Learning About Generative AI in Academic Libraries

Building on previous blog posts on generative AI and professional development, I recently completed the Google AI Essentials Coursera course. Course enrollment was sponsored by the Pitt Libraries and all library employees were encouraged to complete it. The course was an introduction on how generative AI can be used to simplify tasks and how to refine AI prompts to improve outputs. The limits and ethics of using AI were also discussed, but the lens was certainly not a critical one; Google’s Gemini AI tool was embedded in the course, surely in hopes of hooking new customers. Despite its obvious bias, I did find the course helpful. It was my first experience using any kind of AI chatbot and it’s very easy to see gen AI’s usefulness and why many are enthusiastic about its potential.

That said, I remain deeply ambivalent about gen AI. My main reason for not using gen AI tools before is their alarming (and growing) environmental impact. I also don’t see its practical application to my daily work at the moment. The issue of AI hallucination is very serious, particularly because there’s no way to eliminate it: “fundamentally, LLMs aren’t designed to pump out facts.” Finally, the possibility of using AI to short-circuit learning and development of critical thinking are potentially destructive.

Like all of your institutions, gen AI is already a big topic of discussion here, where Pitt GPT is being developed for use by Pitt affiliates. This doesn’t appear to be an instruction-focused tool at the moment, so its impact on library work may not be significant. There was also a recent discussion of Gen AI’s use on Pitt campuses. One faculty member mentioned gen AI usage is “an additional thing we have to police” without having great support for doing it. Focus groups with students indicated extensive use of AI, particularly to do what they see as course “busywork.”

Gen AI is here to stay, so it’s my responsibility to continue to learn about it. For further learning, here are some articles and podcasts on AI I recently read/listened to:

• AI in Academic Libraries from C&RL News (part 1 and part 2) – It’s helpful to hear how other university librarians are grappling with these issues. The participants are either skeptical of AI or outright resistant to it.
• Fostering Critical Thinking in the Age of AI: Why Information Literacy Still Matters – My PPIRS colleague Kim MacVaugh argues that the ACRL framework can help students critically evaluate the use of AI tools. Political scientists have been very interested in her thoughts on instruction, which is encouraging.
• Capstone 2.0: Elevating Research through AI Literacy – Political Science professor Shannon MacQueen integrated AI into a political science capstone class. It’s a small sample size, but the experiment went well and the students thoughtfully used AI.
• Will the Humanities Survive Artificial Intelligence? – An engrossing essay from Princeton historian D. Graham Burnett recounting his journey with AI and how it will upend higher ed.
• Ezra Klein Podcast interview With Rebecca Winthrop – Winthrop, an education researcher, believes AI could further erode already-declining student engagement. When used thoughtfully, though, it could be a great benefit for some students.
• Ross Douthat interview with Daniel Kokotajlo – Kokotajlo is a coauthor of the AI 2027 report, which attempts to game out how AI will evolve over the next few years. Kokotajlo believes AI will completely upend society and that something like SkyNet is a possibility. Nightmare fuel!

Finally, I’d like to build on a blog post about library renovation and celebrate the completion of the Hillman Library renovation! After 8 long years of complete renovation, it’s now a totally different building. If you’ll be in Pittsburgh any time soon, please stop by and see it.