Artificial intelligence (AI) is hailed for enhancing teacher effectiveness by expediting lesson plans and classroom materials creation. However, a pioneering randomized trial in real classrooms revealed that AI might hinder learning. Students with teachers using an AI teaching assistant reported diminished motivation to learn. The trial, conducted by University of Pennsylvania researchers, highlighted AI’s potential drawbacks in education.
Particularly impacted were students taught by weaker instructors, as identified by pre-experiment performance. These students scored lower on standardized exams. “Teachers, just like students or coders, might be using AI as a crutch,” said Alp Sungu, lead author and assistant professor at the Wharton School. The study, “Generative AI Can Harm Teaching,” released in June, has yet to be peer-reviewed. It builds on Sungu’s earlier research on AI’s adverse effects on student learning.
AI’s use as a “material generating machine” may replace teachers’ efforts in crafting personalized lesson plans and syllabus, leading to lower quality output. Conducted in Turkey with 193 teachers and over 2,800 students, the study assigned teachers to either use a ChatGPT-based assistant or continue traditional methods. The AI tool was primarily used for generating lecture notes and assignments.
Students found AI-aided classes less engaging, particularly if their teachers were previously heavy AI users. While overall academic achievement remained stable, students of lower-performing teachers saw reduced achievement and confidence. The study suggests that AI might strip teaching of personal voice, leading to uniform and less engaging content.
Sungu proposed that stronger teachers adapt AI outputs, whereas weaker ones might use them unchanged, affecting educational quality. The study also compared teachers’ access to customized AI against their individual choices of AI tools, suggesting a potential underestimation of AI reliance risks.
Despite these findings, Sungu warns against dismissing AI’s educational potential. Instead, he advocates for training programs and guidelines to optimize AI use while preserving human creativity and judgment. Sungu uses AI to develop interactive teaching tools, though he emphasizes the necessity of personal involvement to refine AI outputs.
—
Read More Kitchen Table News
