI help, where does the hue where is the human needed, and where do we want the human involved, and really think through that kind of marrying of the AI and the human work together and think about where it can be best. And that’s good having some conversations and testing it out, the workflows, and see what really works in your context.

I’m gonna come back to this metadata generation and cataloging in another slide, so kind of pause on that for now.

But then also thinking about optimizing workflows, think about search and discovery.

You use AI, you know, research AI tools to locate content. You can partner with AI to refine your search queries, especially individuals who may not be as familiar with, research as we are in special libraries, it can be really beneficial for them. And so I think having conversations, especially if your library is one that does instruction and maybe does some workshops or does video tutorials, creating tutorials on how to use AI to refine search queries, to locate content, especially if your organization is one that has an AI tool embedded into the, you know, the suite that you’re using, those are going to be really important because your individuals in your organization are are going to start using these AI tools for research, and I think getting ahead of some of that conversation so that they’re used well and that the librarians can be involved in some of those discussions about how the AI tool is used for research is going to be useful and I would encourage you to do that.

So I talked a little bit already about notebook l m, but just, this is kind of just a visual overview of what it looks like. You could upload your documents, ask it to do, you know, various things, and then it will give you the output, over on that kind of right side. So I’ve got my sources on my left, the chat where I’m talking to Notebook Ellen asking it to do various things in the center, and then the output, which is the studio. You’ve got an audio overview.

You know, it can give you a transcript, those things. So be thinking about, just the various documents you could upload, how you could create tools. I was just talking about, for research, if your organization has, you know, kind of embedded, whether it’s Gemini or Copilot, embedded in your org, in your, office suite, then think about maybe using some of these tools like NotebookLM to create a podcast or to create, some summary documentation on how these AI tools can best be used in your context, then you could upload specific, you know, examples of specific documents, documentations that that you want to just to be this the sources that and what’s great about notebook l m is it will only look at the sources you provide. It’s not going to be pulling in from a large language model where you’re not sure where the information is coming from.

Okay. So I said I was going to circle back to, the conversation about metadata.

So I asked, you know, ChatGPT to create a sample mark record for a book, and this is what it popped out. And, you know, it did specifically say the details are fabricated to illustrate a typical MARC format, so please verify with the actual source.

So that’s really important. It is telling you that you need to verify that we can’t just trust, what is created from ChatGPT for the mark record, but it will create these tools and these, you know, resources for us.

And so just be aware that these things are happening and can be done with the existing generative AI that is out there.

And then this is another one. This is, an example of a marked record for Cat in the Hat by doctor Seuss.

And this specifically said, you know, the record is based on a typical cataloging format, and you should verify the actual book for accuracy.

I will say that I kind of played around with this a little bit.

I spelled Seuss incorrectly.

You know, I tried to see what it would give me if I didn’t put in some of the information correctly, and it would still bounce back, a mark record. It fixed the spelling for me at one time, for Seuss, and I just thought it was really interesting.

I’m not a cataloger by background.

I have done some technical services work over the years.

So but I just think it was for checking all of this, it was interesting, and I like how it said, you know, check it for accuracy. I thought it was interesting how I could it would switch spelling for me.

And there is also a, like, a a plugin for chat GPT, like cataloger GPT that you might want to explore, for creating more cataloging records as well.

Just being aware that those things exist out there, I think, is really important.

Then a little bit more about optimizing workflows in your special library.

So, another thing to do is for data analysis. I’ve already talked, about the ethics involved if either data that you have is subject to, you know, trade secrets or ethics review boards, things like that.

But, what can be if it’s not, it can be really useful for doing some analysis of even, like, library services. So, for example, let’s say you’re wanting to analyze, you know, library usage statistics or what what are the types of queries that that you’re getting in your library, what’s common.

Maybe you’re wanting to see what sources are primarily used. Maybe you’re wanting to see, what at what time of day, a lot of requests are coming in. And is there anything, you know, specifically happening in the organization at that time of day or at that time of the year that would be impacting, the types of queries that are coming in so that you could better prepare, you know, kind of rework schedules so that the library is staffed up correctly to be able to support the needs, that are happening at that time. So you could, you know, give a lot of information to, these AI tools to analyze so that you can make some informed decisions about what’s happening in your library, what’s happening when, and how you might want to adjust based on the data that you have available and that it is able to analyze for you because it can look for those correlations.

It can give you possible reasons of why something might be happening at a certain time, and it can of course it can create graphs and charts and tables based on data.

It can also, you know, create those personalized recommendations, especially if you can give it, information about your library, about, kind of the organization that you work. It will customize the recommendations that it gives. That is one thing that I found, you know, the more information sometimes that I can give these tools, the better the output is and the more specific. Otherwise, it just the the output seemed to be very generic, and there can be there can be value, of course, in getting general information, but if I’m really wanting to optimize my specific workflow or understand my specific library and organization, The information that I share with generative AI, the more specific it can be, the better it’s going to be able to give me, you know, those personalized recommendations.

So just think about that.

Think about what you want to give it, what you would want to hold back, you know, from an ethics standpoint, what you’re not comfortable sharing, what, you know, the organization might not be comfortable having shared, but also think about what could be really useful to share to be able to look for, you know, key trends in some of the data that you have and how you might be able to change what you’re doing to better meet meet the needs of your organization.

And that’s ultimately the goal, right, that we are looking for things that we can do with generative AI so that we can better serve, better work with the people, and provide better services, for those that we work with. And so with that in mind, think about what you do and identify ways that Generative AI can act as a partner.

Just where does it come alongside? Where can it get included in your workflow so that it is a partner in what you’re doing?

So I mentioned that, you know, that it can create, some different tools.

So I, you know, I I said it helps me create an Excel formula.

I this is from Claude. I wanted to give it some things.

You I could upload a spreadsheet, and it would give me some data.

I’m gonna kinda give you a little bit of time to kind of read through what that says, but how you can ask very specific details to make sure that you are getting what you want.

It says break down the formula and its components explaining the purpose and function of each part and how they work together. Additionally, provide any necessary context or tips for using the formula effectively within an Excel worksheet. So all those things that you can do is going to be helpful in getting the output that you want.

And then, you know, ultimately, what we’re wanting to do here is to deliver greater value.

You know, provide increased efficiency and productivity, think about cost savings for your organization, think about enhancing data driven decision making, you know, enhancing the presentation of content and resources that you have available within your library.

And when you’re when you do these things, the the goal is that it allows for more human work to be completed. And what I mean by that human work is that it allows the you as librarians to make you know, have more maybe more time for, discussions with individuals at your organizations to really understand their information need, more time for that, you know, engaged in-depth reference interview.

It could, you know, allow new work to be created, and maybe you’re thinking, I don’t have time from new work.

But maybe the gen the generative AI is doing some of these other tasks that there would be space to think about, okay. What else can we do as a library to provide services? Are there places, or things that you’ve been wanting to do, as librarians for your organization or within your library, but just haven’t had time to be able to do because you’ve been focused, on some of these other very important tasks.

And so if there are ways to have AI as a partner in some of those tasks, would that free up space to be able to have, you know, whether it’s more, you know, those more in-depth reference interviews, providing more, you know, specialized tutorials, providing even more in-depth research, and being able to really dig even deeper into the research that exists, putting together, you know, more, you know, systematic reviews and, you know, looking for literature that, maybe isn’t as freely available, but is, you know, more gray literature, things that isn’t is readily available, in databases. So so you could have time to conduct more, research for your organization, packaging it in new ways, you know, sharing that data in a different format than maybe what you’ve been able to do previously.

These are just some ideas for what you could potentially do with time that generative AI would free up. And so I encourage you to think about what it is at your in your special library that you would like to be able to have time to do or you know is an important part of what you could be doing, that you could do with a partner of generative AI.

And so that kind of I the I don’t know if it’s a great segue, but I wanna segue in the segue into kind of that importance of human expertise. I’ve already hit on that some with, you know, thinking of the value of human work.

But whenever I talk about generative AI, I never want it to come across that.

I think the tool can do all these things, and we should offload a lot of stuff to the tool. There is things I think we should offload it to generative AI, because it does it really well. But I also think it’s incredibly important that we, you know, value that the humans that we’re working with, and so we bring a lot to the table.

So here are some things that as you’re having conversations and you’re thinking about generative AI in your context, you know, what is what do we do as humans that’s particularly good?

We can ask the right questions because we can understand the emotions of a room. We can kind of read the room. We can hopefully ask the right questions and ask the generative AI what we really need. It’s not going to give us an output if we don’t ask a question. It we also have knowledge of the organization. We have historical knowledge.

Hopefully, the, you know, the knowledge of your organization is being managed in a way that that information comes to the forefront when decisions are made and the information that you’re giving is within a context.

Critical thinking. This is something that I really like to talk about and I’ve written a couple blog posts on critical thinking and with generative AI, but I’ll just kind of mention one thing here that, you know, ultimately I think the best users of AI are going to be individuals with a pretty good depth of content knowledge. And the reason I say that is because if you don’t have a lot of content knowledge, you’re not going to be able to evaluate the outputs of AI to know if something is correct or something isn’t correct. And you might not even be able to know if what you need. You might not be able to go back and ask that right question if you don’t have a good understanding of the content or the organization.

And so if we do have a good, you know, understanding of the content and the organization and the context in which you’re working, you’re going to be able to ask the right questions of AI and get the information that you need. So that is kind of the value of human, expertise in some of these conversations that we’re having right now.

Here’s some practical considerations for human oversight when we’re thinking about AI.

So within the US, you know, content created by AI cannot be copyrighted or trademarked.

Copyright law is very different in different parts of the world, so I can’t speak outside of the US context, but that is the current state in the US. And so, for those of you working in other kind of jurisdictions, I encourage you to stay current on what the copyright laws are, within your area regarding AI.

There’s also, applicable laws in the US. Again, it would FERPA, HIPAA would be two that would be really important dealing with education privacy and medical privacy, that impact content that can be uploaded into generative AI for analysis.

So information, you know, that has to do with student records, medical records still needs that human oversight because of, potential violations of law if those were uploaded.

And then, I’ve mentioned trade secrets and competitive intelligence already, but just to kind of reiterate that here, you know, protecting trade secrets may also limit your desire to use generative AI in your context.

And then, of course, just having additional conversations within your organization about what may be and may not be appropriate, for using generative AI.

So kind of in summary, generative AI has a lot of potential and it’s advancing all the time. I’m able to do a lot more with it today than I was able to do, you know, certainly six months ago, certainly a year ago, maybe even three months ago. So just to as I’ve encouraged you to do, during this webinar is to try it and keep coming back to it to try different things over time.

That said, I think it’s important to pause before we get too far, use down the road of using AI to determine the best ways to use generative AI in our particular contexts and setting some of those, you know, kind of norms upfront, so the people so, you know, the new individuals coming in to your organization can be clear on what is acceptable and what’s not acceptable and why, those things are the the way that they are. So I think it would be good to have conversations as early as possible about the use of generative AI in your organizations.

And then, you know, kind of lastly, just kind of reiterating that human expertise.

You know, humans create a generative AI. I think it’s important that we reflect on the value of human expertise and implement strategies to underscore that as we find ways to partner with generative AI in our libraries.

So with that, I’m going to turn it back over to Bradley to share some information.

you so much, Lauren. There are so many creative and powerful ways to use AI, including the generative examples we saw in today’s session, but also as a tool to improve daily workflows for special librarians and their users.

Those of you joining today who are Lucidea clients know that we’re always looking for ways to make your portal more efficient, user friendly, and intelligent.

I’d like to take a moment to share that we recently introduced AI search, a new feature available for all software as a service clients using Lucidea Core, ArcivEra, Argus, GeniePlus, and SydneyDigital. Please note that Lucidea’s AI Search is not generative AI and never changes or creates content. What it does do it made is make it easier for your portal users to find what they need even if they’re not quite sure what to search for. Once activated, AI Search broadens search queries using semantically related terms drawn from your database. This helps all users to discover relevant comprehensive content regardless of their subject knowledge or research expertise.

AI search is user friendly, fully configurable, secure, and private. And if you’d like to set up a live demo, our team would be happy to chat and show you how it works. You can reach us

at six zero four two seven eight six seven one seven. And if you’re a SaaS client, you can activate AI search today at no extra cost. Simply contact client services, and we’ll help you select the fields you’d like to include.

And if you have any more questions on our company or any of our software solutions, please feel free to reach us at any of the contact details listed on the screen.

On behalf of the Lucidea team, I thank you all for attending today, and until next time. Thank you.