Researching, Teaching, and Learning: History Students Using AI

Abstract

This is a pedagogy-focused paper that incorporates what I learn (and students learn) from two assignments in an advanced history course. The students will be required to use AI in a controlled manner (I set the terms of the assignments: for the first, their task is to read a book and write a book review—not summary—of it, then ask an AI to write a book review of the same book, then analyze and assess the two outputs; for the second, they will ask the same question of multiple AIs to compare results generated on a given topic—what primary sources will AI point them towards? they'll then also write an assessment of that exercise). These assignments are taking place in a fall 2024 semester course, so I cannot know what the results are as of September 2024, but my theory is that a) most humanities students will have very limited previous experience using AI or none at all, b) many may assume that an AI can write a better book review than they can and be surprised by the results (even going so far as to assume that an AI-created book review would receive a higher grade), and c) students will assume that all AI is the same (and therefore all outputs to the same question will be identical). I intend to adapt what I learn from these exercises to later research-based courses I teach history students at advanced levels.