- [ Resources ](/resources) Share- [ ](https://www.facebook.com/sharer.php?u=https://about.ebsco.com/markdownify/node/163182) - [ ]() - [ ](< http://www.linkedin.com/shareArticle?mini=true&url=https://about.ebsco.com/markdownify/node/163182&title=Libraries at the Center of Generative AI and the Future&summary=&source=>) - [ ]() ## Libraries at the Center of Generative AI and the Future [Webinar](/taxonomy/term/5178) | Original broadcast date: 10 June 2025 Please join Atla and EBSCO for an insightful presentation by Michael Hanegan. In this webinar, Hanegan explores the transformative impact of generative AI on learning and work. He also discusses: - Ways generative AI is transforming learning and work that we have never seen before. - Why the experience, wisdom, expertise and ethics of librarians are essential to building a new future of learning with generative AI. - How librarians and libraries can position themselves at the forefront of this new era of innovation in the early days of AI. [Explore AI Resources for Libraries](https://about.ebsco.com/artificial-intelligence/resources) #### Transcript | [ Download ](javascript:void(0)) ### Libraries at the Center of Generative AI and the Future Ref Link: https://about.ebsco.com/markdownify/node/163182**Elyse Antrim** Hi, everyone. Thank you for joining, today's webinar, libraries at the center of generative AI in the future. I'm Elyse Antrim, marketing manager at EBSCO, and I'll be your moderator for today's session. Before we begin, let's cover a few housekeeping items. You can submit questions anytime during the presentation using the q and a box at the bottom of your screen. Our presenters may share helpful links via the chat box throughout the session. The presentation will run approximately 30 to 35 minutes with plenty of time reserved for audience questions at the end. This session is being recorded, and you'll receive the recording via email in the coming days. Joining us today is our presenter, Michael Hanegan, along with Gillian Harrison Cain, Atla’s director of membership and engagement. Gillian oversees membership initiatives and promotes Atla’s databases, products, and services to librarians, students, and scholars. I'll now hand it over to Gillian for some brief opening remarks. **Gillian Harrison Cain** Thank you, Elyse. Atla is so pleased to partner with EBSCO to have Michael Hannigan share this session. I first met Michael through a longtime Atla member and mutual colleague, Chris Rosser, nearly a year and a half ago. His warm, engaging style of sharing complex and new information was refreshing and very much welcome as we chatted about this new thing that was on everyone's minds. Michael is the founder of Intersections, a learning and human formation company. He is an adjunct professor of AI and the future of learning and work at Rose State College where he teaches undergraduates in the AI and Machine Learning program and at the University of Central Oklahoma where he teaches MBA students in the College of Business. His work focuses on the renegotiation of the relationship of learning and work, how generative AI complicates, accelerates, and enhances that future, and the ways in which we can collectively build a world that is good for the whole human family. Since we first met, Michael and I have had the opportunity to work together on AI focused seminars, consultations, workshops, webinars, and a soon to be released micro credential for librarians sponsored by Atla. Our collaboration included a presentation at the ALA annual conference in San Diego last year, which led to he and Chris's authoring a book that has just been published by the American Library Association entitled Generative AI and Libraries, claiming our place in the center of a shared future. I love that Michael loves libraries and is passionate about the important and critical role librarians can play as our institutions investigate, consider, implement, and leverage AI. I can imagine no better conversation partnering guide as we navigate the path forward. And now I turn things over to Michael. **Michael Hanegan** Thanks, Gillian, for that kind introduction. In the chat, you'll see everyone a link today to a page with some resources, from today's presentation. You'll have access to the deck, some additional materials, that you may want to look at in, subsequent to our time together. And then once the recording is made available, I'll be sure to make that available here as well. I want to say thank you for having me today. I'm excited to talk to you as a non-librarian about libraries, and the and the generative AI in the future. So, again, that link is in the chat. Feel free. That that's, going to stay there going forward. I want to open up with this quote from David Graber, the late anthropologist, in part because it is both for me, kind of my ethical vision of AI in the future of learning and work and I think a reminder of the role that libraries can play in that future. That the ultimate hidden truth of the world is that it is something we make and could just as easily make differently. And I think this raises for us one of the important fundamental questions of the unknown and uncertain, and for me, the optimistic future of learning and work in that we are not constrained, to simply receive the world that is coming, but that we have both an opportunity and a responsibility to actively shape it. My work lives in these four worlds. I work across K12, higher education, library world as an outsider, but a well-cared for and welcomed outsider and in the world of work. And what we know from, the last century or so of the world of learning and work is that learning and work often do not fit together. And in fact, really, since the beginning of the industrial revolution, learning and work have been split intentionally. You go to school and then you go to work. And you may return to school, but it is only for the purpose of then subsequently, returning to work. And this has had two major consequences, that you'll see kind of as an underlying reality in what we're going to explore today. The first consequence is for the world of learning, which is that as learning and work have been divorced now for over a century, the world of learning has become increasingly disconnected from the world of work. We see this in conversations all the time about the relevance of higher education or whether or not some of the skills that we're building in school are meaningful for the world of work. For the world of work, the consequence of this disconnection between learning and work is that the world of work has essentially atrophied to almost nothing its infrastructure and capability for learning. It has been almost entirely outsourced, to the world of education. And the reason that these two challenges confront us at the same time is because generative AI forces us to renegotiate both of them simultaneously, not in a generation, but in a handful of years, at two to five at most. And so what we're on the brink of right now is both the transformation of the world of learning and, the largest upskilling and reskilling in human history. And so my work really tries to explore, how we navigate all of that. As Gillian graciously mentioned, Chris and I just published our book that came out last week, through ALA Editions and core generative AI in libraries, and I hope that, you'll see some of that kind of underlying, idea in there this week. We hope that or we hope that you'll be able to check this out in the future. So let's talk a little bit about generative AI. First, we need to understand that generative artificial intelligence is, an arrival technology. An arrival technology is a specific rare kind of technology. Most technological innovation, is what we call adoption technologies, which is we largely decide whether or not we participate, and our life is shaped by that decision. Maybe you remember, the first time you got a cell phone or the first time you moved from, or a traditional cell phone to a smartphone. Right? Those were all, choices. An arrival technology is more rare, much rarer, and it is a technology that changes the world in which you live whether or not you adopt it. So think about electrification or the steam engine, the railroad, the telephone or the telegraph, the Internet. Right? We think about the ways in which these have reshaped the world in which we live irregardless of whether or not I personally participate. Generative AI is an arrival technology. The difference between it and all previous arrival technologies is that its impact on society will not be felt in decades, but in years. We also have to recognize that it is accelerating exponentially. Let me just show you this in a couple of quick, technical graphs. This is from last summer. This is now a year old. This, and, currently, the current capabilities would not be on this, slide. They would be off the off the top here. But we can see that just in a little over a year, we had three orders of magnitude of improvements, across these AI models. We're if you think about cell phones, for example, we're using five g for our cell phones. That's fifth generation wireless technology. That took us fifty years, to kind of get these, like, five orders of magnitude change. We're talking about these changes, every few months, as we move. And this summer looks to be, even faster than this timeline. Or this is one of the most difficult, benchmarks for trying to figure out AI capabilities. And you can see between 2019 and, mid 2024, we made virtually no progress. And then all of a sudden, you see, the straight up. That bend in progress is a ninety-day window. So each of these benchmark scores are complicated. They measure particular things. But the key here is that the trajectory for all of them is not incremental progress, but an exponential curve. In 2020, scholars, through a detailed questionnaire, were, asked whether or not they thought we would reach what's called general artificial intelligence, and they often thought that it was between, five and eight decades away. Just three years later, we're forecasting four, three to five years later. So you see this, like, exponential drop in how quickly we're moving towards these things. Or maybe you're familiar with the Turing test, which is, an idea developed by Alan Turing about whether or not a machine, could be perceived as human. Recently, a model from OpenAI from ChatGPT successfully passed the Turing test, being judged human more often than actual human outputs. And we have plenty of transformative applications that we already see. So for example, at Princeton and the Indian Institute of Technology, this last fall, they asked one of these tools to build a chip, that was counterintuitive in its design. They gave it no limits, no freedom or no boundaries as to how it would build this chip and suggested that it break away from traditional ways that humans have made chips. This is the chip that it produced. And now, people at Princeton and IIT are trying to figure out why this chip is so effective. The tool was able to produce this design, and then it was manufactured, but we don't know why this works so well. Or this is AlphaFold two. This is a protein modeling, AI from Google DeepMind. The creators of this tool won the 2024 Nobel Prize in chemistry for protein modeling. Protein modeling is essential for drug discovery, and we've been working on this question computationally since the nineteen seventies. From 1970 to 2020, we had accomplished one percent of known proteins in science. In 2023, AlphaFold completed them all. That is one billion, billion with a b, one billion years of PhD level scientific work that is now free, open access, open source, and available to the entire scientific community. AlphaFold has about one point seven million scientists who use this every day, to advance medicine. Or one of my favorite stories recently, from the New York Times, this patient had a very rare blood disorder that could only be solved by, a bone marrow transplant, but he was too sick. And the doctors simply said that there's nothing we can do. We're so sorry, and offered to send him home on hospice. His girlfriend reached out to a doctor that she had met at a rare disease summit who had been using AI for, off label, prescription drug plans for people with rare conditions. He said, I'll call you in the morning. He called back the next morning, with an unprecedented combination of medications, that the patient took over the next three to four weeks, became healthy enough, and received that treatment, and saved his life. He did that by having a tool that could look at all off label use of all medicine and, and for interactions, which is just something that the human brain is not capable of. Or this piece from MIT in the material sciences and engineering department, where they created a knowledge graph with a thousand, material science papers and then asked it to look at art for inspiration for new materials. It looked at this painting from Kandinsky and said it was inspired to create this new mycelium material, which is now being built and experimented with, at MIT. But one of the things we have to remember, especially, as we think about technology, is that speed is not always movement. Faster is not always better. Ruha Benjamin's latest book, I think, reminds us beautifully of this. And so, while we have undoubtedly experienced significant progress, we have to ask deep questions about what this means for our shared life. And this is why we must be reminded again of what David Graeber reminds us, the world is what we make and that we can ultimately choose to make it differently if we choose. So here are three long standing problems of our shared life that are, both complicated and in some ways, made more urgent, with generative AI that I think libraries, and librarians have a huge opportunity to contribute towards. The first is this idea of precarious and precarity. So this comes from feminist ethics of care. Precarious is just this reality that life is fragile. Any of us could be in a car accident anytime soon or have a medical emergency. Like, this is this is just a fundamental part of being human. But precarity is, when someone's life is put in a position where life becomes precarious because of social and collective choices to either cause or sustain circumstances. So for example, we see this especially in people who, must remain below the poverty line in order to maintain disability benefits. Or we see this in ways in which, Americans throw away more than enough food every year to where there need not be hunger in the United States of America. And yet we still have these problems. Right? So precarity is this way of understanding that there are parts of our life that are not as they should be and not as they could be, yet for some reason or another, they remain. And this is one place where, the future of learning and work, I think, can really begin, to either make this much, much worse or much, much better. We also have this degradation of our ability just to connect and engage with each other, and to think about the myth of the public square. This is a wonderful, wonderful book to help explore this idea that, while it has been increasingly difficult in the last, you know, decade or so, to talk across, political and ideological and religious and social divides, that algorithms and artificial intelligence have the propensity to make that much, much worse. But there are ways forward, and this is something that we will have to navigate. And the third is what Alison Pugh in her wonderful book, The Last Human Job, calls connective labor. This is the labor of, seeing and being seen by another human being. As we begin to experience the increasing and accelerating, automation of work, both knowledge work and actual labor, what will happen to connective labor is when I think of the essential questions, of the future. And so there are also three new problems introduced by generative AI. One is that humans do not have a history of metabolizing this pace of change. Humans are beautiful at processing incremental change. My grandfather was born in 1919. No plumbing, no water, no running water, no electricity in rural Oklahoma. And in his last days, one of his favorite things to do because he could no longer drive, was to zoom through the Grand Canyon on Google Earth on his iPad. A phenomenal amount of change in one lifetime. But there was never a point in his life where in a very short period of time, he looked back at the world and thought it largely unrecognizable. All of that progress over his ninety years was incremental. Even though it got faster over time, it didn't feel, kind of unmooring in that way. That is not what we are about to experience. Humans are terrible at extrapolating exponential change. There's a wonderful piece, that illustrates some of this, called Machines of Love and Grace. It is written by the CEO of Anthropic, Dario Amadai. And he talks about what he calls the compressed century, which is that we believe with these tools that in the next ten years, we will be able to do a hundred years of science and medicine. It is difficult for us to imagine what a hundred years of medicine looks like. If we think back a hundred years of medicine, we say, you know, no vaccines, no antibiotics, all kinds of things that are fundamentally different that are just commonplace and ordinary, for us today. It's hard for us to extrapolate forward a hundred years. It's really hard for us to extrapolate forward a hundred years and then presume on the calendar that it's actually going to be, not our children and grandchildren who will experience that, but us, in a decade. The second problem that's new or at least is, renewed is that these technologies have already been used to further degrade connection, and we have to navigate what this means. Right? The way in which algorithms shape, our social order, our communication, the way in which these tools are consistently shown to be more persuasive than traditional forms of communication. We're going to have to navigate what this means. There are, again, the structure of ideas is a wonderful book. The book that I'm reading right now that's keeping me up at night is this one constant disconnection, and it talks about how, life chronically online is in some ways, an escape from being chronically online. It's an interesting, you know, paradox in which we find ourselves, but these technologies, are something that we will have to navigate in that reality. And then the third is just that the implications of this open up very real and very unanswered questions. We just don't know a whole lot of things about the future that we're walking into. For example, this is one paper, that came out this last year from the National Bureau of Economic Research, and they were asking when we reach the point where, what they call artificial general intelligence, which is when an AI tool can largely do, whatever a human can do, at least in some areas, we don't know their grand conclusion whether or not wages will explode or disappear. And the grand conclusion of the paper is, we don't know which direction it will go. We don't know when it will go. We know that it will go. And that's it. Fifty-five pages of saying, it's complicated. It's going to be wild. It's going to be soon, and that's and that's all we know. And I think this is true in a number of ways. We don't know necessarily what this means for mental health and social cohesion. Recently, a report was released that shows that currently the number one use case for generative AI is not writing code or generating images. It is, in a kind of, emotional support to work through ideas, to talk through how I'm feeling, for companionship, for, kind of, quasi therapeutic applications, that people are using these tools, to just navigate their own kind of internal, emotional, and social world. We also have to renegotiate the relationship of labor and identity for our entire lifetimes and really for much longer than that. Our jobs have been a part of who we are. And as that gets renegotiated very quickly and very, significantly in the very short term. We're going to have to figure out how we talk about and how we think about who we are outside of, what we do for a living. And then the third, I think, is this. Will gains in productivity and efficiency and automation, free us up to do things that are, maybe more interesting or more valuable, or will they be used to simply, exponentially increase the expectations in the world in which we work? We've already seen historically that gains in productivity and efficiency and automation have not always been good for us. They've just we've just been demanded more of, that wages are not always commensurate with gains and productivity efficiency and automation, or that those gains are realized by a few at the expense of the many. And so, again, we have open questions about what will happen here, and we will have to decide, what kind of world we will, allow and construct because we could just as easily make it differently. So here's what I think we have to do in the next, five years. One, we have to recognize that disruption is an opportunity to build what we want, and then we don't always have to rebuild what we've had. We don't always have to have, you know, version one point five- or two-point o of what we had. We can just build something different, and I think this is a wonderful opportunity to do just that. The second is that we have to cultivate fundamentally human skills. We have to think about creativity and critical thinking and imagination, information, literacy, the ability to communicate with one another. Not that these skills will not also be, replicated in artificial intelligence because in many ways, they will, and in some cases, they already are. But for our shared life, in the midst of this technology, we will still need to cultivate and sustain these fundamentally human skills. When you look at, for example, the World Economic Forum's, core skills, you'll notice just briefly that almost none of these are content skills. They are fundamentally kind of, human skills or what we might call soft skills. We also have to provide a vision for the world that we want and for the one that will outlast us. Here, I think librarians historically have made great contributions, and now it is time, for us to more aggressively build, what we have been calling for a long time. We also have to renegotiate a new relationship of intelligence and expertise. Historically, intelligence and expertise have been rare, expensive, and slow. It has taken time, money, and opportunity to create intelligence and expertise in the future. These will be cheap, fast, and readily available, and so we're going to have to navigate this change as well. Here's where librarians really come into play. We are about to experience an explosion of information. I love this from, the Heinrich Foundation. They talk about how in the next three years, we will produce more data, more information than all of human history combined. So we live in a world in which information is not only ubiquitous but overwhelming, and how we will navigate that that potential, what that means for, our shared life and for the world we want to build, will require, the expertise, I am convinced, of librarians. So what does this mean for librarians, libraries, and library world? It means that the things that you've already that you already hold are essential, to the world that we are about to inhabit and the world that we want to build, your experience, your expertise, the library ecosystem, and the ethics that drive the work of libraries and librarians. And so what I want to conclude with before we have time to just have conversation is to explore briefly in three categories, what we, the collective human family, need, what I think as an outsider, what I think librarians need, and then what this does, for each of us. As far as it relates to the experience of librarians, what we all need and that librarians really bring to us is this, commitment to equity, diversity, and inclusion of both people and information. What that's going to require of library world is to evolve with the changing knowledge production and consumption ecosystem. So for example, in the resource page that I've given you, I've listed a lot of books. The reality is that books are not primarily where the information is at, right now. The system moves too fast. It moves, too regularly, for those spaces. And so if you want to really be, caught up on the bleeding edge of what AI means for the future of learning and work and for libraries, it's going to be a lot of papers. It's going to be, spending some time in the hellscape that is Twitter. It is going to be engaging in preprints and other kind of nontraditional forms of kind of, knowledge production and consumption. If we get this right, if librarians bear the influence from their experience that they can, this, I think, leads in many ways to a much more democratized, access and opportunity around knowledge and information at scale. As it relates to the expertise of librarians, Chris and I believe that metaliteracy is the essential human skill for the next five years. This, you know, this is a great gift that library world brings to the larger world as the rest of the world comes and says, how do we learn how to learn? I just keep, standing up and screaming and pointing back to all of you saying, library world knows. Please listen. What that's going to require from library world is that we you're going to have to make the connection to trends in future of work. So we're going to have to move kind of outside of the library bubble and catch on to what's happening in future of work and then make that more accessible, in ways for with people who do not have, the kind of, grammar and training and expertise the librarians have. And then when that happens, librarians, I think, become increasingly, essential infrastructure for both learning and work at scale. I do think that this can actually be the golden age of libraries. As it comes to the ecosystem, we all need, a reintegration of learning and work. What libraries and librarians and library world are going to need is to continue to build up the ecosystem and then to engage with other ecosystems as ambassadors and stewards. When I work in higher education and in industry, I find all the time that problems that are, well within the wheelhouse of librarians, those places struggle. And so I'm hoping that the library ecosystem can begin to really assert itself in spaces outside of library world at the bleeding edge of this kind of age of intelligence. And then in doing so, we can we can collectively model how these changes can be navigated in ways that align with the kind of human centered values and ethics of, librarians. As far as ethics, we all need, to build a world that is unapologetically and explicitly human centered. For librarians, we need you, to meet people with language and frames and practices where they are, to recognize that, while we should continue to be unapologetic about our explicitly human centered approach, that we truly mean that this is for everyone, that we have to do so in a way that doesn't confront people into an alternative or argue them into an alternative, but that nurtures them into an alternative. As people navigate the disruption and change and disorientation of the next three to five years, nurturing spaces who provide a better, more robust way of living in the world, will be the ones that change the world. And then the impact is that we can cultivate and sustain a space that is both unapologetic and truly inclusive for people who are learning, relearning, and unlearning the world from which they come into the world in which we want to share. Three final images for, all of us today. One of the essential kind of underlying visions of our book is this Japanese proverb. And I won't I won't say it in Japanese. Chris is fluent in Japanese, so he can he can share that with you someday. But it translates to the strong one under the floor. This is a person or persons who, while often unseen and unthought of, hold up the house. And we feel that this is a wonderful metaphor for who librarians are, that librarians are the ones who are underappreciated, under recognized, and yet without them, the house is unstable. And so we hope that the strong one under the floor, all of you, can remain in that role, but also come out from under the floor, and engage in the space where everyone sees, what you have held up. The second that pervades our book is to think about libraries as centers of gravity, both a convening approach and that we hold things together, but also in an animating approach that we, we move those things around us in ways that bring, order and stability and, from chaos. The third and final is that sometimes libraries and librarians, when we talk about AI, are asked, to leave where they're from, to leave behind some perceived past, and move into the new future. And what we believe is that librarians should stand squarely where they are, rooted in their ethics and their experience, in their expertise. And instead of looking out on horizon for where we should go next, we believe that librarians should look up. So grounded in where we've been, holding true to the ethics and expertise and experience and wisdom of the ecosystem, but that above us is kind of an infinite space of possibility from which to build. And we think that what this means then is that librarians' real contribution to the future of learning and work, is to be themselves, in a world that doesn't necessarily understand the many, many gifts, that libraries and librarians bring. Because that is the ultimate hidden truth of the world, that it is something we make and could just as easily make differently. Alright. I hope we can use the rest the balance of our time, to just explore some questions, and we'll see where this takes us. **Elyse Antrim** Thank you, Michael. There have been several thoughtful questions, in the q and a already, so that's been really nice to see. One of the questions that came in towards the end, and I think it was right as you were on right before you were on your final slide, but I figured we may as well reiterate this question, since it's so relevant to your final slide. It is what specific fields do librarians need to insert ourselves into, and how do we do that? Do we need to expand our job or move out of libraries to make sense of these changes? **Michael Hanegan** This, I think, is going to be a both and question. So there are things in which there are areas in which libraries will be invited, to expand beyond the walls of the library. Let me just give you a practical example. Last summer, I spent a lot of time working, with an engineering team, who was really struggling with engaging with the manufacturing team for which so engineering team would design, manufacturing team would build the things that the engineers have put together. The primary reason that they could not communicate is that the knowledge management on both sides was terrible. And instead of saying, take a communications class or let's get some meetings together, I literally connected them with a librarian and just said, let's build some systems to keep all of this together. And magically, those questions go away. And so it's not that, like, do you need to hire a librarian and bring them into this team and, like but it's just to say that there is this like, that information science background becomes so important in a world where technology depends upon data. When we think about what industries in which libraries need to speak into, the answer is yes, in a couple of ways. One, I think that librarians are going to be essential in helping industry learn how to learn. So universities and other places don't also do not have the infrastructure for the scale of upskilling and reskilling that's necessary. Upskilling and reskilling is going to involve just about every person over the age of sixteen going forward. So there's not enough classrooms. There's not enough professors. There's not enough space in just traditional education spaces for that to happen. Libraries as trusted partners, as trusted conveners, as trusted sources of wisdom and information can play a huge role in building that kind of metaliteracy and the capacity to learn how to learn, both for individuals and at scale. The other thing that I would encourage you to do is for libraries to think meaningfully about skill acquisition and pedagogy. So, librarians often have the expertise that they need, to speak meaningfully into these spaces. Sometimes we need to build our capacity for how we speak into those spaces. I experienced this a lot as an outsider to library world. One of the things that I appreciate so much about going to conferences like ALA with people like Gillian and with Chris is because I have my library translators with me that help me make sense of all the things, that that the internal dialogue is useful, amongst fellow amongst colleagues, but that wisdom, experience, and expertise is going to have to be translated to a space that has no background, in the kinds of things that library world talks about. And then the last piece I'll say is that I would recommend two books here. One is the skill code, and these are all linked in the page, for today. The skill code, which is about skill development, and the other is The Last Human Job, by Allison Pugh, which really talks about the importance of that connective labor. I think librarians flourish in both of these spaces as we move. What else? **Elyse Antrim** Thank you. Gillian, was there a question that you were particularly interested in hearing answered? **Gillian Harrison Cain** Yeah. Michael, one of our attendees, wrote something that I am near and dear and familiar with, that one of the hardest parts of information literacy is convincing others that they need information literacy. She says, I work at an academic library, and it's so hard. Do you have any ideas for how to let others know that they don't understand it at all? **Michael Hanegan** Yeah. I do this in a couple ways, especially in higher ed, but I think this translates to other spaces, which is when someone who has low information literacy sees someone replicate their work who has high information literacy, they go, oh my gosh. I want that. Right? So to see what's possible, it's harder to get someone to be reflective of where they're at. It is easier to show someone where they could be. There's a wonderful paper. I don't remember the title right now. I'll find it afterwards. I want to make myself a note, and I'll put it in the, in the page. But it's about, the relationship of ability, belief, and the configuration of, the calibration of those two. So it's like, how good am I at something? How good am I actually at something? And it relates to the way in which generative AI and learning occurs that often people overestimate how good they are at one thing, and this is why it's hard to convince them that they must do better. So oftentimes, I'll go in and, like, when I'm talking to faculty in higher ed, I will often ask them to submit in advance, a project they're working on or that they're stuck on or a place where there's heartburn, and then just try to show them, hey. Like, I don't know anything about this, but, like, look at the progress I made. Imagine what you can do as someone who actually knows what you're doing here. So I think really just showing them not, not taking a deficit approach, like, hey. You don't understand, but just to say, like, hey. The ceiling is much higher than you think it is. I would love to help you kind of make that progression. There's a lot of ego in places like that too. So sometimes we have to navigate. You can move forward without saying how low, we might perceive someone to be. **Elyse Antrim** Okay. Another question. Do you have any recommendations for getting libraries to commit to GenAI as an institution as well as teams of people with drastically different opinions on the future of this technology. Those differences appear to be holding some libraries back from exploring Gen AI meaningfully and holding up the house, to use your metaphor. **Michael Hanegan** Yeah. This is exactly why Chris and I wrote our book. We recognize that it is difficult, to try and navigate these questions. We realize that there is a lot of fear and anxiety, about a whole host of things, and we recognize that, we in some ways, in the early days, we engage this technology as if it was any other technology. So what we find in a lot of spaces where we go with libraries is someone will say, I tried it last year, and it wasn't very good, and so I'm not going to use it. And I and I have to say, well, but it's, like, five hundred percent better than it was last year. Like, your experience of a year ago in traditional technology and software is usually pretty you know, your iPhone of a year ago is not much different than your iPhone now. But with generative AI, we have because we're on this exponential curve. Like, the things that you did one time that didn't work very well, are not, are not necessarily a snapshot of where we live in the current moment. And so part of this is we have to build, like, frameworks and shared language, and I think our book tries really hard, to help you do that. But the other piece is just to become more familiar with where it's actually at, right now. I think that's, one of the other hurdles there. And then the key here is that skepticism and hesitancy should not be shamed or sidelined. There is wisdom in caution, and there is wisdom in innovation and flexibility. What needs to happen is we have to build cultures that enable us, to hold those perspectives together and then try and figure out how to move, in ways that align with our ethics and our commitment to build the world that we want. It's not easy. It's messy, but all good work is messy. **Gillian Harrison Cain** Michael, there's been several, questions and comments related to the tension between ethics and the impact on the environment. Can you speak… **Michael Hanegan** Yeah. I think there are a couple of questions here. One, it is really hard to get a handle on what the environmental impact is for a couple of reasons. One, just from an infrastructure perspective, the footprint that we see for AI is not just generative AI. Right? So oftentimes, like, for example, if we're talking about, you know, a data center, not only is your chat g p t running through there, but so is your Instagram and so is your Netflix. Right? And so part of it is, like, parsing out how this works. The other piece is that I do think that for people who are using these tools and technologies at a scale, there is an there is an increasing responsibility, to try and figure out how to mitigate that. I will say that the average person who is not using it either for something highly technical or something just extremely high volume, you will probably have a larger carbon footprint from flying to ALA than you will for three or four or five or six months of your AI use. So part of it is to kind of put those in context. The other piece that we have to recognize is that the tension here is that while the consumption, the energy consumption of these tools is on the rise, the use of these same tools is also part of the solution to resolving that problem. So, you know, I often say that, like, the air conditioning that we use will not solve the energy crisis, but the generative u the generative AI tools, will. In fact, for example, we've made more progress in fusion energy research in the last three years than we have in the last thirty because of this technology. So the irony here is that while that consumption accelerates, it is also, the source by which we can solve this problem. The other piece is that it is difficult to take a snapshot and figure out where we are at because it changes all the time. A current a current query in ChatGPT uses a hundred and fifty times less power than a query from ChatGPT a year ago. We have gains in efficiency in hardware. We have gains in efficiency in infrastructure. We have gains in efficiency in the models themselves. So, while I do think it is important, while I do think it is a meaningful concern, and I do think there are some approaches to navigating this for people who make up the bulk of the consumption of these technologies. I think I think it is important that we hold that in context with other forms of energy consumption and environmental impact as well. And I wish it was easy to put our finger on all this, but it's really complicated. **Elyse Antrim** Okay. Another question. Changing gears. Thoughts on copyright. If AI is trained on copyrighted information where access was not granted for such usage, how does that impact what the future might hold? Different economic models or possibly the loss of rights for rights holders. **Michael Hanegan** Yeah. These are all these are again part of those kinds of open questions. Right? We have now, three guide three pieces of guidance from the US copyright office about generative AI. And kind of the current answer is, like, it depends, and it's complicated. Like, this is, like, the official guidance. Right? So part of the challenge here is that we don't know. The other the other question the one that actually makes me interested that I don't see a lot of us talking about is this. There's a lot of conversation about, how we don't want certain information to be used for training. And I understand that for, like, confidential reasons. Right? Like, I don't I don't want my medical data necessarily to be used to train these things. But one unintended consequence that I don't think we take seriously enough is that in some ways, by withholding certain kinds of material, we actually bend models in ways that we don't want them to be bent. So for example, if we were to withhold, a whole lot of material that is kind of deep ethical quality, of high integrity, of high academic quality, and then these tools are forced to train exclusively on stuff that is publicly and freely available on the Internet. Is it if it's essentially, what ends up happening is that information, disproportionately shapes the way in which these tools are built. My concern is that, as we navigate these questions, that some of the material that we would want to shape these tools is withheld, and the information that we would not want to be a part of these tools is disproportionately represented among them. I do think at some point, eventually, there will be some kind of retroactive compensation, as it pertains to copyright, but I don't see that getting resolved anytime soon. So for me personally, I want the kind of work that I'm doing to be represented in these tools. I want that to be shown there. My concern is that if it's not, it is, people who have a very fundamentally different view of the world are happy for their material to be a part of that. And so, it's a mess, for sure. I would encourage you to look at the guidance from the US copyright office, though, for those. **Elyse Antrim** Thank you. I think we have time for maybe one more question before we wrap up. Gillian, do you have is there a question that's sparking your interest? **Gillian Harrison Cain** Yeah. Michael, there's so many good ones, but let's end on this one. What advice can you share with the older librarians to be confident in using AI? Are there tools available? **Michael Hanegan** Yeah. Here's what I would do. The best use of AI is not to pick a tool and learn how to use it. The best way to engage is to think about what you do really well and then find tools and uses that enhance that work. So don't feel like you've got to bend in a totally different direction, but begin to try and think about, what it is that you make, what your contribution is, currently, and then how can we find things that help with that. The other thing I will say is this. My hope for the earliest days of generative AI and libraries is that some of the mundane, routine, monotonous work, can be quickly changed in order that that connective labor, that, that innovative labor, that that cutting edge experimental kind of labor, can be a larger part of our work. That the things that we've always wanted to do or, where we have felt kind of divided in our time and attention because this kind of more mundane, automatable work still demands our attention, my hope is that we can begin to take some of that, set it aside, and then focus more on this human component. No one is a master of this technology. No one. Not me, not anyone. And so free yourself from the commitment to being on the bleeding edge, to being first, and being the best. The goal, I think, is just, to improve. **Elyse Antrim** Okay. Thank you. Any final remarks, Michael, before I close this out? **Michael Hanegan** Just thank you so much, everyone. I hope that you'll, that you'll reach out. I am going to take a copy of all of these questions that we didn't get to today, and I will try, to answer them in some form, in that page. **Elyse Antrim** Wonderful. Thank you. So thank you all for joining us today. You will receive an email, within the next few days with a link to the recording of today's session. This webinar was sponsored by Atla. So to learn more about Atla databases, you can visit, ebsco.com or atla.com. I'll put links to those pages in the chat. Links to relevant resources will also be included in the follow-up email. And be sure to follow EBSCO on social media for updates about future events and resources. So thank you again for spending your time with us today. Thank you, Michael, for your time. **Michael Hanegan** Yeah. Come see me at ALA, please. **Elyse Antrim** Wonderful. Thank you all. **Michael Hanegan** Thank you so much. *Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors.* #### You might also be interested in ![](https://assets.ebsco.com/asset/c35d1362-3c7f-4c5e-9521-1db840c407df/meet-mosaic-video-image-780.png) ## Meet Mosaic, by GOBI Library Solutions [ Watch video ](/resources/meet-mosaic-gobi-library-solutions) ![Mosaic logo with illustrated balloons celebrating the product's second anniversary](https://assets.ebsco.com/asset/e0133b2c-0947-4159-a761-26fd94b248d9/Mosaic-Turns-Two-Infographic-web-780.png) ## Mosaic Turns Two: Two Years of Progress [ View infographic ](/resources/mosaic-turns-two) ![](https://assets.ebsco.com/asset/8b5ad26d-b064-41b7-8d76-7307e00aece1/ebsco-timeline-video-image-780.png) ## Celebrating EBSCO Milestones in Partnership with Libraries [ Watch video ](/resources/celebrating-ebsco-milestones-partnership-libraries)