Catch up on Tech Talks from CNI’s 2025 Conference

Three presentations on AI applications in libraries

Posted on July 16, 2025 in Blog Posts

The Coalition for Networked Information (CNI) held its annual conference in April this year, featuring a number of cool talks on—what else?—AI in academic libraries. One participant observed that the conference struck a good balance between exploring the current state of library technology and studying the emerging landscape, offering practical guidance for navigating this space. You can read LibTech Insight‘s coverage of the event. In this post, we wanted to share recordings of some of the talks we didn’t want you to miss.

---

🧠 “Artificial Intelligence Literacy: Building Competency, Confidence, and Collaboration”

This talk features an all-star lineup of Leo Lo (University of Virginia), Niamh McGuigan (Brown University), and Oren Beit-Arie (Clarivate). They discuss practical strategies for integrating AI literacy initiatives into library services, upskilling library workers, and forging partnerships between the library and other institutional bodies. This talk emphasizes that librarians can be leaders at their institutions and offers a pathway for building this authority.

📚 “Unlocking Web Archives Using Retrieval Augmented Generation”

One term that all AI-literate information professionals need to commit to memory is Retrieval Augmented Generation or RAG. We wrote a primer on this concept, but in short, RAG refers to a specific approach to AI output generation. Using RAG, the AI model creates its outputs based on external knowledge sources (for instance, a specific database or archive) rather than the whole of its training data. This has immense relevance to libraries interested in integrating generative AI into their archives and digital collections as a way for users to interact with these datasets. In this talk, Corey Davis from the University of Victoria explores the affordances and pitfalls of using RAG for web archives. If your library has floated the idea of coupling its digital preservation efforts with AI, this talk will provide a vital foundation for that conversation.

✍️ “Artificial Intelligence for Transcription and Metadata in Special Collections: Three Projects”

Sometimes, what is lost in conversations about generative AI is that it, in theory, should make our working lives easier. This video features three case studies of using AI to create metadata and improve transcripts—important but highly time-consuming tasks. (Full disclosure: I used to copyedit long video transcripts, and it was awful.) Many archives have massive catalogs of data, which would take years of human labor to make accessible. The three presenters outline the successes and struggles of integrating this technology to take on these tasks.

🌟 Choice and Clarivate have launched a free, eight-week micro-course on AI literacy. Register now!

Related Posts