Elizabeth A. Drummond

PDF of Dr. Drummond's Article | Dr. Drummond's Reflection

When Machines Write: Thinking Through Some of the Challenges that Generative AI Poses for the Nature of a Liberal Education 

Written by Elizabeth A. Drummond 

As a faculty fellow this semester, I likely played an important role in supporting the academic community at my institution. Depending on my specific responsibilities, I may have engaged in a range of activities, such as mentoring students or junior faculty, participating in curriculum development or assessment, leading or participating in faculty workshops, or contributing to research or creative projects. I may have also had the opportunity to collaborate with colleagues from different departments or disciplines, or to engage with community members or stakeholders outside of the institution. Through my work as a faculty fellow, I likely helped to foster a culture of learning, collaboration, and innovation on my campus, and made a meaningful impact on the academic community. 

That generic and uninspiring paragraph came to you courtesy of ChatGPT, in response to me asking it to “write a paragraph about what I did this semester as a faculty fellow.” So, what did ChatGPT get right? Yes, I led and participated in faculty workshops, as well as collaborated with colleagues from different departments and offices. In particular, the CTE Faculty Fellow for DEI, Professor Mairead Sullivan (WGST), and I facilitated a discussion about how we can better assess teaching. We discussed what kind of feedback we want and need from students and colleagues, how we might develop more meaningful forms of assessment, how we can use them in more constructive and equitable ways, and how we can foster a culture of reflective learning among both students and faculty. I also engaged with community members on and off campus (I doubt that I would ever use the word “stakeholders” without scare quotes). These conversations will continue in the future, and I hope that I will contribute to a culture of learning, collaboration, and innovation on campus.   

I begin with ChatGPT because I spent more time than I had anticipated when I applied to be a CTE Faculty Fellow thinking and talking about generative AI this semester. As I mentioned at this spring’s Teaching with Technology Day and Innovation Symposium, I am an AI skeptic – skeptical of both its possibilities and its dangers. Discussions of ChatGPT, Dall-E, and similar tools dominated “the academic discourse” over the winter break and into the spring semester. Some were excited by the possibilities of generative AI, but more – at least in my feeds – were concerned about what generative AI, and ChatGPT specifically, might mean for our teaching, whether it might provide incentives for students to take shortcuts in their writing. 

For me, generative AI is a pedagogical challenge, rather than a technological challenge – and not a new pedagogical challenge (indeed, the conversations about ChatGPT are reminiscent of previous conversations about Wikipedia and even the Internet in general). How do we best teach our disciplines, their methods, and writing? When we think about research and writing as process, instead of outcome, and when we stress patient and careful work with analyzing sources, building arguments with evidence, and drafting and revising, then we are already starting to build the scaffolding that works against an easy reliance on ChatGPT. When we create space for imagination and student agency, allowing students to choose their research topics and presentation formats, then they often get excited about their work (and take pride in it!), rather than see assignments as just obstacles to overcome. When we think about the reasons why students take shortcuts, then we might be able to develop approaches and policies that undermine the incentive to do so, rather than policies that police and punish students, in the process treating all students from a position of distrust.  

That is not to say that we can just pretend that these tools don’t exist. In the fall, I plan to discuss them with my students, considering them in the context of why it is important that they learn to develop research questions and to struggle with writing. I also plan to invite them to co-create guidelines for their use, as part of a broader discussion about class expectations, academic integrity, and what it means to be part of an intellectual community. I was inspired to do this by reading about a course at Boston University, where the students drafted a policy about AI use that was subsequently adopted by the department.  

I’m skeptical about AI’s dangers, but I am also skeptical about generative AI’s possibilities. Rather than see ChatGPT and similar platforms as useful tools in academic writing, I am concerned that generative AI contributes to the undermining of the humanities and the arts. In reducing knowledge and creativity to formulaic writing and mimicry produced by a machine, it works against curiosity and wonder and beauty and experimentation and failure – all those things that make us human and are essential for true creativity, for true innovation, for the production of knowledge or art that can be truly groundbreaking and revolutionary. In doing so, generative AI contributes to the trend towards instrumentalizing higher education, towards seeing education merely as pre-professional training and educating students for the workforce, towards higher education as the acquisition of a credential.  

But that’s not the purpose of a liberal education in the Jesuit, Marymount, and Catholic traditions. Rather, a liberal education is a journey of learning, curiosity, and exploration where we grapple with complexity and challenging ideas. It is an education for democratic citizenship, for thinking about the kind of society in which we want to live. And when we undermine that kind of education by saying that AI can do our thinking for us, we undermine our ability to engage in democratic citizenship, to evaluate truth claims, to question our leaders and the system, to see through disinformation (much of it fueled by these very technologies), and more. As we as instructors grapple with emerging, and rapidly evolving, technologies such as generative AI, I hope that we think not just about the immediate issues related to academic honesty and assignment design, but also the fundamental issues related to the mission and purpose of a liberal education.