By Niki Tavakoli ’27
To Dr. Keri Lambert, Instructor in History and Social Sciences at Andover, artificial intelligence is a tool worthy of exploration. In fact, she has incorporated artificial intelligence into her lesson plans to prompt students to explore both its advantages and limitations. Dr. Lambert looks forward to learning more about how to incorporate artificial intelligence into her class while teaching students critical thinking skills.
Though artificial intelligence, specifically large language models (LLMs), is excellent at providing factual information on a wide variety of topics, it lacks the skill a human brain has to think analytically and creatively, a skill Dr. Lambert aims to develop in her class. One of her lesson plans prompts students to examine the capabilities of LLMs.
“I had a lesson devoted to exploring how ChatGPT would be able to write a paper in response to a prompt that my students would have to write a paper about. From that exercise, many students took away an understanding of some of the limitations of historical thinking produced by artificial intelligence. We talked a little bit about the skills that ChatGPT doesn’t currently have, in terms of analytical thinking, critical thinking, strong writing, [and] having a real clear sense of voice. All of those things are ones that we’re going to be working a lot on in my history [classes],” said Dr. Lambert.
LLMs have a limited ability to think critically about history. They tend to lack substantive analysis and sideline the perspectives and roles of marginalized groups in the event. For example, we asked ChatGPT to explain the American Revolution. It only briefly noted the French and Spanish support for the colonists and did not mention the substantive role alliances with Native Americans played in the war. In Dr. Lambert’s History 300 class, a central question is when US History began. Lambert emphasized the variety of possible answers to this question but noted that LLMs tend to give the most basic ones.
“[ChatGPT] is going to answer in a way that most often reflects the most dominant narrative. It might answer 1776, it might answer 1492, kind of the well-known starting points that center white actors. You can prompt ChatGPT [by saying] … can you tell me some starting points that emphasize other perspectives? And it might spit some interesting starting points back at you. But that relies on the kind of vigilance of the prompter. And if a reader or if a person using artificial intelligence isn’t vigilant and doesn’t want to prompt ChatGPT to check its own implicit bias, then the narratives that ChatGPT will create about history will be far simpler than the ones that we want students to be able to grapple with.”
However, Dr. Lambert wants to incorporate artificial intelligence to prompt critical thinking, recognizing its potential to help students learn. She has even taken a workshop series on the topic.
“There’s a lot of cool possibilities about how you could use ChatGPT and other artificial intelligence machines to sharpen arguments, think about counterpoints to an argument, [and] to think about the different ways that we can ask questions about the past. What I’m grappling with right now is… how can we keep critical thinking in the students’ court and creative thinking as a priority for students while also using artificial intelligence to prompt and challenge and maybe inspire some of that creative and critical thinking? … That’s the kind of stuff that I really want to develop both from my own exploration of these different resources and professional development opportunities and in collaboration with some of my other colleagues both within history and outside of history this year.”
Overall, Dr. Lambert’s goal is to give students a balanced and complete view of the advantages and weaknesses of AI. She hopes to help students critically analyze events in their past and present, skills that come through studying and analyzing history. Lambert says that whilst many humanities teachers tend to view AI primarily as a tool for cheating, she wants to emphasize that AI has both risks and potential.
“I think the big priority for me is to help students to see its possibilities, to see its risks, and to see how much the human brain and our individual ways of approaching material has huge advantages over how a robot [or] how a computer can approach thinking about the past or thinking about problems in the present that can be informed by our historical knowledge. I hope that we can shift the conversation from one being or exclusively about academic integrity in the humanities to one being about helping us to have a sharper understanding of what can be learned by studying the past, and what can be learned by practicing the different communication skills that we do in a history classroom, so that we can ultimately use AI but use it in a responsible way and in an ethical way.”
Read more articles like this in our Fall 2024 Issue!
