 Let me introduce myself. My name is Masood Kohkar and I'm the UNST Library and Keeper of the Brotherhood and Collection at the UNST of Leeds. I'm also the current chair of Research Libraries UK and on behalf of Power of UK, let me pass on my really warmest welcome to everyone who's joining us in this conference. I hope you have enjoyed the session so far and I hope that you're looking forward to the brilliant sessions that are coming our way over the course of this conference. I must also say that it's not just an absolute delight but a privilege to be able to chair today's keynote session by two highly esteemed speakers. In just a few minutes, I will invite Kate and Tostan to introduce themselves after which Tostan will give his keynote talk first. This will be followed directly by Kate's keynote talk. Once both keynote talks are concluded, we will open the session for questions and answers and discussion. Both keynote speakers are talking about topics which are highly pertinent to the research libraries agenda at the moment and they are also very close to my own heart. Tostan will talk about the role of libraries in the digital shift and societal impact and how it crosscuts with community, including providing us with an invaluable US perspective, along with some of the key challenges that he has seen and the work that is underway to mitigate these challenges. And Kate will take us through a journey of libraries as instigators of human creativity in the age and world of AI. She will touch on topics and themes of authenticity, provenance, trust chains, media collections, and the role of libraries in civic engagement. I hope you're as excited about this session as I personally am. Please do submit your questions to the Q&A button as the sessions continue because that allows me to be able to theme them appropriately and be able to discuss them at the end in an appropriate fashion. At this time, it's my pleasure to hand over to Kate, who will initially introduce herself, and then Kate will hand over to Tostan for his introduction and keynote talk. Thank you so much, and Kate, over to you. Thank you. Thank you very much, Masuda. Thank you very much everybody behind the scenes for all the support in putting forward my presentation. So just a really quick introduction from me. So yeah, my name is Kate Vobsen-Brown. I am Vice President for Research, Innovation and Impact at University College Dublin, where I'm speaking to you today. I'm a recent appointee here between 2018 and 2024. I was previously Director of the Gene Golding Institute for Data Science and AI at the University of Bristol in the UK. And there I was Professor of Engineering, Mathematics, and Bartlett's Capapology. I was also a Fellow of the Alan Turing Institute for Data Science in the UK, and I retain an honorary fellowship there. At the time at Bristol, I chaired the University's Library Transformation Committee, which had a part physical and part digital component. And I am currently Chair of the Space Academic Network, and I'm visiting Professor at Strathmore University in Nairobi as well. Thank you very much. Thank you, Masuda, and the UK community for the invitation. I'm really, really glad to be, I can say, back. I'm currently, and I'm now starting to share my screen if you bear with me for a moment. So I'm currently speaking to you from Chicago, where I'm the University Librarian and Dean of the University Library. It's in that sense a two-part role where I'm overseeing the University Library system, and I'm also part of the university's academic leadership team. But as I said, this for me is a way back because I have been either part of or worked closely with the RUK community for about 15 years before I came to Chicago. We've worked in various roles at King's College London, more on the digital humanities side, at JISC, the UK's digital service provider for higher and further education, and then more closely aligned directives library topics, first at Imperial College in London, where I worked with Chris Banks, on open access, research data, rights for retention, and then for the last five and a half years or so before coming here as part of the leadership team at the British Library. And I'd like to sort of segue into a mix of an earlier part of my life that is very closely related as I just realised to my current life here. Moving countries means that you come across new terms and new concepts. One new term that I've learnt recently is, and it's a term that is in fact describing me, that I'm a first-gen or first-generation college student. As I keep growing up in Germany, I didn't really think very much about this because it didn't seem unusual even though in my family background, not only am I the first person to go to university, I'm also the first person who's managed to get out of the secondary school system with the kind of degrees and options that gave the opportunity. But at the time, it didn't really seem very unusual to me because I've had the privilege, I have to say, to grow up in Munich, a city that has a well-funded state school system, so great schools of teachers that encouraged me and that then allowed me to move straight on to the University of Munich, an excellent university, study there and get a doctorate in history and philosophy without having to pay a penny or I think a cent I don't need to say for that. And the older I've become, the more I've realised that this is a huge privilege and I've also realised the transformative power that education and the access of knowledge had for me. And so I've become a strong believer in widening access to information resources, widening access to education and having a chance to debate and to explore. And this 40 years later to when the photo that you just see was taken takes me to, I think, a moment that partly explains why I'm talking to you today and that I regard as a sort of pivotal moment in my career. What you see here happened in October last year. It was a press conference, a press conference for, I think, a very sad and alarming topic, which is the massive increase in books being banned in libraries across the United States. And I had the pleasure of hosting this press conference at the University of Chicago library. I was fronted by three other people you can see the two here, the president of the University of Chicago and then to my right, I'm looming over my shoulder, JV, the governor of Chicago. There was a fourth person who's not on this photo who I'll show you towards the end of my presentation. I just want to acknowledge that this moment. And in an environment where freedom of speech, the freedom to research is increasingly under threat. But we were very concerned about what's happening in libraries across the United States. It's not just that books have been banned. It's that libraries have been threatened in the weeks running up to the press conference. There have been several bomb threats against libraries in the Chicago area and over time, not just in those areas. And it's also a time where librarians are being threatened because they don't want to remove books from their shelves that often speak to the experiences in the lives of minorities, not exclusively, but this is a large part of books that have been under threat. So in response to this, what we announced on that day is that the University of Chicago library is not working on building a comprehensive collection of to the best of our knowledge, every book that's banned anywhere in the United States. We're making this collection available to anyone who comes to our libraries, and I should say, particularly our main libraries are open to anyone. You don't need to be affiliated with the University. You can just show up. And you can have access to our special collections, many of our services and not to the collection of books. We're also making these books available through interlibrary law and partnership libraries across the country. We've committed to running an events program that will raise the profile of this topic, hopefully encourage more research and more debates on censorship and bookbounds. And then crucially, we also announced a partnership with the Digital Public Library of America, where initially for one year, we make at this point roughly about two thirds, hopefully in time, all the books that are found available electronically to every resident of Illinois. And we're also working with the DPLA to increasing access so that everywhere in the United States, books are banned from libraries using the DPLA online out in partnership with us. People will have the chance to access and read those books. That's been the departure for University of Chicago library system. The library has never done anything even remotely like it. And I should now maybe briefly say a few words about the library system that I'm now leading. When I came to the United States, I've been for a while introduced as leading the ninth largest library in North America. Like many library starts, it's partly true and partly wrong. I've put it in that explains based on what assessment over what timeframe we meet those criteria. I partly put this in here. I think about making a statement that the way how libraries often describe themselves in a more structured way. I don't find particularly helpful, but more importantly also to say that it's not just maybe some of our stats that aren't always accurate or need various footnotes to understand. I'm not sure in the environment where we currently live in the pure size of the collection should be the only determining factor to judge the quality or significance of the library. But I'm mentioning it here to say that the University of Chicago was built this library system with a particular and I think justified pride in our extensive collections. But we're now in an environment where this library is changing, like many other libraries, responding to new needs. And we're now looking at describing ourselves in a different way and having a different program. And I want to talk you through some of these elements linking them to the current challenges and then bringing them back to what we've done with regards to a topic of bookbounds. So how we're describing ourselves now looks a bit different. What you can see is the draft mission statement of our draft strategy. It's just a few weeks early for me to be able to point you to the official document and some of the language on here might very well change until we finalize it. But we are now redefining ourselves from being the host of a large collection to an organization that's strongly aligned with the values of the university free inquiry and free expression to the fundamental values of the University of Chicago. And we are also redefining and widening the community that we serve, not just saving our campus community, but our wider community in the Chicago land area and beyond. And we're adding deliberately a global element to this because we believe that to be effective at a research library, we need to work in partnership with other libraries and organizations. We need to look at how knowledge is created and shared globally, and we should do our best to widen access to that knowledge and to be as inclusive as we can. So this is a very different way for at least this university library to describe itself. And I want to launch into a few challenges that I think we are facing across the library and I think why that cultural community taking up one and just to avoid any confusion. I'm not saying Elon Musk is the prime challenge that we are facing or I know some people might have a different view, but I'm taking Elon Musk's acquisition of Twitter as an example for something that really alarms me. And I think that needs a global response, not just from libraries, but from society, which is a significant amount of cultural creativity of research outputs and of public debate now lives in the systems of effectively almost a handful of companies. And if one of them goes on there, if one of them changes his business model and decides that that content isn't valuable anymore, or that opening that content isn't valuable anymore, and we stand to lose a significant part of what as a historian thinking about my future colleagues for decades or centuries, I would call our cultural heritage but right now is much more than that it's an essential part of think of who we are as society. So I think, as libraries, we need to not just think about our own collections at this point, we need to consider that we have some responsibility for a global collection, and as a sector for our regulatory often natural reasons, we are currently not well equipped to deal with this. And in order to be able to do so, we need to work much more as applicants, we need to engage with lawmakers of policy makers with our institutions, and have a debate that raises the question of how can we make sure that key parts of our future cultural heritage are not going to be lost and that they are also accessible for researchers for learners and for people. So we've deliberately, I think, in our new strategy, acknowledged that we need to tackle some of these challenges and step out of our comfort zone, but we're also approaching this in a more traditional way. Last year we received a grant from the NEH, the Industrial Endowment for the Humanities. The grant itself is for only a million, I say only a million because the challenge is really big, but we've committed to raise at least an additional four million. And what we're going to do is in the first instance, integrating many of our legacy collections at the library with our central system for managing collections so that we have one space where we can hold and preserve and then make accessible all our digital collections. But we're building this, not just as a single system, but as a network of connected nodes, the project is called the UChicago node. The first one that we're currently actively working on building is with researchers in the Humanities division at the University of Chicago. They use a range of tools for processing, modeling, digital content, and we're going to integrate what we do centrally with what would be their node in a way that allows them to pull all the content out of our collection and richly enhance it and then push it back into that central node at the library. Hopefully in a few months, we'll be able to as part of this partnership, launch a research data publication service for peer reviewed research data in the humanities, let bio humanities division and run on this joint piece of infrastructure that we're developing. But we are also thinking about this much broader. We are already holding material here that speaks to the history and also present the future of Chicago, and we are now working with community organizations and building more partnerships to support them in managing their collections, either we can offer to host them or we can work with them to help them and better enable them to manage their collections. So we're moving away from the idea that there's just one single library collection to our connected model where we work together locally and globally to access to make collection accessible, but also make them preserveable. And crucially, we're also going to open up this platform to artificial intelligence research. And that brings me back to a point that I made five years ago at this conference. And this is the slide that I used in a talk about the future of research libraries where perhaps somewhat foolishly I made some projections or prophecy, if you will, and what research libraries would look like in the 2030s. And one of my main concerns that articulated in 2019 is that we in our collections are the risk of disappearing behind particular i driven commercial systems who would intersect themselves between us and our users by just offering attractive services. And not only would we partially disappear, but we also would not really be able to understand what these systems do with our content, and that there was a real risk to openness and transparency is values that are caught to my recent research process. Now, about five years later, I think we've seen with chat GPT and similar language models, the way of how this could work, and how some perfectly fine library content ends up being spent out to users in ways that may or may not be accurate. I think there's another major change and challenge that we as a community need to tackle globally. But again, would require us to raise profile of these issues to work with policymakers to work more across our institutions and raise the profile. I think it can only be part of the answer, but I'm convinced that one crucial part of the answer for making responsible and meaningful is that we need to really up our game when it comes to skills. We need to have our skills at the library. And we need to expand the skills training that we offer to the communities that we serve. So that's something that we've committed to do on our campus. I'm not picking one example for a project that will also do this in the wider community context. Just last week, we were awarded a grant by the National Science Foundation. The grant itself is sort of meaningful, but also not particularly significant. It's for building a four petabyte research data store in partnership with the research computing center and the physical sciences division here at the University of Chicago. If you hear physical sciences division, you'll realize four petabyte might be reasonably big for my big collections, maybe not that big in the context of some of those disciplines. But this project has been crucial in how we're thinking about the library and its role. It was partly because the project is part of the NSF's campus cyber infrastructure plan. And to be able to apply to this, you need a campus cyber infrastructure plan, which the University of Chicago didn't have quite in the way it was required. They were separate elements. And so for this proposal, we put together a group in morning this year or research computing center and faculty and we brought all of this material together and we wrote a campus cyber infrastructure plan. That now, amongst others recognizes the library as a core part of campus cyber infrastructure and for an organization that largely has been defined by particularly its print collection. This is a really significant strategic change. But the part that I mean, some of us most excited about this project is that we also agreed that we would make some of the data that we are creating in this infrastructure available and feed it into teaching activities. The University of Chicago Data Science Institute runs the data for all workshops where they work with high schools to bring data science skills to schools and to inspire kids locally. And we'll make curated data sets available as part of this and I very strongly feel that whatever possible, future grants that we go for will include a skills element, hopefully for us at the library and for different communities that we serve. And that it will also contain a deliberate element of community engagement and supporting local communities. And I want to briefly step back from the library world and take you into Chicago. What you see here is a map of a city that on paper is really diverse in the most wonderful ways. On the third of our population is white. What a third of our population here is black. And then the other third obviously contains a large Hispanic community, but also people from all sorts of backgrounds. And that sounds like a wonderfully diverse place to live in. But the reality is that Chicago is highly segregated. The north side is predominantly white. The south side is predominantly black. And if you look at this map, you might not be able to read all of this, but the essence here is the more red you will see the poor and area on some of the darkest red areas have a poverty rate of around 50%. And somewhere in the sort of south central area of that map, you will see a lot of red and a little bit of white grain between that white grade pile is where the University of Chicago sits. So we are sitting surrounded by some of the poorest, most underserved parts of Chicago, where you can really see the effect of decades of systematic racism of underserving communities, poverty, lack of jobs, lack of opportunity. It's a real challenge for the city overall as one where the University and this library system are committed to being a meaningful partner to ask him what we can do to help and to play our role like others should also do in the community. And so in response to this, I've set up a community engagement program. And a few months ago we fired a director. Looking into the theme of skills, Danielle, who joined us in September, previously worked on a lot for profit organization to help community college students on career paths. And this will be part of the program that we're now currently building. Hopefully in the final stages of setting up an apprenticeship model where we'll open up a community and particularly promoted to residents of the south side to come and work with us here but also if industry partners and gain some skills that will hopefully help them maybe get a job with us or study or do whatever their career might take them. We're also now working with some of the business incubators at the University of Chicago providing information resources and training for people locally who want to start businesses. And I'm hoping that as part of this we can also ramp up our digital literacy training and perhaps also training areas like AI. We're exploring partnerships of Chicago Public Library and we're also talking to local schools to really expand what we're doing in the world of skills and this is an important departure for us. I also points to a lot of change happening at the library. And I could fill a whole keynote on the question of change management and things that I've learned, things that have worked well, and some where we're still learning that maybe haven't always come off that well, but there's one aspect that's very much in my mind, living in such an university. Working in a profession that for years has talked about diversity, equity and inclusion, but I think in some ways looking at the stats, where I think we failed. You all, if you are from the UK know the stats in the UK, I'll just put up here. Parents snapshot of the library profession. 86 to 88% whites currently only expected very slowly. This is significantly larger than the amount of the white population is in particular in Chicago, but also across the United States. And I found the same here when I started at the University of Chicago Library. My whole leadership team was white and with my arrival it was predominantly male and also I think have a great some of my colleagues are and have been individually. We really didn't score very low maybe criteria for diversity and if we want to be a new organization that works in different ways and engages communities in different ways we need to look differently. So we've taken a very deliberate approach to recruitment, partly based on my frustration over the last few years always hearing that they're not enough diverse candidates. And I think well, maybe we're not looking hard enough, maybe we're not training them hard enough. And since I've joined here every single position that we appointed a senior level has been built by an outstanding candidate with a diverse background. And the best way I can summarize some of those changes is sharing a picture from our pre Christmas celebration that shows how this is not the entire leadership team. It also involves a few other colleagues who report to me about how we look now, as opposed to how we look like at the leadership two years ago. This is just one example I think of many changes that we need to tackle in the profession, we want to really become more diverse and more inclusive and be able to work beyond the core community and beyond some of our core competencies. And that now brings me back to the beginning and also slowly to the end of my talk. You hopefully now understand a bit more the environment that we're in and the challenges that people feel why I felt so strongly about the topic of band books. But I can best, I think, talk about this challenge with the slides that are the way mostly talks for itself. This is the American Library Association tracking the number of unique challenges so every single item on here does not stand for one book band. It stands for a book being challenged for removal from a library somewhere in the United States. That numbers go on massively and as I said these are individual cases if a book is challenged a thousand times, it only shows up once here. And you see a rapid increase over the last three years that was already a dreadful statistic to look at in 2022. We didn't have to form numbers for 2023 when we start the band books initiative, but we could see where it's going. And I think this is really a major threat, I would argue, to freedom of speech to freedom of expression. And I hear some people say, well, this is mostly school libraries. And that's true, but it's also public libraries. And it's coming to university libraries as well, maybe not to us private institutions, but it's starting to happen at the state level. So this is a major challenge and is one where I felt we just couldn't stand by. And I said there was a fourth person with me when we made this announcement. We see her here. My name is Juliana Stratton. She's the retirement governor. So effectively the deputy here in the government. She stayed for a while after the event and talked to some of our students and we looked at some of the books that are banned. And if you look through the band books list, there are some classics on there. And there are books that you might read if you follow the model of the standard white middle aged educated white person. But most of the books that are banned are children books and books for young adults that speak to diverse identities. And I've picked one that really I think touched me in several ways. It's a book about a little girl that's lost one of her library books and with her dog, she goes out trying to find the book so that she can use the library again. I've read the book several times and I can't find any reason why anyone might want to ban it. But if you look at the skin color of the girl, that's the only explanation that I've come up with and I can't tell you how much this made me angry and sad. And so a few days before we had the press conference, thinking about the challenge, I wrote down a few words and I want to use these now as my closing statements to make the case that now is really a time for us at libraries to stand up. Book bans are usually not just aimed at an individual book. They're aimed at what a book stands for and what libraries stand for. Books are more than containers of knowledge or sources of inspirational enjoyment. They are a symbol of knowledge and its impact on society. In a similar way, libraries are more than a container of books. They are a symbol for progress and a promise. A promise of a space where we can get lost in thought, get inspired and engage with the world's knowledge. A promise that a free society accepts and cherishes a multitude of views, even if we personally may not agree with all of them. And a promise that we stand by those who cannot afford access to knowledge and that marginalized communities can still use their voices. In this context, libraries are the promise of freedom, of the free and democratic society built on knowledge, of the freedom to dream of a better world for everyone. We stand with everyone who feels there shouldn't be a dream of reality. And so in closing, I would like to invite all of you, I think not just in the United States, everywhere else, libraries are under threat. And I think we need to think about who we are and who we want to be. We need to expand the idea of the communities that we serve and we need to be prepared to take a stand. And that's what I would like to encourage all of you to do. So, thank you very much. And I'm now handing over to our second keynote speaker. Thank you very much, Torsten. Wow, that's a hard act to follow. What a fantastic talk. Thank you so much. And much of what you said resonates with me personally. I'm not somebody with a librarianship background. Fundamentally, I'm an engineering mathematician, a computational biologist. But I have had for many years very close collaborations with libraries and librarians and funded projects and through my role as an educator. And in particular, libraries and librarians at the University of Cambridge, the University of Bristol and University College Dublin and the British Library, all places where I've worked. So thank you very much of what you've talked about resonates and I think there are some themes and what I'm going to say I think which will kind of interweave with some of the themes from your talk. So, I'm going to talk to you today really about libraries and librarianship in the, in the world of AI and I, I do so from the point of view of a self confess technophile I guess. First of all, a little bit of context when I use the term AI I'm thinking in the very, very broadest terms so one could use a definition like technology that enables computers and machines to simulate human intelligence and problem solving capabilities. And it has a toolbox, the term AI really is describing a toolbox in which there's a whole range of, of methodologies and technologies that we might want to use for different purposes. And this is a really interesting moment I think to be considering the opportunity and challenge landscape for us all. The UK held our landmark AI summit at Bletchley Park last year as many of you will know, which attracted global leaders from the tech and policy sectors. And the UK has since established an AI safety Institute, which sits in a crowded landscape of AI organizations, including the Annaturing Institute, the Ada Lovelace Institute and others. In October last year, the US president released an executive order announcing guidelines for AI. And these include a direction of travel for research for regular the regulatory landscape and for policy development. And then we have the EU Act, EU AI Act, which when fully implemented will address risks described identified described and posed by AI applications, it will prohibit risky practices. It will set requirements for AI systems, and it will establish a governance structure at European and national level for AI. In a world where the topic of AI pervades our public discourse with, I think, almost equal measures of excitement on the one hand and alarm on the other. There's huge momentum building for industry to tap into the commercial opportunities these new tools offer, and for the public sector similarly to seek out the benefits of embracing the change ahead. But along with this momentum, I think there's also anxiety growing around what exactly these opportunities might be, how to make them work for us, rather than against, and how to recruit the right talent or upskill the workforce to do so. But I say that in the knowledge that of course, AI is not new, even if we individually might not consider ourselves to be technology experts, most of us do use it every day, and have done for many years. So just taking the example of the apps you might have on your phone, you probably got to recommend the systems route planners, online shopping apps, that kind of thing, all of which employ some kind of AI engine under the hood. So the fact remains that the large language models the LLNs, like chat GPT, which was released by open AI November 2022, and followed swiftly by a cluster of other similar products has catapulted generative artificial intelligence specifically, and therefore AI solutions have brought this more broadly into our shared cultural consciousness. And I think this has opened up an opportunity for us all to energize the conversation about how to harness these tools for good. In this landscape, it's interesting to me that the views and experience of librarians who are information management professionals have to date rarely been invited into the front and center of this public discourse. In my own experience, I found the council of librarians and cultural collections curators and managers to be enormously insightful. So at University College Dublin where I'm the VP for research innovation and impact, the library team are developing hugely ambitious plans to transform the structure and function of multimedia and digital collections alongside enhanced physical and virtual learning spaces that are fit for the coming generations. Previous employer the Alan Turing Institute in London where I'm an honorary fellow now where I worked for many years is physically located within the British Library. And this has always seemed to me to be a bold welcome statement that demonstrates the embedded nature of data science within this wider landscape. It helps to raise the profile of interdisciplinary research programs that they've supported over the years, which develop and explore, for example, digital historical publication collections. And at the University of Bristol, where I was director of the Jean Golding Institute for data science for many years took my move to UCD last month. I worked with the Chair of the Library Transformation Committee and worked closely with the head librarian and director of the theater collection there. And I've learned so much from their deep and rich experience in the information landscape data management information delivery and user engagement within digital learning environments. The question librarians can and should I think play a unique role in the wider cultural conversation around the use and impact of AI broadly and perhaps Gen I specifically. So I'm really delighted to have this opportunity to highlight just a few areas where I think this conversation could be particularly productive and I've picked four. There are more and I'd be interested in feedback for other other gaps as we go along the way. So the first I think is to see libraries as implementation hubs for new information technologies. One of the aspects of the implementation of AI solutions which I think generates particular alarm is the risk that models may reinforce and amplify any biases inherent in the training data sets which underlie them. You somehow sometimes hear people saying wouldn't it be great if there were some exemplars including lessons learned about how new information technologies might be integrated into existing data infrastructure or how risks might be assessed and mitigated. How the opportunity landscape arising from this kind of integration might be best realized. And I've always felt that well don't we already have this within our library community. Arguably in the 1990s so when I was kind of starting my my own career. From then the rise and rise of the internet pose many of the same challenges and opportunities to libraries as the AI revolution is posing now. With this experience librarians can show us how to approach such new technologies within the frame of a kind of critical thinking. Holding provenance and trust in the quality of information at the heart of the mechanisms of assessment. Libraries are of course spaces for learning and have strategies for engagement with educators which are mature and structured. And at least I've seen a marked variance in enthusiasm for these new AI tools amongst lecturers, students and professional services as you might expect as they weigh up the perceived pros and cons. Some are technophiles and early adopters at one end of the scale and they perhaps see opportunities to transform ways of delivering content, ways of learning and assessment. And at the other end there are many standing back and waiting for permission or incentives to integrate these tools into their programs they perhaps fear pitfalls and the time it might take to upskill. And then somewhere in between there's a sizable group on the fence, feeling interested and optimistic about these new tools but unsure as how to integrate them or whether they are even empowered to do so. So I'm wondering can we learn from the experience of libraries over the past decades of rap or rapid digital transformation. Can we raise the profile of the instruments of engagement with new technologies, which libraries have used to create have created to tackle issues like these. And bearing in mind my experience is that of an external observer and educator and researcher and collaborator, rather than that as a library professional. So my knowledge is far from complete. I continue to be impressed by the range and richness of methods of engagement that libraries are putting in place. I've seen salon series developed which provide for a for discussing AI and the use of these tools and research and education. Some libraries have generated a communities of practice, which break down discipline and sector silos and bring people into conversation spaces where opportunities and challenges of AI could be addressed, and staff can begin their journey of upskilling. Yet others are developing data generation hubs for multimedia collections. These mice include, for example, the digitization of texts, scanning of 3D objects, film and audio recording augmented reality development and reaching out to students and staff offering courses, facilities, data storage and archiving and ongoing support. And such hubs can serve as a focus for our communities to examine AI social, ethical, economic and artistic impacts and also feed into conversations about responsible and explainable AI. Second area I'm interested in is how libraries can be seen as creators and evaluators of next generation search engines. So anyone in the higher education sector at least who has become to experiment with a readily available gen AI tools on the market today and I'm thinking about tools like chat GPT for all the embedded co-pilot Dali Claude three Gemini. Any of those people who are first played with prompt tests and then tentatively begun to integrate the functionality that they discover into everyday work tasks will probably share my experience of initial amazement at the speed and quality of the narrative or figurative output, followed gradually by awareness of the limitations and then a bit of downright disappointment at the occasional false information or inappropriate response, and then a final routine acceptance of the tool and integration into normal work activity patterns with all the caveats that sit around that. Many people I'm networked with now use these tools routinely for example for editing texts for generating synthesis of texts or templates for documents for generating first draft content for human editing, changing the tone or style of narrative, or creating summaries of documents. So in other words, there is a groundswell of use of these tools in my sector at least as next generation search engines. Libraries have a great deal of experience of developing search tools physical and digital, learning how underlying systems frame outcomes and user reception. It may be that libraries are already exploring how such sources search systems incorporating AI either embedded as part of their information infrastructure and access to existing user interfaces using for example a chatbot technology. Recommender systems or by aligning commercial solutions off by tech prime so I'm thinking here of copilot or Gemini might be deployed and evaluated. The future does hold some potentially transformative opportunities. These kinds of tools allow for free text queries queries using digital images and in future audio video clips. AI could be used to run audits on collections, for example, searching for hate speech or diversity surveys. The tools could provide alternative shortcuts to close reading or comparative literature research, translation, sentiment analysis, all in the context of a specific bounded library collection environment. The tools could also create personalized recommendations and study assistance tools that are discreet or they might integrate with users wider digital environments like Google Maps. But then we have to step back and say how does this refer back to concerns around privacy and data security. The learning from initiatives to engage and assess these opportunities within libraries will be of great interest and potential impact to the wider community. In particular, I think it might align with initiatives already developing, scoping out the future of work and the rise of working from home or away from the main place of learning or employment. Third area, libraries as places of trust and authenticity. Another contested issue for AI solution developers and users is that of ethical concerns raised both by the foundational principles on which the products are built and the applications that they are directed to. So for example, Gen AI LLMs, large language models, require training on model datasets, labelled, mostly labelled modern dataset. These are often derived from accessible material online, including copyrighted work without explicit permission being sought. Both text to text and text to image and other forms of multi-modal Gen AI can produce outputs which imitate the style of specific artists and writers. And we've all become painfully aware of deep fakes, which are high profile in the realms of images and videos, but also appear in the written word. The rise of fake, inverted commerce, academic articles from paper factories using AI is one of the most recent guises of this phenomenon. So circling back to the first topic we addressed, another example is found in inherent algorithmic bias, where models may reinforce and amplify any biases inherent to the training datasets that underlie them. This could also privilege one user group over another. So for example, it's been shown that AI detection tools which are developed to hunt out text that has been generated using AI might result can result in false positives where that text has been written by a non-native language writer. Libraries have particular expertise in generating authenticated collections and designing and implementing governance frameworks for large and complex datasets. A strong and I would argue perhaps unique level of trust is placed in libraries by user communities as custodians of collections, as data stewards and learning environments. And I think this could be leveraged into the development of shared AI governance frameworks within institutions, including the design of mediation processes. But this should also become part of our wider discourse around the meaning of authenticity, which takes me to my final topic. Libraries as custodians of human creativity. In my own collaborations with libraries and cultural collections, my eyes have been completely open to the many ways that existing collections can be used to support, inspire and provide a forum for new creative work. It therefore feels natural to me to ask the question as to how libraries might be able to contribute to the current very active discourse around the use of AI in the creative process. Libraries hold collections of work which are mostly static, I'm kind of saying mostly in brackets, because the only exception I can think of lies in cases of retraction or amendment of, for example, journal papers, but please do let me know if I've missed something and there are others. And I think we have a vision of a text publication as a very definitive piece of work, a bounded piece of work. But I wonder in a world of AI whether that assumption might be challenged. Many, perhaps most of our future library holdings will be digital. It is theoretically possible to unlock an ebook and alter it. Censorship could also find a useful tool in AI to change content. So the question remains, how can we monitor this and prevent it? What are the implications of this information management and for our very concept of a book? Can the experience of librarians help us navigate this changing landscape? Second area here is plagiarism. Gen AI makes it easy for writers, whether they be students, educators or researchers to generate narrative which is inherently based on a training dataset created by others. The issues of copyright which arise from this are being tested in courts around the world as I write, and this is likely to be a long drawn out process. When does this constitute plagiarism? In many institutions, library staff act as advisors and training providers on plagiarism and their experience would be invaluable here. Is the use of Gen AI in the creative process of writing an issue of plagiarism or pedagogy? Can there be a cross-cultural view on this, or are these judgments always dictated by local norms? So in conclusion, I've highlighted here four ways in which I think libraries and librarianship has a crucial central role to play in our collective evaluation of AI implementation today and the design of that human technology interface as we look for the future. So the ones I picked out were libraries as implementation hubs for new information technologies, libraries as creators and evaluators of next generation search engines, libraries as places of trust and authenticity and libraries as custodians of human creativity. And I want to sign off by saying, please tell me there are more. I'd really love to continue the discussion with you all. Thanks very much. May I just start by saying what an absolutely incredible key notes on slightly leveraged topics, but also very, very strongly linked with the common themes around trust, authenticity, role of libraries, community engagement, impact. And I think it's been, you've both taken us through an incredible journey, both at an emotional level, at an intellectual level, but also at a future gazing level. I think it would be absolutely incredible to start some conversation around that. But before that, I recognize that the official time of this session will finish at three, but I'm just so intrigued by all of this that we would eat into some of the break time. So I hope the audience members will be okay with that and enjoy the discussion as well. There have been a few questions that have come in. So many, many thanks to everyone who's put those in. And also, oh, sorry, I've been just told that the official closing time is quarter past three. So I think that's even better now. We can absolutely have a proper discussion around this. Let me start by actually just highlighting a comment that I think will echo many, many people's sentiments and that's related to your keynote post. And I think there's been a comment about how much everyone's appreciated the heartfelt, important and informative angle to that keynote. And I think there's a link thing around that about how did that happen? You know, like, this is a really, really difficult scenario. There's huge political pressure nationally and globally statewide. What did it take to actually say we are absolutely committed to the principle of democratic access to knowledge? Was it straightforward? Was it quite difficult? Was there a lot of political angle to it? Just really intrigued about that side of things. Well, thank you for asking that question. I sort of, when I came to the United States, I knew that it would be highly likely there would be political challenges, including the potential of seeing some challenges of a scale that I, if you'd asked me in the late 90s, if I would expected seeing such challenges to democracy in a democratic country, I would have found it hard to believe that that would happen. I knew that there was likely something to happen, but I was really shocked to see the scale of it. And I think what really inspired this event was a meeting of the advocacy and public policy committee of ARL, the North American equivalent to our UK member of. And I raised my hand and said, well, as the new person here, shouldn't we do something bold on this? And there was a long discussion and I'm not speak for the committee because that was an internal meeting, but I just point to the challenge of being in a country where there are many state universities who have to follow the lead of their governor on all sorts of policies, which including in some states currently means that if you use the word diversity or inclusion in writing, you might get the whole university into difficulty with lawmakers and local institution. So I left that meeting realising that some of our existing groupings would not be able to do something, but also thought we had a private institution. We are the private institution though that has often been classed as conservative and small C in that the University of Chicago has a core principle that says it does not get engaged in politics, because it feels like the university needs to be a place where everyone can express any view that's the principle that we work with. And if the university expresses a political view, not only will it be a mess to come to a conclusion that everyone would support, but also it shuts down the discussion that the university wants to see on campus. It makes one exception to this, which is threats to the core values of the university and particular freedom of speech. I had planned to do something in this space, but there was one issue that I couldn't get across which is, I do not have a technical infrastructure where in a copyright compliance way, I can distribute banned books across the country, and librarians will understand the challenges. Building a local collection is not that hard. And then I started talking with the Digital Public Library of America and have an infrastructure to do it. And suddenly, we had the local means, we had a means to make national impact. And I then did what I sometimes do. I just naively go to the president, the provost and the other deans and say, and they asked me, what do you want to focus on for next year? And I say, here's a topic that the whole university should focus on because I was able to link it to the central value one point where the whole university will stand up and not shy away from a political fight that we were able to get that support. And we also had the advantage that on the state level, the state of Illinois just passed a law trying to restrict the ability of libraries to ban books. So there was a framework set in place. But by the time, it still takes a lot of time aligning all these stakeholders by the time I have the mostly aligned within two and a half weeks to pull this off. So the one thing that I would like to say maybe as a closing point, this was the complex challenge. The fastest I've ever seen university procurement to work should we had about two weeks to agree to deal with the public library from Mary Thomas. There was one thing that helped, which is we had a common purpose, and everyone understood why we were doing this, why it's important by transcends this university. And so my team rallied to put this in this was start of the term the busiest time of the year. Everyone said, this is important. It's important to get used to who we are. And anything of that political magnitude. It can only really work with the impact that it deserves if you align this I could have done this just by myself. I have some freedom here as an academic dean at the university, but it's aligning with institutional values that makes the call. It's really, really helpful to us. And then really interesting to see the dynamics of the challenge and the context in which you had to operate. It actually aligns very nicely with Kate's view about the four areas that you were talking about the libraries can take a lead. And I think in some ways, it's a very interesting fifth day, which is liabilities as political forces for good across the institution. I think that could be a really, really powerful addition in that, which actually takes me to some of the things that you were highlighting Kate and what incredible keynote. Let me start by saying that as well and what a burden and timing for this because these are the challenges that most of us are currently looking at. And I'm really intrigued and there have been questions around this as well about the generative AI impact on trust and authenticity. And one of the areas that you are highlighting and also linked with some of the things to us to understand is libraries as places of trust and authenticity. What do you think our role is in this in terms of generative AI because in some ways they are interconnected if you're not testing the new technologies if you're not looking at what it means. Will we still be considered as places of trust and authenticity and we build on our heritage to build on our legacy and actually say we are we have been places of trust and authenticity for a long period of time. But then there are also people who say actually there are some real issues in the biases we have also accumulated. Are we genuinely neutral? Are we genuinely places of trust, worthiness, etc. So it would be interesting to just seek your view and then toss in a few of anything that would be interesting to see that view. Yeah, no, I think that I think that libraries have a huge role to play in this wider conversation. I don't think that the sector alone can has all the answers or all the empowerment to enforce the actions that need would need to happen. But I think you do have a voice as a community and I think it is listened to and should be listened to at the highest level and it's interesting just watching the chat in the Q&A as well and the suggestions that are coming forward from there around priorities of information science and the shared visions for metadata are kind of shared actions around a scrutiny for AI generated content within the publication industry as well. Maybe that's a really interesting conversation that you might be able to broker that's much more difficult for somebody from a data science background to broker. So actually talking to the landscape of editors and publishers and seeing whether there's any kind of shared mission between those two sectors that intersect so closely together around trust and authenticity and what it means, what individual or collective authorship actually means for the terms of long term curated collections. So I do feel that libraries have a very unique voice and the role that the sector has as both custodians of existing authentic human kind of creative work and supporting new forms of that creation gives you that kind of unique voice but I don't think any other sector really has that you can speak to both sides of that coin. The disappointment for me is that people from the sector have not, we're not in the AI summit in the UK for example, we're not amongst the invitees. And when the expertise and that kind of that depth of knowledge, cross generational knowledge of how to manage technological innovation for information management is something that runs like a thread, not really just over the last 50 years but the hundreds of years that libraries have existed, there is that thread there, and yet that expertise is not has not been visible certainly at the UK at the highest level. And so my agreeing to come and speak today was partly I think to say that the data science community recognises the expertise here and would like to work with you more closely I think to raise that profile to give people an opportunity to be heard and take part in that discourse at the very highest level. And that's absolutely fascinating because it links really strongly with the question that Manuel I was asking about the standards and how we can input into development of either metadata standards or protocols that are cognisant of the changing landscape in the AI but also proven on statements the trust change will become really really interesting going forward but knowledge feels authentic versus not authentic. I think all of those angles are their libraries and librarians can bring some real strengths in that discussion complimenting and supplementing the work that the scientists and AI developers and others are doing. So yeah, absolutely. I can't agree more. Most and any views from your side on that. I mean, I would support everything that you said, maybe one boring element to it that I'm currently trying locally, which is procurement. The universities have been using our tools for a while like all of us have and they're buying more and more of those and licensing them. And what I'm currently doing here locally is having more discussion also stakeholders in different areas, co-ordinating between the library, IT services and others to see can we maybe come to a general agreement on what kind of principles you would like to apply for procurement because I mean, ultimately. I know in my base day is sometimes a tendency to complain a lot about capitalism and vendors want to make money. We can either change the system or we can try and work in the system and the most effective way I can think of within the system and if someone wants to change the system I'm not saying you shouldn't do it. I'm just currently speaking about this one perspective. The market determines where I will go in many ways. And if we as consumers articulate what we're willing to buy and what we're not willing to buy and if we coordinate and we have clear principles, then that will push vendors in a certain way. That's not going to be easy. That's not going to fix the whole lot. But if we align across our campus, and maybe if we can find a way without breaking any procurement laws and others to coordinate as a sector, then I think we can make a major difference for our users. And that then requires the second part having the ongoing debate on campus about raising the issue raising awareness students I find are super interested in this when we run our first day I went for students. We had like 40 of them sitting on the floor in our largest room. It was one of the busiest events that I think we've done the last decade or two I've been told. They want to understand what's happening. They want to work with us and their many faculty too. So the library can act as a convenient campus to raise those issues and bring people together. We can't fix it by ourselves, but we can be I think part of the answer. And she linked with that to us and then also Kate both of you highlighted to us when you talked about the commercialization of the system and market monopoly is driving where things might move in the future. And Kate, you were talking about like this helps for experimenting with new information technologies. Do you think this is time when we actually take more ownership of our destination and procurement is one rule to influence. Are there any other roots to influence this. I mean, one part I touched upon in my talk I feel libraries need to become more. It's sometimes a problematic term, depending on the legal context, but I would say more advocacy organizations in the. There's a commercial element as a regulatory element. Again, it may be our host universities who will want to speak to lawmakers, but we can be part of that voice and we can raise those issues. And we can work with there's a range of interesting not for profits. I think review the market. I just wonder I like to mention partly because I'm on the steering committee spark, but there are others who are looking at those. So I think we can directly indirectly support the debates and raise public awareness. That will also require us to up our game and increase our awareness nothing partnership with faculty. I also thought the cheering Institute model co-operative of the British library was a really interesting one. Building more of those partnerships is a really good way and working through a set of institutions on all level of regulation and public debate. But a lot of this, I think, needs to come down to having broader public debate to gather the come and support the consensus and what I sometimes find. When I speak to people and explain these problems from a lively perspective, suddenly they go like, oh, I sort of realized there was something I haven't quite grasped what the problem looks like. And then they're interested in talking to us about what the solution would be like. So we don't need to necessarily have the answers, but we need to be able to ask the questions in a way that's compelling that others will want to talk to us. I can jump in there because there's a related question in the in the Q&A, which about kind of empowerment, how do you become involved? So for Kersti, you linked that University of York has posed the question. The University of York specifically, so that university is a member of the adventuring Institute network. And so where universities in the UK are members of that network, there is an opportunity to just become engaged in the activities that are going on at the Turing Institute. For example, through the special interest groups. So I would recommend that you find out, I don't know who it is off the top of my head. I'm sorry. Find out who the academic lead is, which who manages that membership relationship with the Turing. And and ask them for an introduction to the team at the Turing that deal with those with those specialist interest groups and just just become just become a little bit involved. The downside of this, of course, and it's time consuming for people and you're, you're making your sector visible by being visible yourself to the data scientists. I understand that. But there is, for those of you who are who are listening in from from the UK, there is that re-routed to the national debate through the through the entire institute. And nicely links with Kerstin's comment about our evolving role in advocacy and visibility and I think that's where we need to talk with other organizations like the energy institute. But also our counterparts in your size and other parts, which is looking at the infrastructure is around this. I think there's a link question before I move more into the academic side of the discussion, which is about, are we now reaching a time, particularly within the UK? But I'm just also thinking about your previous one posted about global collaboration, global collections and global partnerships, whether we should basically build our own digital infrastructure. Should we think about curated training data sets? Should we think about our own large language models that work on those data sets, which are more ethically governed, which have more careful consideration about biases? What about your feeling slash view on that? Or is it is it more of a by-train at this time? I'll try and go first. I'm very tall about this. I'm a strong believer in shared infrastructure. I also know from experience building shared infrastructure is really hard and takes a lot of effort and we are competing with companies with huge resources that can decide to throw everything if need be a warm problem in the way that the higher education sector often struggles. I'm currently co-chairing a committee with the Vice Provost for Research to develop a future strategy for research computing and some faculty members outlined where if you wanted to be a world leader in AI, we need to go. We did some calculations back of an envelope. The cost for electricity for building the infrastructure that they talked about, based on current data, obviously that can change, came to $1.5 million a day. Now, these ambitious faculty who had world leadership aspirations. So I think there is a scale where I just don't think we can compete. I think what we can do on a more moderate scale is getting our house in order, making sure we look after our digital content well and we can expose it in a standards compliant, easy to side way and develop some mechanisms of trust that as a minimum, you know where library content is coming from, where you can find it and that it has some level of authenticity to it, although that's also probably the context. For anything bigger, my sense is we might need partnerships with industry and we have to choose very carefully how we pick those partners because I'm not entirely convinced governance are really set up. And even, I mean, take two of the large sort of national bodies that I'm familiar with having worked at JISC and the British Library, they're big by many standards. They're all really small based on the scale. So we can coordinate, we can do some of this, but for the wider part, I believe it's regulation and good partnerships of industry that might be more effective. But maybe I'm too defeatist on this one. $1.5 million per day might make you feel that way. So this is, it's technically possible already to do this with a bounded data set. So you're not, you're not making it, you're not giving your data away in order to train a model. It's technically possible. Absolutely right. It's still quite expensive. And I think it's really interesting to awesome that you mentioned the energy use. So if I could predict the next big discussion that we're going to have data science and AI community, it's going to be around sustainability. And I think we're, as a community, we're trying to get ahead of that curve and to drive the conversation so that we can propose ways of managing that. But yeah, any kind of exascale compute system is consuming all data center will be extremely greedy in terms of the kinds of electricity. So there are two things that one can do about that. You can, you can obviously limit searching for training. You can improve algorithmic efficiency, which is genuinely an engine, an energy saver. And, and you can, you can kind of transition to green energy, but then there needs to be some sort of Venn diagram of overlap where we find the sweet spot between those between those things in amongst those things, which I think would really support the kind of focused solution building that that we're talking about here. So it is technically possible. It is still very expensive in terms of shared infrastructure. I agree that is that is difficult to do. But in terms of business information within organizations, it is medium sized companies are doing that now. I was involved in a very interesting project looking at using AI for automated contract development for in within the landscape of the legal profession. So that so there are there are some areas where it makes it makes financial sense to explore these and think about how that how that might be done. But yeah, it's a it's a it's a big investment. And I think there's a really interesting dynamic merging between the need for a shared infrastructure, but also the need for a distributed infrastructure, particularly for cyber resilience. And I think there's two different things, but equally important in different contexts. So absolutely very grateful for those comments as well. Let's just move this conversation a little bit back to the academic domain and Kate you mentioned fake papers, you mentioned academic integrity, all things that we are grappling with across institutions at the moment. Any thoughts on why that's happened, particularly considering that there are strong peer review processes copy editing processes. So how has this slipped through and also on the academic integrity side. This might feel like we need to completely rethink some of the way in which we assess people. Is it is it now going to be a savings assessments is that's even still a valid concept anymore. So any thoughts on that from your side. As to why is this happening. I think, I think you've sort of answered your own question there almost in terms of academic, the academic career pathway. One of one of the main key performance indicators that have been demanded of academics is numbers of papers and citation and disease and and and I think rather than quality, quantity over quality if you like. Now there are one knock on effect of the ref actually in the UK has been to shift that that conversation to quality over quality and impact over quantity, which is quite interesting. But I think if you were to if you were to draw a graph of retractions from academic journals you'd see a very sharp uptick. Now, I think sometimes this is this is this is both alarming and reassuring at the same time is alarming in that there are that many retractions, and they are particularly focused on some disciplines. It's reassuring because my take on it is that those disciplines where we're seeing a lot of retractions so there are parts of the medical publication landscape at the moment where we're seeing a lot of retractions. I think that's because they started looking for them. And I'm doing the right thing in terms of research integrity. And those areas where discipline areas where we're seeing very very few retractions I would say those thoughts are probably to have a look at again. In terms of the these engines that are producing truly fake papers that that's that's still them that still seems to me to be the minority far far far the minority of papers but it is happening. And our detection tools probably need to get better to catch up with that but that's always going to be a chicken and chicken and egg. Whatever sort of version of turn it in or there are other platforms available that are able to to to be used to scan to scan text to look for to look for plagiarism or AI that that's always going to be a kind of red queen running running running to stand still. But nevertheless we have we have to do it. So I think you're right I think it's a combination of of looking at the underlying drivers that have put us in the position where it's published or died within within a career pathway. And also thinking thinking about the how we can keep up how technology can keep up with the human ingenuity to get around technology all the time. Absolutely. I'm just going to ask one last question and we'll try to wrap this up and it would be great to have some some concluding remarks from both of you if you want to say anything particularly through some of these questions or comments in the chat that we may have missed out on. But there's a question about actually started to me but it would be really great to use on this, which is whether there's any danger in the UK about censorship and sandbox emerging within the UK higher education model. And in particularly for our UK libraries, or is it more of a US challenge. And I can start by giving my own thought, which is a lot of this is influenced by the political and economic model through which institutions and therefore their respective libraries are funded. And one thing that we do have an advantage of the UK, but whether that's an advantage or it's a very fragile model, whichever lens you look at it with is that we are actually very in a very limited base funded by state funding. We are primarily either a government level funding for research or international and local student funding that that brings in the primary sources of income, which that which does gives us autonomy to be able to do things in a different way that state funding places might not be able to do so particularly when they are politically divided or motivated. However, it does bring the fragility of the international markets and therefore if and this is happening in many many higher education institutions in the UK and beyond UK that fragility is actually causing the institutions to reconsider different models. And if it turns out that they are linked with state funding or government funding more strongly than I think the governmental policies can have more of an impact in influencing higher education views slash libraries within higher education views. I would also say that Torsten's concluding remarks were absolutely critical for this, which is we absolutely must make a stand on this now, if not very soon in the future, which is what we are about as libraries. We're about freedom of speech, freedom of expression, we are about democracy, we're about access to knowledge for everyone. And those are absolutely critical things if you want to stay in that mode of being able to provide information of all kinds to all people who need them. So that's just my personal view around the fragility of the funding model. It could lead towards that but it's highly unlikely in the current circumstances, but we need to be actively looking at that and pushing the agenda of independence of libraries in there. Any views Torsten or Kate from your side or not? One thing to add, having now lived and worked in three countries, I think we see some I would argue similar drivers and attitude changes but they manifest themselves in different ways. There are questions say in the UK about the role of state funding. If you look at decreasing funding for public libraries, I think we sometimes in relatively well funded research organizations forget how crucial public libraries are infrastructure in particular serving so those types of students that we'd like to see more of that without having access to public library resources might never make it that far. So there is I think an important part thinking about the whole knowledge ecosystem, if you will, that's partly why we use that language here and I can also reflect from the US. Having spoken to representatives of public libraries they feel left alone by the big research libraries in particular by the private institutions who could speak up and either don't speak up or at least are not perceived to do this. I mean Germany at the moment has its own political challenges that raised some questions about the future of democracy in ways that I've not expected that to be. We've also seen challenges against the idea of the expert. So in the Brexit referendum and before, and I think a significant criticism of the role of universities overall and parts of society that sort of seem to move away from the idea that universities are forced for good. I mean, that's reason to be very alarmed in any of those countries. I don't think that's going to go away quickly. And it requires us somewhat to rethink what we do and to rethink some of these challenges in global scale. So the challenges are different. Not manifested, but there's some underlying drivers that I think would make me concerned in any country that I could kind of live in or have lived in before. There's some excellent comments coming in the chat about the role of public libraries, but also pyros that are directly funded by government and the implications of the politics around that. Any final thoughts from your side on that particular note. I think that we in the universities need to do better in communicating the societal benefit. That derives from the institutions we serve and the communities we serve. And maybe in the past, we have looked too much inwardly into the kind of social, social health and wellbeing within our universities and not enough, how porous we are with the communities we serve in a, in a civic sense. I know that I know that that's something that both the university I work for now in Ireland and work for previously in the UK had been taken very seriously over the past 10 years, kind of raising the profile of the work that we do within communities, making that making those borders a porous and actually engaging with the community's external to the university in such a way that it drives activity within the university so it's not all one way as well. It feels to me to be very tangible thing that we could all do, which would at least at least kind of help engage with that with that wider conversation that is, is so contested, I agree. I must say, this has been one of the most wonderful sessions I've been part of. I've learned an amazing amount from both of you and I know that we will be excited shortly so I just want to say on behalf of the whole RLUK team, the executive, the boards and the audience. Thank you so much for your absolutely amazing keynotes for your thoughts for your comments for your insights. I think this will give so much food for thought to so many people. I think we will be discussing this for multiple months. If not years going forward, and some of the insights will actually help us reshape some of our thinking around this.