The Newsletter took a back seat during a bustling week last week. Part of this was due to attending a Canadian Urban Libraries Council event in Montreal hosted at the La Bibliothèque nationale acquiert (BaNQ).
I participated in the Public Library Leadership program there—as an alumnus and a mentor—so I was fortunate to rub shoulders with many other professionals from across the country. As you can imagine, these meetings are always a time for sharing great ideas and discussing imperative topics such as the library’s role in society.
At the event, I continued to learn more about our industry, made valuable connections, and delved into various imperative topics. One of the most intriguing discussions was on the role of artificial intelligence in our libraries. It was interesting to hear how libraries were experimenting with AI technology.
I thought I’d share three interesting pursuits below. If the library hasn't publicly shared its project yet, I’ll keep individual names anonymous.
Using AI to Build a Handwriting Analysis Program
While on a tour of BaNQ itself, our group came across a kiosk on the main floor where staff were collecting handwriting samples. Each participant—anonymous to the project—would receive several pre-typed paragraphs, which they would transcribe onto a separate sheet. The project team would collect and scan the samples into Kraken to teach the machine to decipher handwriting. The goal was to train the technology to understand better, identify different generational handwriting, and interpret writing more effectively. One of the objectives was to build a model advanced enough to read artifacts and historical documents.
A Library Chat Bot
The common ‘chat’ feature is often found on the home web page, which can help customers tackle simple questions about the library. Humans usually monitor these, but some organizations have been turning over their FAQs for bots to handle. A representative from one library system shared that they were working with a local college to develop a proprietary chatbot for the library’s website. The college students were developing the tool as a project for their program. They are currently working on its development with the library, gathering information about the library along with common questions to feed to AI.
The City University of New York launched a similar project several years ago. Students worked with the library to feed the bot information about it and simulate queries to train it to answer questions. The 2022 study noted that the pilot provided data on student behaviour and interactions with the library while freeing up some time for librarians to tackle more in-depth reference questions. It will be interesting to hear about more libraries experimenting with similar tools.
Cataloguing AI Materials
My own system has not yet crossed this bridge, but a discussion on AI-generated books has me in planning mode. Another system shared how they were receiving AI-authored books to catalogue, which was an entirely new but not unexpected frontier of AI adaptation.
Luckily, the Library of Congress has a committee on top of this discussion. The Program for Cooperative Catalogue Committee has published some FAQs on cataloguing resources generated by AI, addressing key questions identified as related work rather than an agent, even when the resource attributes creation to the AI. When known, the human agents who program or operate the AI should be recorded as related agents.
These discussions and initiatives highlight the exciting ways libraries are embracing new technologies. As we continue to explore and integrate AI into our services, we are enhancing our capabilities and ensuring that we remain at the forefront of innovation in serving our communities. The future of libraries is undoubtedly intertwined with these advancements, and it’s thrilling to be part of this transformative journey.