“Death, AI, and Librarianship” Keynote. Internet Librarian 2025 Conference
Abstract: In a time of rising social isolation, ideological division, and technological upheaval, libraries stand at a crossroads. Lankes shares thoughts from his new book and delivers a bold, urgent, and deeply human call to action for the library profession. From confronting the rise of book bans and the erosion of intellectual freedom to navigating the ethical minefields of AI, he explores the evolving identity of librarianship in the 21st century and champions the power of storytelling, the necessity of joy in the face of adversity, and the radical inclusion of “feral librarians”—those who enter the profession through unconventional paths but embody its deepest values.
Transcript
Transcript (AI generated):
00:00:01
Good morning. Thank you, Brian. Thank you, Jane. I really appreciate the opportunity to come speak to you today. I am always thrilled to be part of this conference. It’s a great group of people getting to know them and working on a regular basis. So, without further ado, because hearing me mutter on is not necessarily that helpful, I’m going to do the infamous look at my slides for the next 45 minutes or so, and hopefully then we’ll have time for a conversation over on feed loops. So, let’s make sure everything is having a good moment. There we go.
00:00:36
So, can libraries save lives and communities, not just inform them? And as always, this is one of those you put the presentation abstract and title. Together, months in advance, and I just come out with a new book called Triptych, and Jane contacted me and said, do you want to come talk about some of these issues? And I said, yes, absolutely. And so, here’s this lovely. system. However, given the times that we live in, I think it’s really hard to.
00:01:07
begin your conference presentation at least with a nice sunny pretty blue background and so actually these are, this is the real title of my slide, Death AI and Librarianship, and I just want to warn you as we get into this I’m going to start in somewhat dark places and it’s coming. The good news is it’s coming. There’ll be a nice pretty slide at the end that will make everything feel better, but that’s the goal because I think that what we’re seeing not just under our current or rather past ten months, but as we’ve seen over the longer.
00:01:42
term, we have a lot of issues that we’re dealing with. I also understand we probably have ample international audience members there and I will talk a little bit internationally, but a lot of my comments will come out of the context of the US. And what we’ve seen over, the past years are a lot of big things that we’re dealing with from technological upheavals environmental upheavals we’ll get into some of those and a division within the country which we’d like to think of as political and.
00:02:16
ideological but I will actually argue is more socio-economic in a moment and this challenges the role of what libraries do how should we participate and be part of really bringing forward the opportunity country and mitigating harms brought on by different social change and so that’s where I want to begin with and we’re going to start let me just start with a quick story I was giving a presentation to the library in the Lincoln system library in Springfield Illinois the.
00:02:50
capital and I got off the plane and you may or may not know this but Springfield Illinois does not have a particular large airport and so in about 10 seconds I walked off the, the plane, it was at the curb, and the last time I’d been to Springfield many years ago, and I went, and please take a moment to roll your eyes on how all this is going to sound, I was assuming there would be cabs there, as there had been 10 years ago, of course, there were not. So I hopped on my phone, called an Uber, and sure enough, a big Suburban pulls up, I hop.
00:03:20
in the back, and it’s about an 11-minute drive from the airport into where I was going, downtown, Springfield. And, as we’re driving, I’m a little tired, the driver, wonderful woman, probably, definitely in her mid to upper 50s said, so why are you here, and I said, I’m speaking to the library, and she said, so what are you talking about, and I just sort of threw out of a line about, I’m going to talk about social connectedness and the importance of social connectedness, and she then said, oh, that’s amazing, that’s an important topic.
00:03:51
She goes, this is actually the anniversary of my divorce, and when we got divorced, he got all the friends. I had, move out of the house I moved into another state I’ve only come back here within the past 10 years and I started uber driving and she goes this is my world she sort of gestures to the thing and in that 11 minute drive she went through how she stayed in the marriage for her kids and what dating was like and how hard it was to connect and she just sort of had a little bit of therapy.
00:04:24
session if you will as we were going forward and I realized as I went then to the next day speak to the librarians to talk about social connection has talked about the role that libraries can play in bringing together these large large, separations in our society that she was the person that the library needed to serve she needed the library she needed to have a place to go and be someone other than an uber driver have a world that was slightly larger than her her.
00:04:55
suburban and so that’s where we really begin that this discussion, which is I think we need to begin moving from the idea that libraries serve and libraries wait and libraries do good things without major direction to the notion of being much more active and much more conscious and much more directed in the kinds of service that we provide to our communities and that’s not only in public libraries but in our school libraries and that’s very much in our academic libraries and our special libraries and so that’s where I want.
00:05:27
us to talk about but to get there to talk about what I think we as a library community need to do, I need to give you sort of two major setups two major discussions about where we get there and so once again in a cheery happy mood to start off your Tuesday morning let’s talk death I always, when I put the slide together I just had this image of Monty Python and the meaning of life and the chapters like death and so we need to talk about what.
00:05:58
that means and why that in, impacts our society and what we’re doing. So let me do that through a series of graphs. Here’s the first graph. So several, several years ago, Anne Case and Angus Deaton, economists out of Princeton, Angus has won the Nobel Economics Prize, began looking at an interesting phenomenon. They began looking in Montana and then they began looking broader.
00:06:28
And what you can see is actually in that lower graph, what they began to find was people were dying that shouldn’t be dying. If you look at demographics, if you look at rules, something was happening. For example, not only we saw that for the first time, in the early 2000s, a dip in average life expectancy of US population. If you were to find what an industrialized nation is, what a modern nation is, whatever word you wanna use, you could use longevity. Because starting in 1900 to 2000,
00:07:00
and just about, is this pretty steady line of us living longer. We began doing things like fixing sanitation and hygiene, and then we did this thing like figuring out heart disease and how we can keep people alive. All my Staten fans, we can share our history with that. And so there was this sort of progression. And when they looked at this data, what they were trying to figure out is suddenly they were finding middle-aged white guys dying younger. They were pulling down this data.
00:07:30
Now if you’re not familiar with the US, frankly, middle-aged white guys dying early is not something you would expect when you compare it to other demographics. And then when they began looking at what they were dying of, they found a series of things that once again didn’t make much sense at the time. They were dying of drug overdoses, now this was during the opioid crisis, but they were seeing this in a broader sense. They were dying of suicide. And the average suicides happened younger and, more educated crowds, and they were finding it complications of alcoholism.
00:08:04
They started with middle-aged white men, but then they began looking at other demographics and they found it in the African-American community, in women. They found it sort of all over. And when they began to think, what is it that holds this together? What’s the common variable that is marking the people who are dying younger? What they found, and we’re going to come back to this in a moment, what they found was, it was the presence of a four-year college degree. Those without a four-year college degree were dying younger than those with.
00:08:37
And this isn’t completely unusual, by the way. I was in Denmark, same thing, I’m sorry, it was in the Netherlands. Same thing happens there, and seeing this in the UK and such, but it was startling. And they phrased this phenomenon as the deaths of despair, the idea that people were feeling, out of control. That they didn’t have access. to healthcare, they didn’t have access to a lot of other services, and they were turning to things that either helped them forget they didn’t have the control, or that they felt.
00:09:10
they could control, and only finding out way too late that they couldn’t, or that they could make at least the final decision. So this is a phenomenon that’s occurring now. We see this particularly in rural settings. We see this in, has a different political bent and ideological bent, but like I say, the thing that really held it together was a lack of a college degree, and a lack of the college degree, you know, as a university professor, I would love to say that my lectures are what saves lives, but that’s not the truth.
00:09:41
The truth isn’t that a four-year college degree somehow ideologically or would have you immunize people from this effect. It was socioeconomic. People who can afford a four-year college degree in the United States can normally afford healthcare and access to healthcare. They can afford how… Having… One job or two jobs and not five jobs. They can afford and be mobile in where they live and what opportunities they take. There’s this sort of designing line between them. And Deaths of Despair is about a whole group of people who felt like they didn’t have connections and control.
00:10:17
And that was leading to desperation and that was leading to these ultimate untimely deaths. Now, this has been talked about in a lot of different contexts. Often this is being discussed as a notion of the epidemic of loneliness. And I want to take issue with that because I think it’s important that we look at the data. So, looking at the data. More graphs. I want to talk about social disenfranchisement. And that’s a big mouthy word.
00:10:50
And we can talk about lack of social participation if you want. Lack of power in how communities are run and organized. And I want to make in a moment the difference. differentiate this idea from loneliness. So social disenfranchisement is the idea that people, and frankly, people without the four-year college degree, have less ability to shape the communities that they are a part of. They have less political voice, they have less engagement in social wellness organizations, less opportunities to volunteer,
00:11:21
less opportunities and to have a say and organize and debate and connect to people that have power. And Robert Putnam, who many years ago, wrote a famous book about bowling alone and talked about how we see in the United States, this sort of disintegration of social wellness organizations and social philanthropic organizations, so-called funny hat organizations. We have less Shriners, we have less Masons, we have less people who are going bowling on Thursday night in bowling leagues.
00:11:52
We’re seeing a lot more people living, in their own smaller world. He updated his book, he updated this concept, he also worked with another author, Romney Gannett, who began to talk about the notion of social connection. And he had this, what he called the upswing, and that’s what this top graph is. This idea that there’s a sort of I versus we concept that pervades society, that you go from a I society in the late 1890s up to 1900, the so-called Gilded Age, where there.
00:12:28
was a big push on individual achievement, individual responsibility, individual freedoms, and greater social disconnection in things like wage disparity. We saw robber barons and the train folks and the steel magnets having a huge percentage, of the wealth of the nation, an outsized say in government, an outsized say in what would be provided by the government, what wasn’t, a lack of social welfare and a welfare nation. And so we’d be able to do that. How do we begin this and what are the arguments?
00:13:00
Putnam makes is starting in the 1900s with the progressive movement, with FDR and with Haft and the introduction of welfare, child labor laws, the introduction of labor standards, the introduction of a whole suite of things. We began swinging to the idea of a we nation, this idea that we could come together and have community goals, community outputs, that we could talk about a community responsibility and a government responsibility. And he talks about how we really began to sort of bond, and this was during World War.
00:13:31
I and then World War II, and this idea that we came together and we talked about public education and availability and affordability. We began talking about social mobility. We began creating a structured middle class. We radically decreased wealth disparity, and we began to have a common view. And he’s very clear in this book where he talks about this was not universally achieved across demographics, but even in places like, women and women’s rights, even in civil rights and African-American and Native Americans,
00:14:05
there was progress and progress could be seen. Not enough, but progress was being made. And so his argument is when JFK talks about what you can do for your country, when he talks about going to the moon, when you talk about LBJ and the great society, we had these commonalities. And he puts the peak of this commonality somewhere in the mid-1960s. And starting in the mid-1960s, for a whole host of reasons, we began to see the swing back, swing back towards individual.
00:14:36
freedom, greater wealth disparity, and frankly, a new gilded age where we have people with tech billionaires, we have people with outsized voice in how society works and outsized control in that area. And so the first thing we have to look at is, he says, it comes with consequences, right? The Bowling Alone book was a consequence of a lack of, national unity and a feeling of national cohesion. Now, this has been picked up by lots of different people,
00:15:06
whether they talk about deaths of despair, but this is often also referred to as a loneliness epidemic, that more people live alone, that more people, whether they use social media or what have you, they’re sort of on their own and they’re more interested in their own lives and they don’t feel and therefore act on a connection with a larger society, that people are increasingly lonely. But I want to give a special shout out to my doctoral student, Jane Orr, who didn’t necessarily, that didn’t sit well with her, and she began looking at the data.
00:15:37
And we have data from European studies, we have data from U.S. studies, we have lots of data around loneliness. And what you find is that it is a bad framework. Because number one, the percentage of those who identify as lonely within a group and within a culture has not been identified as lonely. It has not significantly changed. The numbers have gotten bigger only because population has gotten bigger. The notion that social media makes us more lonely or more alone also doesn’t hold up.
00:16:08
because what you find is that social media is an amplifying effect, meaning if you feel lonely and you go to social media, you will feel more lonely. If you feel well connected and social, you go to social media to be better connected and social. So it’s an amplifying effect. The other thing is the phrasing of this, that epidemic is some sort of medical treatment and it’s a malady. And rather than thinking of loneliness as an illness, think of loneliness as a natural.
00:16:38
drive. She said the equivalent, and I love her metaphor here, is loneliness is like hunger. You need hunger to know when to eat. Hunger is not a negative thing if you can satisfy it. That idea that when you get hungry, it’s not a disease and it’s not good or bad. It’s something that’s supposed to guide action. And loneliness is the same thing. It’s an emotion. People feel lonely and the value of it, the benefit of feeling lonely, is it pushes you to connect, to find your network. Also, when you look at loneliness, there is absolutely a pathological sense of loneliness that leads to depression, that leads to needing treatment and therapy and really serious issues. But the majority of loneliness is situational.
00:17:23
The people who report the highest level of loneliness are new mothers, people who’ve just had a child and that’s disrupted their access to work, that’s disrupted social networks because now they’re focused on this new baby, people who change jobs, people who move across the country, people who have these sort of periodic events that disrupt their social network that leads to loneliness. And once again, that loneliness is a prompt to recreate or create new social connections. In fact, the people who report the highest level of loneliness are new mothers, people who’ve just had a child and that’s disrupted their access to work, that’s disrupted social networks because now they’re focused on this new baby. And once again, the people who report the highest level of loneliness are new mothers, people who’ve just had a child and that’s disrupted social networks because now they’re focused on this new baby.
00:18:00
network or the place to go and create those networks and so that graph that you see on the bottom there that graph is not an increase in loneliness it is an increase in articles about loneliness we have a loneliness epidemic epidemic we have people who are taking this social disconnection which is much more structural much more economic much more about freedoms and rights and policy than it is about an emotion and we are ascribing that to an emotion people are absolutely having issues on social media people are absolutely being bullied on social media but social media is.
00:18:35
not causing loneliness it’s exaggerating it if it’s there and once again can be a way to reconnect to other people to minimize it and so when you come see things like a Surgeon General’s report from several years ago that says loneliness is equivalent to smoking 15 cigarettes a day it’s interesting data but when you look and push through the data you’re really talking about social disenfranchisement people who live in a community may have a community may have a community that has a community that has a community that has plenty of friends, may have plenty of people they connect to on a regular basis, but oftentimes when they’re getting together, they’re talking about how they have no power, how they have nothing.
00:19:08
that they can contribute and change within society. So the first thing that I want to deal with is, as we talk about what the role is with libraries and librarians moving forward, is that ultimately, this is an important question, and it’s places where we can make an impact. And once again, it’s really easy to see this as, oh, public librarians will do that. They’ll, they’ll have some book groups and life will be solved. But it’s not very, it’s not just public libraries, it’s not just school libraries, it can be academic libraries, because I want to give you an example of how this shows up in society.
00:19:40
So more graphs for you. So when we talk about social cohesion, we need to talk about how people feel and don’t feel a part of it. And so this is another lovely book, I’m a library professor, I like to point to books. After the Ivory Tower Falls by Will Bunch, and he sort of didn’t write this book. It’s a knowledge of the upswing. He absolutely did a brilliant job of talking about how the higher education sector in the United States fits perfectly into that swing and perfectly into that problem. And he talks about the idea that if you look at how higher education has evolved, over.
00:20:14
its history, it has at one time become more socially connected and more of a means to progress through social mobility, and eventually now to less. So, he starts at sort of pre-World War II, and pre-World War II, those that went to the universities were a small number, primarily male, primarily from wealthy groups, and there was a sense of merit. In other words, if you went to a university, you were meritorious. You had great merit.
00:20:45
After World War II, actually at the end of World War II, when suddenly we had this massive workforce of able-bodied men coming back from war and into an economy that wasn’t ready for them, a lot of the policy makers tried to figure out what do we do? with this policy, because after World War I, when soldiers returned without any dealing, with the notion that everyone suddenly needs a job that didn’t have one, it led to the Depression, it led to marches on the Capitol, it led to social disruption.
00:21:15
This was what they wanted to avoid. So after World War II, every returning soldier was given access to something called the GI Bill. Now, every returning soldier, so long as they were male and white, let’s be honest. And so what that meant was the goal was to take this returning workforce, that their jobs may have been replaced by women, automation, they weren’t needed after the large marshalling of forces for the war, we could move them into higher education, and we could buffer that transition. And it worked out spectacularly well, because it turns out that this was not simply a move.
00:21:50
to sort of put them in a holding path. It led to education, innovation, creation, job creation. A middle class. And there was a social. well which is we had just fought fascism around the globe we wanted people to come in and have a liberal arts education that would train them that fascism was not a good thing and that we could prepare them to look for a better creator society and then what happened is starting in the mid and late 60s where the majority of colleges at least public colleges were free because states.
00:22:25
paid for them the states began to change the narrative began in California and it began looking at the notion of people are coming in they are getting an education that education is allowing them to get better jobs so isn’t education sort of part of job development and so let’s begin if it’s going to help you get a better job maybe you should pay for part of it and so you begin California and then across other states the amount of money being spent from the state to support higher education began to decrease this allowed.
00:22:57
universities to start charging tuition and so, It went very quickly from college is free to college is cheap, but you got to have your job and you could work a part-time job and pay for it, but we see over time, and that’s the lower graph here, that what happened is that states continued to pull back money, universities continued to add tuition, and then they began adding tuition beyond the gap, and we began to see tuition rise beyond inflation, and universities became higher.
00:23:28
and higher cost, less and less accessibility, and the mythos of merit remained, so suddenly people were like, well, if you can’t, if the state’s not going to pay for college, get a scholarship, get a part-time job, and eventually no part-time or full-time job could pay for college. It’s like, all right, now you can get, we’ll put in these scholarships for low incomes, but even those were limited, and then it still went up and up and up. We’re talking about, you know, Vanderbilt charging $100,000 a year to go to college, and we.
00:23:59
began to look at… debt. But once again, in this notion of people having less and less financial access to higher education, the merit narrative remained. Well, if you just worked harder, if you just were smarter, if you just could do this. And so what that meant is that people who no longer had access to higher education weren’t there because they didn’t have, they weren’t smart, but because they didn’t have the money. But the narrative told them because they didn’t have merit. And so that’s what the upper graft is. If you look at the share of population by income level, you can see that.
00:24:31
the highest quartile of income are where we get most people coming to higher education. And so as Bunch tells in his book, we began to build a group of those who had a degree who were meritorious, those who did not have a degree that were told that they were not meritorious. And you meet that up once again with the deaths of despair, you have this horrible effect where those of lower income that don’t have access to higher education also don’t have access to health care. And so you have this horrible effect where those of lower income that don’t have health care also don’t have access to all these different services. And so they begin to feel powerless. They begin to feel social disconnection.
00:25:03
And so they turn, once again, to deaths of despair. They turn to drugs. They turn to alcohol. They turn to suicide. And I should make a demographic note. There was some good news that came out of the demographics recently on drug deaths. Drug deaths in this country are down. The problem is when you look at why drug deaths in the United States are down, it’s not because people aren’t using drugs. It’s because we have wide availability to Narcan and drugs that can help in overdose situations.
00:25:34
The actual usage of drugs has not minimized. So that’s the first thing I want to talk about, which is when we talk about libraries serving or saving our communities, we have to understand that we have communities that are split in two around socioeconomic status, which has led to ideological differences, populism, different policies, which explains, a lot of the current social situation that we’re in. And that. That. That division. has real consequences for mortality. That if we can begin to heal and provide.
00:26:04
and change the access to those, rebuild trust within a higher education system as one example, we can actually talk about saving people’s lives. We can actually talk about how libraries can increase the lifespan. So, I told you we’re gonna start dark. Dark number one. Let’s talk about dark number two. Now, I have to say I am, as you can see from this illustration, I’m not against AI. I use AI. I like AI. But we have to understand that how AI is impacting society.
00:26:36
is much different than how we are reacting to how it’s changing society. So, I wanna talk to you for a moment about how AI also creates a really stressful environment, not only within this underserved community, but across the board. And to do that, we need to talk about corrosive AI. We need to talk about a degradation of trust. We need to talk about the notion that, what AI, not just potential, but we are currently seeing, is quickly, its rapid deployment is pulling trust.
00:27:10
from our institutions. Now, AI, it happened long before AI hit the mainstream. We were already seeing on that I-we curve, this idea that people were more and more interested in individual rights and capabilities, and so therefore we’re becoming much more skeptical of institutions, particularly government institutions. So we were seeing that to begin with, but now we throw into this fire the gasoline of AI that begins to pull trust even further. We’re seeing, for example, rapid deployment with limited testing of these systems.
00:27:40
We’re seeing systems being put out there with very few guards. Sora 2, OpenAI’s latest text to video generation system when it was first deployed, allowed you to make lots and lots of videos with Martin Luther King Jr. doing whatever the hell you wanted him to do or to say. We saw, public figures, celebrities, etc. And it wasn’t until there was this sort of outcry that said, wait a minute, this is not cool, that OpenAI began to put in more guardrails. We see this on a regular basis that this race between Google and Meta and OpenAI and Microsoft and whatever to get the latest model out to get the latest attention is going at such a rapid pace that they’re putting out dangerous product. They’re putting out things without controls. And even the controls that they then add in later are being done.
00:28:30
at a high cost to society. We’re seeing gig workers in Venezuela, gig workers in Africa that are being paid pennies to look at the most horrific products of human intervention, human imagination, and human behavior so that they can farm out destructive information. The bigger effect, though, is once these systems go out, and they’re really good at generating human-like, text, of creating images, of creating video, et cetera, that we’re seeing this introduction.
00:29:05
of skepticism into media. Images and video that were sort of seen as trustworthy as a general class is now being degraded. This is one of my favorite articles ever in The Verge by Sarah Jong. No one’s ready for this. And what she talks about is, she talks about images. And she says, you know, if you’ve bought a cell phone in the past two or three years, it comes built into all the camera apps and the photo app and whatever, easy AI.
00:29:35
So that you can take an image and you can say, oh, I don’t want that in the background and make it disappear. Oh, I forgot to add my brother. I can click here and it’ll add it. Oh, I want to add drugs next to this person lounging around because it’s funny. And it’s easy to do with no preparation on a consumer phone and then shared widely. One of the things that terrifies me when OpenAI came out with Sora 2 is that it wasn’t just how to do text. video it was built into a social media app they called it a creativity engine they called it the.
00:30:06
ability to take this information and share it and share it in instagram and share the tick tock and share it around the world easily and they say what you know we put a little watermark into it but does anyone know what that watermark means does anyone see what it’s looking at and then they created the ability that when you open the app that literally says look here look here look here say these three numbers and now i’ve created a digital avatar that i can include in any of my videos or i can share with my friends or the world to end in videos and they can put me in any situation that they want and so what this means is that we have created this plausibility dismissal.
00:30:43
of scandal and we saw this recently with a um with a inspector general nominee who uh we found racist and nazi tweets and he said oh he didn’t actually deny it but said ah ai is everywhere and you can do anything we’ve seen our president, do it. We’ve seen everyone do it on multiple parties that simply say, oh, if I don’t like it, if I don’t want to be seen, if I don’t agree, it’s AI. And that was hilarious two years ago. And now it absolutely could be AI. It absolutely is the idea that suddenly we can break down trust.
00:31:17
in things that we used to. Show me the picture. Show me the video. And suddenly we can’t necessarily trust it. And as I was saying, that degrades trust within society at large. It creates this mechanism of doubt that is in everything we look at. And for libraries, it creates an interesting conflicted mission. It creates this idea that on one hand, we want to provide a tool and use this as a way. We know that, for example, AI and generative AI is extraordinarily useful in the disabilities community and accessibility to create images to text for description for those with.
00:31:52
visual impairments. We see the idea of transcription of vocal words into text for those with hearing impairments. And we see the idea of transcription of vocal words into text for those with hearing, lot of great advantages to it, but at the same time, we’re seeing our authors and we’re seeing our artists feel like they have been taken to town, that their information has been stolen to train up these systems, and that’s still an illegal, interesting area. But will our libraries become refuges of trust where we don’t use AI, or will they become training grounds of AI literacy where we do use AI, and how do we find that middle ground?
00:32:27
Because that’s very different than how we thought of search engines, where we could sort of debate what’s our sense of how that tool helps us do our job, and people use the job, and now it’s turning into, does AI make me look evil, and to whom, and is that a people that I’m willing to sort of not serve any longer? This is often summed up around the notion of AI literacy, and AI literacy to me is the wrong answer to the right question. The question is, how do we build greater trust? How do we build trust?
00:32:57
How do we build greater ability for people to navigate? an information world where AI is present everywhere and possible but the problem is that AI literacies is currently being presented doesn’t actually answer the real problem this first bullet this is I kept it PG 13 this was an answer that we got when we were doing a survey of a local public library on AI readiness we sent out a survey we asked the whole staff we got 260 responses back we asked questions like what do you think about AI what do you know about AI do you know.
00:33:31
how to use it what are your concerns etc and then we give this nice little other thoughts and this was one of the first ones at UT here’s the University of Texas and so when we then followed up with focus groups you walked in the focus groups and the answer was we know that University of Texas is coming into our library they’re gonna force us and teach us to use AI we’re gonna have to implement it even if we think it’s horrible and we spent a lot of time say no no we’re here to understand the problem and what we found was that the problem was that we are treating AI as if it were Google, as if it was just another search engine or another search tool.
00:34:03
And so under the rubric, the aegis of AI literacy, there are these two big assumptions which make me sort of back off on the phrase. The first is that assumption is inevitability, that we have to teach librarians. Librarians have to know how to use AI. They have to know how to use ChatGPT. They have to know the difference between a Markov model and deep learning and generative AI and commercial generative AI because it’s going to happen and the world has changed and so we need to be on board with it. Even though we are very, very early days with things like, are these systems sustainable.
00:34:38
economically or in the middle of an AI bubble? Where is regulation on this? How are we going to deal with the idea of deep fake porn being easily produced now that open AI is going to allow romance and adult interactions? How are we going to deal with the notion of therapy, et cetera? So, this assumption that it’s going to happen, it’s going to happen. to change things, but it’s not necessarily going to look like it does now. And if we assume that it’s going to keep going whether we’re there or not, that sort of takes away our agency to shape that as the library community.
00:35:09
And the other one is that AI is simply another tool. That AI, oh, I use ChatGPT for this, right? When we ask people what’s the one word you put with AI, one of the common things was, ChatGPT, right? It’s a commercially generated large language model that produces human-like text, right? AI, we know, are things that drive our car under intelligent cruise control, that keep track of our fitness on our phones. We know that there are specialized AI being used, for example, in medicine to approve diagnoses of breast cancer and review of PET scans and such.
00:35:43
And so what’s happening is this assumption that AI is these commercial companies primarily, working off the same money or figures around social media that exchanged anger for connection. That monetize attention. attention, that want to keep people engaged in their system so that they can monetize it, regardless of the effect it has on the individuals, is scary. One of the questions we got on a regular basis from the librarians who were doing the focus group is, who benefits from this?
00:36:13
Who’s making money on this? That’s very different than when the web first emerged, and we looked at web as a major disruption in the mid-19, late-1990s, where everyone could build a website and do, should we have one, shouldn’t we have one? Well, you could have a website with very little investment. You can’t have your own generative AI system, because with billions and billions of parameters requiring gallons and gallons and gallons of water to maintain data centers that are, increasingly power-hungry, et cetera, we can’t build that infrastructure, so it’s not the.
00:36:48
democratizing effect of technology like the web, it’s much more of a concentration of resources and power within a few individuals. That becomes a question. That becomes a question about, what do we do about it? this and why AI requires more than just information literacy plus. Oh, everything’s deceptive. My 83-year-old mother can’t tell the difference. Well, we just give her a quick lesson and she’ll be able to recognize AI images. And the answer is no. No, she won’t because no one can. Like, for example, this handsome, handsome gentleman on the side, that’s not me. That was produced by an.
00:37:23
early version of Sora. Can you guys hear me? I see when someone says lost sound. I want to make sure that you’re still okay. If you could just, in a text, give me a thumbs up. Okay, very good. So we had this situation where, thank you very much, I just wanted the thumbs up. We had this situation where, if you look at it, AI is really good at producing images that fool people, with the most trained and cynical people finding it. And what this means is when information literacy.
00:37:53
and AI literacy is currently brought together, it’s got a central concept of being triggered, scattered. Skepticism. Meaning, you’re going through your day, you’re doom scrolling, you see an image, you see a story, you see something, and you go, is that right? Can I believe that? And so your mind goes from passive consumption to suddenly skepticism. And then we give you all these tools to follow that skepticism, to look at its accuracy, its authorship, look at notes, look at these other things, and we build information literacy on this idea of triggered skepticism.
00:38:26
But what happens when you’re always skeptical? What does society look like? What does our service population look like? What do professors look like? What do students look like? What do kindergartners look like when everything they see promotes skepticism, when they can’t know if this is real or not? We’ve seen it already affect the political sphere. It’s not just having pictures of a president dropping droppings on individuals. That’s funny. Ha ha. It’s also the notion that we’re seeing.
00:38:57
In 2015, AI produced thousands of individuals. digitalized ads that they could put on social media. We’re seeing that in how we are putting together our products and how books are being produced. When Amazon had to limit authors to three books a day to be published because AI was generating them at such a rapid pace, we have a situation here. AI slop is everywhere and in our collections. And we’re starting to see work slop. We’re starting to see the idea that one study showed that what’s happening is bottom.
00:39:31
level workers are doing basic analysis, basic review, basic reporting, et cetera. They’re using AI on a larger basis and it’s allowing them to do their job faster and better. They’re then pushing these reports and work products up to a second tier that suddenly discovered that AI does a great job of providing a lot of verbiage, providing a lot of median access, but doesn’t necessarily provide the deep analysis. And that’s what we’re seeing. And that’s what we’re seeing. And that’s, their workload is increasing. So AI is shifting the workload. not decreasing workload. And we also see MIT coming out with studies.
00:40:03
that show that people are experimenting with it and 95% of companies that they looked at didn’t find any value add to it yet. And so there’s this search, desperate search, for a use case scenario, including the idea of how we can turn it into our browsers. All right, I gotta keep running. So I suggest the notion of AI conversant. Our goal should be as a group, as a library, as a library society, we need to be AI conversant.
00:40:33
That is ready to discuss, right? It’s an institutional approach above AI literacy. It’s a necessary skepticism as well as optimism and indecision all put in the same group. The role of AI literacy is to take people who don’t have skills and give them skills and in doing so, once again, with the implication of inevitability and tool, to bring them around to how AI can be useful. And I think that that is a step too soon. We need the skeptic. But we need the skeptic to understand AI. It can’t be a simple matter of I heard AI is horrible or it looks bad or I read this article.
00:41:07
They need to know what they’re talking about. They need to talk about guardrails. They need to talk about watermarking. They need to talk about information being used in the notion of how hallucinations work, but now delusions work. Large, long-term access to these models suddenly can have people thinking that they’re mathematical genius and et cetera because systems are built around synchropony, that notion that AI is supposed to make you think it loves you, so you will continue to use it. That means not correcting you when you come up with the wrong information.
00:41:39
We need people to understand that, including at a technical level. So we see this idea of commercial generative AI versus deep learning and targeted AI. We need a critical analysis of AI. And? And the mission. So I think our goal should not be AI literacy. It should be AI conversant. We should be conversationally able to engage in. deep facilitated dialogue, which brings me to the notion of, well, how do we, you promised us a bright day, Dave. How do librarians deal with all of this?
00:42:11
Well, there’s the obvious. Great libraries build great communities. Librarians can be instruments in mitigating harm and saving lives. We can be the place where that Uber driver goes after hours to sit and talk and listen and can meet with other people. There’s a book group in Gerald, Texas, this tiny little postage stamp of a community about 45 minutes north of here. And it’s a book group of elders, and they come together once a week, and they supposedly talk about a book. But it turns out all these elders are women, and it turns out they’re all widowers.
00:42:44
And so what happens is it’s actually a support group. They come in supposedly around a book, but spend a lot of time just making sure they’re okay and how they connect within it. Even in COVID, this great disconnection and disruption, you see places like Toronto Public Library that when they were physically closed… closed, their reference staff would call up elders and at-risk populations just to do check-ins and make sure they’re all right. So we can be instrumental in mitigating harm and saving lives. We can be a place of connection, a place of empowerment, literacy, learning, access to higher education, access to jobs, access to social services, and places of social participation.
00:43:19
In Denmark, the Aarhus Public Library in Dock 1 has started having conversations with politicians and politicians on major issues of the area. This has also happened in Finland. We have this idea of people coming in and saying, I can be part of this conversation because it’s a safe place for me, because I have information within it, because it is not harmful and ideological triggering as much as social media is. And we see this both in school, public, academic, and special libraries.
00:43:50
I talked about the idea of access to higher education. Access to higher education is still important. So university libraries, how do you know that there are people? There are people in your libraries that are, homeless that are couch surfing or sleeping in their car because they can’t afford tuition and a place to live. How do we engage those folks? I’m in Texas, Austin, it’s great. Our graduate program is one of the few remaining in-person library science programs. People come here, they get a great education, but they have no ability to get into student groups or socialize because they’re working three jobs just to afford their rent.
00:44:23
We’ve looked to our donors and we said, we don’t need scholarships for tuition, we need scholarships for rent. And so we can be a place that begins connecting people together. We’re being asked to do more under greater stress. We have long-standing stress such as technological disruption, health crises, climate change, changing demographics as we serve. I know I’m just about to finish up. Libraries are also being asked to do more, we’re being asked to do this, and I’m asking you to save lives. And I’m asking you to do this right in the middle of book bans and legislations that.
00:44:54
are trying to curtail and make librarians liable for obscenity and potential harm. of being in jail. And the worst thing, and frankly the most dangerous thing that’s happening right now is the idea that one federal court has found that public libraries and school libraries are government speech, that they’re not covered under the First Amendment, that governments can decide what’s in those collections given any criteria they want. That’s something that’s going to be dangerous and probably moving to the Supreme Court. We’re dealing with a fragmenting definition of what a librarian is. We’re dealing with the idea that ALA is being pulled out of requirements that.
00:45:29
define these things. We’re dealing with the idea that pretty soon there’s going to be a Texas librarian and an Illinois librarian and a Georgia librarian and they’re not going to be the same thing. And we’re also dealing with the fact that we have created this myth, just like the merit myth in higher education, about the library science degree. That notion that the library science degree is the only acceptable credential. We need all hands on deck. We need pharaoh and guerrilla librarians who are instrumental in forming it. We need people who can’t afford to go to libraries.
00:45:59
We need people who can’t afford to go to school. So, get that knowledge, get the principles, get the values, and get to work in rural communities and other communities. I’m not advocating against library science or devaluing degree. I like that. That keeps me employed, but we need a greater expansion of professional library ecosystem, that deals with rural librarians, continuous education. So even if you have your library science degree and you’re 30 years away from it, you can bring it up to date and alternative paths to credentialing. Ultimately, what we have the opportunity to do as librarians is to bring joy, to bring.
00:46:32
capability, to save lives, to mitigate harms, to restitch our communities together. And that’s something that we should be called to do. And whether it’s using AI or fighting against AI, having the AI discussion is great, but we also need to talk about, once again, Uber drivers. We need to talk about couch surfing students. We need to talk about our students, and we need the freedom of access for people to find themselves. And our collections and our services themselves. We need to fight against the notion that we are somehow, extensions of a government speech, we are the corrective force for government.
00:47:04
when it overreaches when people are educated and informed. I want to thank you very much for the time and thank Jane and Brian for giving me access to this. If you want to know more, I wrote a book on it, of course. Here’s the book and a QR code towards it. Thanks again. Brian, back to you.
