Academics and AI: Calming the farm.

Recently I met with a student who needed to complete a rapid systematic literature review in 8 weeks. The students can refer to Cochrane Methods and our library’s systematic review guide. A review is a difficult task for any novice. But the students were encouraged to use generative AI (Artificial Intelligence) for ‘scoping of the topic’, but not for extraction or analysis. Whoa! Calm the farm!

I aim to collaborate with academic teaching staff, putting together some guidance for them to create assessment tasks that practically embrace AI, so that it's feasible for students to understand how:

1. To produce systematic types of reviews in short timeframes

2. To use suitable AI tools when doing reviews for the first time

Using this particular unit assessment as a test case, I’ve contacted the academic and highlighted how this assessment may prove overwhelming for the students. I’ve suggested we collaborate to develop a more structured task. I’ve started collating examples of what we should be trying to avoid (wholly AI-generated reviews) and possible helpful AI tools that students will find easy to use for free.

At the time of writing this abstract, I’m unable to report on results or conclusions specific to this case, but we have a growing list of units in health disciplines being taught at Monash, setting reviews as assessments. It is logical that students will turn to AI with short timeframes, but we need to guide them in doing these tasks with academic integrity. There is an abundance of information available on doing reviews using AI. As librarians, we need to determine what will work best for our students and staff now, and possibly down the track.

Presented at the ALIA Health Libraries Australia (HLA) Conference 2024. Aug 29-30, 2024, Caboolture Hospital, Qld. https://hla.alia.org.au/alia-hla-conference-aug-29-30-2024-caboolture-hospital-qld/