
This past August I took over as Director of the Hanson Center for Communication in the College of Engineering after teaching graduate students how to write and present their work effectively for the past six years. The Hanson Center serves hundreds of engineering students every year and is dedicated to teaching undergraduate and graduate students best practices in writing and presenting through workshops, writing groups, competitions, and formal courses.
With AI tools for communication like ChatGPT and Co-Pilot becoming ubiquitous, students and instructors are asking more often than ever, “Do students even need to learn how to write anymore?” (In just the last two months I have been invited to classes and workshops to discuss the use of AI as a writing tool on three occasions.) My answer is a resounding “Yes!” For scientists and engineers, the value of writing has always been much more than the written product. The process of writing is a distinct form of thinking. When we write about our science, we can see clearly what we know but, even more critically, the gaps in our knowledge become abundantly clear. This clarity in what we know and don’t know then allows us as researchers to dive into the literature or design new experiments to fill those gaps. Furthermore, since we have used writing as a way of thinking, we know which gaps to focus on. After all, we each have an infinite number of blank spots in our knowledge and they don’t all need to be filled to make a contribution to a specific field. Writing exposes the gaps but also helps us identify our priorities and this can give students confidence and clarity about how and where they should devote their energy.
AI writing tools can write drafts for you but it cannot think for you. While many people bemoan how formulaic and generic AI writing is – and it is – I would argue the larger cost to scientists is that AI tools will not force you to think critically about your subject. They will not ask you to consider the basis for the claim that you are making, how your claims relate to each other, what you really know to be true, or what is really the most important question. No, these considerations come to mind as you write, pen to paper, fingers to keyboard, and cannot be outsourced. In summary, writing is thinking and AI cannot replace the learning that comes with struggling with your ideas in your own mind. In summary, even if you never show anyone else what you have written, the value of thinking while writing makes it a worthwhile skill to develop and hone, even in the age of AI.