Episode 581: The AI4Libraries Conference

So universities have been aware of AI and how it could affect things for a while. But last year, the release of ChatGPT caught universities and, I think, all of us a little bit off guard. Dylan Ruediger: Obviously, libraries are very important in shaping how universities are responding to generative AI. And they're also likely to, themselves, be shaped by the technological transformations that we are witnessing.

Libraries have to face this thing. It is something that we ignore at our peril.

You are listening to WREK Atlanta. And this is Lost in the Stacks, the research library rock and roll radio show. I'm Charlie Bennett in the studio with Fred Rascoe, Marlee Givens, and a guest to be named later. Each week on Lost in the Stacks, we pick a theme and then use it to create a mix of music and library talk. Whichever you are here for, we hope you dig it.

Our show today is called The AI for Libraries Conference.

What!

AI.

Fred?

Yeah.

Again.

Uh-uh. CHARLIE BENNET: This is too much AI. This show-- Too much AI.

--you have done I don't even know how many AI shows. I think you've reached your quota.

OK. I promise this is the last AI show for this year.

There's only three more sho-- OK.

[CHUCKLES] But Charlie, it's a good one because our topic is a conference about the intersection of AI and libraries. CHARLIE BENNET: You're going need to sell it a little bit more, Marlee. Oh, OK, it was hosted online by Georgia Tech just last month.

OK.

And it all happened because of the efforts of our colleague and guest, Anu Moorthy.

OK. I will roll with AI for this show, but only because it's Anu.

Mm-hmm, that's the spirit. Today, we'll learn what the AI4Libraries conference was all about, why it was important to hold, and how it was put together on a really short timeline.

And our songs today are about rapid change, dealing with entities that have uncertain ethical standards--

Like you.

--and conferences naturally. This conference was a big thing-- I heard that. I'm just ignoring it.

Yeah.

This conference was a big thing to try to pull off, so let's start with It Takes a Lot to Try by Aviram right here on Lost in the Stacks.

That was It Takes a Lot to Try by Aviram. And this is Lost in the Stacks. Our guest is a Anu Moorthy, the electronic resources librarian at the Georgia Tech Library and founder of the AI4Libraries conference. Anu, welcome back to the show.

Thank you.

And welcome to the studio.

And I'm so excited to be here.

That's right. Last time, you were a virtual guest, like in a Zoom room or something.

Yes.

But you're here now, live, not artificial.

Yeah. CHARLIE BENNET: [CHUCKLES] So you're here to talk about the AI4Libraries conference. What's the elevator pitch for-- when you tell people what's it all about and you have just a few sentences, what do you say? We have been hearing about AI almost everywhere we see. And libraries are the place where the students, the faculty, and others come to to explore the things and learn from us.

Yeah. Was there a moment that you sort of reached peak frustration and said, I have to figure this stuff out? Or did it kind of slowly grow? Like, you just kind of kept thinking we need to do more. We need to do more.

It kind of-- it was not a-- I had been thinking about it for almost four years, going on the fifth year. But I was hoping that someone will start a conference so I can just go learn from it.

I didn't know and did that. And suddenly, July, mid-July I thought, why don't I start this? And I can call all the people. And maybe some people will come. Maybe 100 people will show up, and then we can learn from each other and implement a few things in our institution.

It happened in October. You said you went from idea. It just kind of occurred to you in July, well, you were the one that you were waiting for. You were waiting for someone to do it, but you were that person.

Yes.

Did you ever run a conference before?

No.

That's a short timeline.

Yeah. How did you begin this process?

First, I ran it with my boss--

To check if it was OK.

--whether it's OK. And she said, go for it. And I didn't know how to build a website, so I talked to someone, and they suggested a few things. And then one weekend, I told my friends, colleagues, I'm going to play with this website thing, and I'm going to build a website. And they said, I'm going to go play tennis. Go ahead and build your website.

And so I built the website. I bought the domain names. I learned a lot, again, that weekend. And I don't-- it was the last weekend of July. So the website was on. It really didn't look nice, and I was really frustrated because I didn't know how to do graphics. And I still don't have a logo. I don't know how to create a logo. I know exactly what I want, but I don't know how to do it. I haven't had the time to create the logo for the conference website yet.

So that seems like that's the right order of things. I got the idea for the conference. I've got the conference website. And then you have to actually have the content for the conference, right? So how did that come together?

I just started cold-calling people through LinkedIn, Twitter, or X. CHARLIE BENNET: Are you surprised that this actually worked out? Because the way you're saying it to us, it's almost like, I can't believe it happened. Is that the general feeling? Yes. I'm still in awe of the interest. And I don't know how it all happened.

I mean, you did this because you were just hoping that someone would do it. And then you just realized that no one was doing it, so you did it. Did you find that other people were kind of waiting around for someone to do it? And then when you called them, they were like, oh, thank Goodness you're doing this thing.

No, they were all gracious. They all responded, number one, which I wasn't expecting. And they were really gracious about doing the opening keynote. "Just put me in." "I blocked the whole day for you." "I'll do anything. Just put me in any slot." And that was the response I got from all these people that came.

And you got librarians that were interested in AI, faculty and various departments, scientists that were interested in the ethics of it.

Yes. Librarians, they also submitted proposals from different libraries. And we had technical services staff who work in the back end. And then industry leaders in systems, they did the closing keynote. And also the ethics person, he talked about AI and ethics. And the opening keynote was copyright. So we could learn about copyright and the impact that AI is going to have on copyright issues.

So I have in my notes here some of the themes-- leveraging AI for efficient information organization and discovery, transforming technical services through AI, preserving and digitizing cultural heritage with AI, ethical use of AI in libraries. Do these just represent-- not just. Do these represent the interests of people who agreed to be part of the conference, or did you have stuff laid out first?

No. Those were my ideas because my tagline is "cautiously embracing it" because there's always going to be things that are gray area. So I wanted to say "cautiously embracing the future of libraries." And those were the things that I thought that we will talk about. I thought about ethics because that's like the most important thing for me. And I did not think about-- I thought about archives and digitizing things that was in my head. But systems is automatically in the--

In your interest area?

--interest area. Yeah.

So we're hitting the end of the segment. So we just have to finish with, I think, the size here. How many attendees did you have? This was a virtual conference, right?

Yes.

So all in line, how many attendees did you have?

We had 1,142 attendees and 499-- I may be wrong in the 499, but they were unique attendees.

Oh, OK, gotcha.

And they were from 45 different countries had originally registered. That's 1,650.

So this is your superpower-- putting conferences together. This is Lost in the Stacks. We'll be back with more from Anu about the AI4Libraries Conference after a music set.

And you can file this set under QA76.9.N38K83.

You just heard Static Traffic by Career Woman; and before that, New Reality by Patio. Those were songs about trying to keep up and cope with drastic changes.

This

Is Lost in the Stacks, and we are back with our colleague Anu Moorthy, the electronic resources librarian at the Georgia Tech Library. Anu organized the very first AI4Libraries conference held virtually this past October. And I'm really curious because it's not like you're the AI librarian at Georgia Tech. And I don't know if AI impacts your work in electronic resources. So I mean, what was the interest for you?

I just wanted to learn more about it. My work doesn't revolve around it, but clearly it's going to impact all of us in the library world. Either we're going to help students learn, or our workflow is going to be changed because the publishers are adding AI to their platforms. So I was just curious, and it was a cool learning opportunity for me.

Some people just go on Wikipedia if they're interested in a subject, like, as a starter. They don't organize a conference.

What did you-- what did you learn? What were some of the things that-- sort of the big ideas that came out of organizing the conference and attending it that you pulled?

I really learned a lot from about copyright. If you put things in ChatGPT, how are you getting the answer? And is it OK to use the content that an author copyrighted? Can we train AI with the data or the content that a copyrighted material was trained in?

Would say the copyright session at your conference was your favorite or the one you were most interested in?

Copyright and the systems one. Copyright session was done by Dave Hansen. He is one of the-- he used to be the copyright librarian, and he's doing a lot of work on author science. He's educating authors about the upcoming changes with AI and how we should promote creativity and innovation. At the same time, we have to take care of the copyright.

Yeah. And he was at Duke before authors? And the systems session, what was that one about?

System session was done by Marshall Breeding. He has a website called Library Systems. He does all the things for systems and libraries, the systems that we use. I will not name all of them. We have a lot of systems that we use in the back end for the users to get to that one PDF. All that magic that happens in the background, which will be automated I think in the future because of-- which will be more automated, I should say, because of AI. So that was also my very favorite session.

That's one thing about libraries. We have lots of platforms that don't necessarily talk to each other, lots of acronyms. So, that's libraries. Why is AI going to be important in your specific role, electronic resources librarian?

I think all the redundant work that I manually do could be done by AI. And that will free up the time to do more creative things in my area.

Like conference planning, for example.

So the AI will take away the mundanity--

Yes.

--of electronic-- or not that your job is mundane.

It could sometimes get mundane. So if you're doing usage statistics of the resources, it's really like a tedious task. You gather stats. You organize it. You calculate costs for you. So those are all library jargons that I'm talking about. But those things can be done by automated.

So if an effective AI could be programmed to analyze all the usage statistics and analyze them in hundreds of thousands of millions of ways, there's an opportunity to learn a lot more than just, like, humans analyzing it.

Yes. And also I played with it in ChatGPT. I was playing with it.

You said in the last segment that you had this idea-- the seed was planted for this idea of having a conference about four years ago. ChatGPT is about a year old. Was that kind of like the kickoff point when ChatGPT came out as a product? Did you think, oh, man, this is going to start moving at a rapid pace?

Yes. I think four years ago when I interviewed for a job, I presented about AI and how it's going to be So. Important for medical libraries. I worked in a medical library before Georgia Tech. And I presented it. I still have my slide from that presentation. And that's when I started thinking about, oh, I wish someone will start a conference so I can learn more about it because it's so complicated. I don't know what generative AI is, what is large language models, AI ethics, the algorithm bias.

AI is very much like the band Radiohead. I just don't want to know any more about it, but it's all anybody will talk about sometimes. It just seems like it has come at us so fast. And I mean, I'm hassling Fred a little bit about this, but also--

It's showing your age.

Yeah, right. ChatGPT was announced, and now we're digging into ethical issues in AI. It seems to be speeding up.

Yes, a lot. And the pace is so dramatic. And we just have to keep learning so we can be useful for the students and the people we serve.

Well, you're listening to Lost in the Stacks. And we're going to talk more with Anu about AI and the AI4Libraries conference on the left side of the hour.

Hi, there. This is Mark Riedl of Georgia Tech's Computer Science Department, or is it? Maybe my voice has been deep faked, and this is just a digital forgery created by a neural network. Either way, you're definitely listening to Lost in the Stacks on WREK Atlanta.

CHARLIE BENNETT: Today's show is called The AI4Libraries Conference. We just heard Mark Riedl, or his simulation, give a show ID. And he was also a speaker at AI4Libraries. Mark often writes about Large Language Models, as mentioned just a moment ago, or LLMs. In one article he wrote, titled, A Very Gentle Introduction to Large Language Models, Without The Hype, he writes, "LLMs have seen it all. They have seen billions of conversations on just about every topic.

You just heard Damage by Automatic; and before that, Would I Lie to You? By the Eurythmics. That set was a conversation. Would AI lie to us?

Yes.

And the result would be damage.

This is Lost in the Stacks. And today's show is called The AI4Libraries Conference because Fred figured out a way to have another AI-themed show this year

Yeah, that's right.

--because my blood pressure is not high enough.

[LAUGHS]

We're speaking with Georgia Tech's electronic resources librarian, Anu Moorthy, who founded the AI4Libraries conference. And we have sort of two shows running in parallel here because you just whipped up a conference in, like, three months because you were interested in a topic. And that topic is one of the flashpoints for information science in the world. I hardly know which way to approach this, so let's try and blend them. Do you wish you'd had some AI help to run your conference?

Yes.

And I'm hoping that in 10 years, if I still have the conference, maybe the AI will run it.

It'll program itself.

Yes.

So you do want to continue the conference? You want this to be, I'm assuming, an annual thing. Is that the plan? ANU MOORTHY: When did you-- when did you decide that you were going to keep doing it? Because it does seem like there would have to be a moment where you would decide whether it was successful or not.

I think even if we didn't have that 600 attendees, I decided that I'll continue to do it because it's a learning opportunity for librarians from around the world. They can learn about AI if they don't have the resources. This is an opportunity. They can just go and learn about AI from this conference archives.

I've almost forgotten what it's like to do conferences. I've become so comfortable in not traveling and not going to conferences that I kind of forgot that that's the key, is that you can exchange-- it's like reverse AI. A bunch of people choose the information they're going to exchange with each other in order to build more knowledge and more context.

Yes. CHARLIE BENNET: That's the last thing I'm going to do against AI. I'm going to back off now.

You're surrendering?

I didn't say that, Fred. And you know what? Now that you've said that, I'm back.

All right.

What do you want to do with this conference? You said if I have it in 10 years, people can come and learn. What are your big sort of strategic goals with the conference? How do you want people to find the information? How do you want to build from it? What are you imagining? ANU MOORTHY: Actually, I just think that everyone should learn from this, if they want to of course. It's there. It's a resource for librarians to learn from the experts and from each other.

I didn't want to do the conference propos-- it won't be peer review if I did not have more peers reviewing the conference proposals. So I had colleagues help me with the proposal review-- Heather and Jeff Coat and Martin Patrick, they both helped me with the reviewing the conference proposals. And Heather also helped me with on the conference day with recording sessions, because the mission control had three lab computers, emails, teams chat, and all the other things.

And I know that you had a lot of Georgia Tech people involved. You had Georgia Tech speakers. You had Georgia Tech technical support. But this is not a Georgia Tech conference. Can you talk a little bit about why it's not a Georgia Tech conference?

Well, Georgia Tech sponsored the Zoom for the conference. And I did it all as a service opportunity. I didn't want it to be tied to the institution, because most of the work was done during the weekends or nights. So I wanted to keep it separate because we don't know where-- because we all move on.

It's something that's important to--

Right. Well, it's kind of future-proofing. It's sort of imagining, I think, that the life of this conference will extend beyond your time at the Georgia Tech Library.

And its importance extends beyond just being at Georgia Tech.

Yes. That was the thought process. If I continue to do that for a long time, or if AI runs it.

And I kind of wonder because there's all these-- you have the number four in the name, and we have code for lib and some of those other-- and there are communities that have grown up around this. Do you also imagine a community growing outside of the conference?

Yes. I created a Slack channel where people could just join and chat about AI and what they're doing. I have, honestly, not promoted enough it after the conference. So I should probably-- there is a community of Slack where people can come and post ideas and they can learn from each other on top of the website.

And anyone interested can join the Slack?

Yes.

They'd probably just need to get in touch with you?

It's on the website.

OK. All right. I have to say just before-- I know we're right at time. But you mentioned kind of jokingly that you want to use AI to help you plan the conference. I just imagine a fun experiment might be asking ChatGPT or whatever platform comes next how to plan an AI conference and just track the responses from year-to-year how good it gets, or bad, or whatever.

Oh, that's a great idea. I did it for this year.

And of course, ChatGPT will not have access to any of the materials for another two years, so we'll have to--

Until they decide to open it up to past 2021, I guess.

--kind of wait and see. Well, this has been a delight. We have been speaking with Anu Moorthy, electronic resources librarian here at Georgia Tech and founder of the AI4Libraries conference. Anu, thank you so much for being on the show.

Thank you so much for inviting me.

And you can file this set under AS6.B85.

That was Afternoon Speaker by The Sea and Cake; and before that, Conference At Water House by Jah Stitch, songs about conferences and speakers.

CHARLIE BENNET: Today's show is called The AI4Libraries Conference. This homegrown, almost-DIY conference makes me wonder what subject might drive one of us to start a conference? I don't know that we'd be as successful as Anu.

Definitely not.

It seemed like a charm experience. Yeah. I imagine it would be something that we explored in A Lost in the Stacks episode, but not as thoroughly as we might have wanted. So let me ask you two, what episode of the 580 other episodes we have produced would be your conference? Fred?

I, probably as no surprise, am going to mention something that's fossil-related. I think the fossil librarian show where Kallie Moore came on to talk about how she catalogs her fossil collection and how she accesses that makes it accessible. I think expanding on that idea would be really interesting. I would probably be the only person to attend, maybe a couple of other nerdy paleontologists. I don't know.

Could you expand it out, though, and be like, museum collection items-- like, not just fossils--

Fossils only. CHARLIE BENNET: Just Fossils only? Yeah.

Yeah. OK.

This is my conference plan.

It is your conference plan. Marlee, do you have one? Maybe a more applicable one?

Yeah. I guess an entire conference on the museum of obsolete library science would not really fly, but--

You can have a session on each individual object.

Yeah, probably. Yeah. Or the bricks. I really enjoyed the one about bricks and the material oomph. And no, I think to be completely serious, it would be disabilities and libraries. Yeah, absolutely.

That would take a lot of work.

Yeah. What about you, Charlie?

I was looking over the episodes just today. And one that stuck out to me was actually episode 549-- How Do You Solve A Problem Like Twitter? which was kind of a joke at first because this dude bought Twitter and now he's doing stuff to it. What are we going to do about that? There's professional communities built into it. It's a public square, all that. But now, the problem of Twitter is the problem of all social media and instant response communication. How does it model citizenship?

So there's overlap with Anu's conference. CHARLIE BENNET: Yeah, yeah, yeah. AI is not going to kill us. It's just going to make us super stupid. OK, roll the credits.

Lost in the Stacks is a collaboration between WREK Atlanta and the Georgia Tech Library, written and produced by me, Charlie Bennett, and Fred Rascoe and Marlee Givens.

That's me.

That's you.

Legal counsel and a mobile command station for a virtual conference.

It's huge.

Ooh, yeah. Were provided by the Burrus Intellectual Property Law Group in Atlanta, Georgia.

That's handy. Anu could use one of those next year. Special thanks to Anu for being on the show, to everyone who helped make AI4Libraries a success. And thanks, as always, to each and every one of you for listening.

And no thanks to Fred for all the AI episodes.

Yeah.

Our webpage is library.gatech.e du/lostinthestacks, where you'll find our most recent episode, a link to the podcast feed, and a web form if you want to get in touch with us and suggest a show idea that's not AI.

Next week is a post-turkey day rerun. And we'll be back the week after that to talk about those three little letters that always seem to get a big reaction-- C, R, t.

Uh-oh.

It's time for our last song today. As long as there are academic libraries, there will always be conferences. It's just a fact of our profession that we continuously move from conference to conference. So let's close with Conference to Conference by Iguana Death Cult. I think ChatGPT might have named that band. Iguana Death Cult right here on Lost in the Stacks. Have a great weekend, everybody.