Odra Noel. Altmann’s Bioblasts – The Four Seasons (Winter). n.d. Wellcome Collection.

Generative AI has entered the classroom, poised to profoundly impact both teaching and learning. As educators navigate this digital shift, they’re pioneering the use of emerging tools and resources. Sarah Kiyak, a teacher at Alexandria High School in Virginia, recently piloted JSTOR’s GenAI-powered research tool (currently in beta) in her Dual Enrollment (DE) Composition class. Beyond helping students find and assess sources, JSTOR’s tool empowers teachers to create a more engaging and effective learning experience for their students. In this blog post, learn how Kiyak used JSTOR’s AI tool to deepen student engagement, instill crucial research skills, and streamline her own workflows.

JSTOR’s GenAI comes to AP history

After being introduced to JSTOR’s AI research tool in December 2023, Kiyak saw immediate potential for its use in the classroom. Because its core capabilities—providing a summary of an article, suggesting related topics and content, and answering questions about a text—are designed to make evaluation of sources more efficient, Kiyak chose to introduce it to her DE scholars, who were about to begin work on a 12-page research paper. Like she’d done many times before, she first (re)introduced her students to JSTOR and the platform’s features. But this time she had a surprise for them. 

“When I shared how we had special access to GenAI, the students hooted and hollered!”

Just how it will be adopted in the classroom is a fast-moving subject, with a fair amount of debate. While some teachers have expressed concern about bringing GenAI into the classroom, Kiyak is clear about her stance. “I think it is vital that we teachers model healthy ways to utilize advanced technology. We are using AI as a resource, not a way to cheat the system. I demonstrate how to use AI for good, not evil.” 

Making research efficient and accessible 

Novice researchers who are just developing their skills find it daunting to write long research papers with many citations. Sifting through dozens of articles trying to find the ones most relevant to the topic can be challenging, even for expert researchers. With GenAI, research that used to feel like looking for a needle in a haystack has been simplified.

To introduce the tool to her class, Kiyak said, “I projected my Boolean searches onto my smart screen, chose some articles, and showed students how Gen AI provides a concise summary of each article.” Via the summary, students could rapidly assess the relevance of the article, seeing if it incorporated their keywords. They could also ask the tool directly what the article says about a certain word or phrase. 

It immediately empowered her students’ research.

“Gen AI, again, helped us understand the scope of the research articles and if they were pertinent to each student’s topic.”

“[Three students] then wanted to shift into [researching] Obama’s administration and the rise of the Tea Party in the 2000s. With this specific historical shift, the Gen AI helped the students hone in on exactly what they needed in their resources. Early in their research, if the AI summary didn’t incorporate any of these keywords (Civil Rights, Southern Democrats, Obama, Tea Party), then they knew they needed a different source.”

Academic articles can be written in infamously inaccessible language, and when students struggle to identify what is relevant they often spend time reading articles that aren’t, only to realize later that they’d wasted their time.“Gen AI, again, helped us understand the scope of the research articles and if they were pertinent to each student’s topic.” Wasting time is frustrating, and for novice researchers, that frustration can stifle the intrigue and curiosity that fuel the pursuit of knowledge. 

GenAI lowers the barrier to entry without performing the intellectual work. This frees up bandwidth so that it can be used to enhance understanding of the material and make connections with other texts. The AI-powered summary of the text can even deepen students’ comprehension, giving them a framework for what they are reading, just as knowing the plot before reading Shakespeare can make his plays more intelligible. The AI-generated responses use only the text at hand to generate responses, which minimizes hallucinations and misinformation. To ensure credibility, it displays text excerpts and provides citations via direct links to the relevant part of the text. 

For a deeper dive into the tool’s capacities, read about the many ways JSTOR’s GenAI tool can empower research. 

GenAI as a teaching assistant

Teaching research skills to a large group of diverse learners can also prove challenging and time consuming. While one-on-one sessions can be designed to suit each student’s needs, they are resource intensive. “Gen AI helped me work more quickly with my students during our one-on-one conferences,” Kiyak explained. “I could find sources and then easily verify if the research focused on the students’ topics. Gen AI truly helped me with my own time management in the classroom.”

Teachers everywhere have noticed a decline in student engagement. By eliminating the more tedious part of research and getting right to the intellectual stimulation, teachers have a better chance of keeping students’ attention. So far, users of the GenAI beta tool are staying on the platform twice as long, showing evidence of its incredible promise to boost engagement.

JSTOR’s GenAI beta tool was generated deliberately and collaboratively, aware of the concerns within higher education and beyond. To learn about the guiding principles that govern this work, read “Approaching generative AI together,” an open letter written by ITHAKA president Kevin Guthrie.