by Jesse Millek, Erasmus University Library, The Netherlands.
Since the launch of ChatGPT on November 30th, 2022, Generative AI (GenAI) has woven itself into numerous facets of life. Yet, for university students, the adoption of GenAI has left them between the proverbial rock and a hard place. In this case, the “rock” is GenAI companies and AI advocates who are constantly sending the message that GenAI is the future with some even saying, “don’t learn to code; learn to prompt.” This places enormous pressure on students not only to use GenAI, but it brings with it the fear that if they don’t become proficient in it, they may be left behind in an ever-changing job market.
However, there is the other side to this idiom, the “hard place,” which in this case are institutes of higher education who are broadcasting a very different message about GenAI. Most universities have put a ban, partial or complete, on using GenAI for fear that students will miss out on key facets of their university education such as learning how to think critically, working through complex problems, and knowing how to put research down on paper. To combat this, policies have been put in place which dictate that the illegitimate use of GenAI will have academic consequences. Thus, we have a paradoxical situation where students are told that to make it in the future job market, they must learn to use GenAI, but they are also left to wonder if asking ChatGPT for assignment title recommendations will get them accused of plagiarism.
Two modules on using GenAI in the university
This is where the Erasmus University Library steps in. As a university library, part of our job is teaching students how to research and access trustworthy information. Teaching students how to navigate the complex world of ethically utilising GenAI is a natural step for the library to take. To bridge the gap between the two different messages students are receiving, we created two e-modules to teach students how to best use GenAI: when to use it, when not to use it; when to trust it, and when not to trust it. The two modules are separated into one focused on bachelor’s and master’s students, where the use of GenAI is more restricted, and one for PhD candidates who have more freedom when using GenAI. Both modules are broken into four key sections: 1) The rules to follow; 2) Things you should watch out for; 3) A brief intro to prompting; 4) Some potential uses for GenAI.
In the “The rules to follow” this is where we outline the don’ts of using GenAI including the obvious of not letting it write a whole essay, to the less noticeable rule of not putting in private or sensitive information into the GenAI. The companies will use this as training data for the next version of the AI. Meaning, personal details or private information could leak out if fed into a GenAI. While these are rules that students should stick to, there are other important factors to take into consideration before using GenAI which we cover in “Things you should watch out for.” This includes details about GenAI “hallucinating” as not all facts and references written by GenAI are true or real. Much like any source of information GenAI needs to be double checked. Additionally, having a GenAI doing the thinking and writing for you will erode and inhibit your ability to research and write. As Ted Chiang so succinctly put it in an essay for The New Yorker, “Using ChatGPT to complete assignments is like bringing a forklift into the weight room; you will never improve your cognitive fitness that way.”
With knowing what the rules are and being brought up to speed on some other potential problems that students could run into, the guides then move on to how to use GenAI beginning with “A brief Intro to prompting” where we lay out the basics of prompt engineering. Finally, we come to the section, “Some potential uses for GenAI.” Here we provide students with suggestions for how they can ethically use GenAI such as discussing essay topics, helping to find keywords for searching databases, brainstorming, or asking the GenAI to clarify academic concepts or complex jargon. However, even knowing the rules or the other issues they might run into when using GenAI, we still must remind students that not all these tips and suggestions will be allowed by all faculties or all lecturers. It is up to the student to be informed about what is and is not allowed in their specific case. At the Erasmus University Library, we hope that these guides enable them to better navigate the ever-changing world of GenAI in higher education.
Dr. Jesse Millek is an Information Specialist in Digital Literacy and AI at the Erasmus University Rotterdam’s University Library
