By Georgia Jackson, College of Arts and Sciences
Who are you — and what kind of writer do you want to become?
The question, which is becoming ever more urgent as generative artificial intelligence (GenAI) becomes increasingly prevalent in digital life, animates Anuj Gupta’s course “Writing with AI.”
The USF College of Arts and Sciences hired Gupta, an assistant professor in the Department of English, this fall, to support ongoing efforts to help student writers succeed in a world saturated with GenAI tools like OpenAI’s ChatGPT, Microsoft Copilot and Google Gemini. Drawing on his own research, which uses humanistic and social scientific methods to understand how writers interact with GenAI, Gupta is providing his students with a framework for thinking deeply about these new tools and navigating the rapidly changing environment.
Anuj Gupta teaches in the Department of English.
“It is both an exciting and scary time to be a college student. GenAI tools present the possibility of democratizing access to knowledge, innovation and productivity, and many employers want students to have GenAI literacy,” said Gupta, who completed his doctorate at the University of Arizona earlier this year. “But using these tools presents many ethical dilemmas to writers — copyright, plagiarism, environmental impact and automation of jobs being just some of them.”
Gupta’s task is twofold: help students navigate the benefits and harms of the rapidly evolving technology and encourage them to develop their own principled stances so that they may find meaningful employment and become leaders in their communities.
In his classes, which he is developing in collaboration with colleagues in the Department of English, Gupta begins by situating GenAI in history, alongside other writing technologies that have shaped human thinking and communication.
“There is a vast range of tools that humans have built to help them write, from pens and typewriters to word processors and search engines, and now GenAI. By impacting how we write, the tools we choose also shape how we think,” he said.
Gupta tackled GenAI in his doctoral dissertation, titled “Learning to Talk to GenAI Chatbots: A Corpus Study of GenAI Prompts, An Emerging Genre for AI Literacy,” and continues to collaborate with scholars across disciplines on the subject. At USF, he is developing what he calls the “4 Ps” model of GenAI literacy, which emphasizes the importance of GenAI products, processes, policies and public connections. Students in his class learn to analyze GenAI technologies from all four angles as they strive toward literacy.
Gupta's model emphasizes the importance of GenAI products, processes, policies and public connections. Students in his class learn to analyze GenAI technologies from all four angles as they strive toward literacy.
“First, I invite students to experiment with different AI products. What are the affordances and constraints of different GenAI tools available to writers? How do we evaluate the quality of the outputs they produce?” Gupta said. “Then we think about processes. What are the ways in which these tools produce those outputs? What kind of training data goes into creating them? How is it collected? What is the role of prompt engineering? How do you write good prompts? In what ways can GenAI tools augment or disrupt your writing processes?”
When it comes to policy, Gupta introduces his students to AI acceptable use policies (AUPs) published by various institutions, including universities, academic journals and workplaces. He likes to present his students with AUPs that take different stances on GenAI use.
"You will be expected to use AI in some workplaces, but one of your classes might say absolutely no AI, whereas an academic journal might allow it on a case-by-case basis,” Gupta said. “Being literate in GenAI means being able to navigate this spectrum on a regular basis.”
Finally, Gupta asks his students to consider the public's connection to GenAI.
“What kind of court cases are going on right now about copyright issues related to GenAI? How are communities pushing back against big data centers that could harm the local environment?”
His goal isn’t to persuade his students to use or renounce GenAI but rather to equip them to make an informed choice themselves.
“AI literacy does not necessarily mean that you should be pro-AI or anti-AI. Rather, I see it as a set of functional and critical skills that help individuals feel confident as writers navigating a very complex world mediated by emerging AI technologies. Irrespective of whether you use GenAI tools, being GenAI literate means you should have the vocabulary to explain to your employers, your teachers, your coworkers — and, most importantly, yourself — your principled rationale for your choices and how they make you a better writer.
“I think that in some shape or form GenAI is poised to become as ubiquitous as word processors and the internet in the fabric of our digital writing lives,” Gupta said. “What role it performs there and how writers engage with it, though, are still hopefully very much open questions that our students will answer with the writing they do in the world.”