A Brown University economics professor observed a significant drop in student performance when an AI-assisted take-home exam with an average score of 96% was replaced by an in-person, proctored final. The average score plummeted to 48.6%, with 18 students dropping the course and nine not attending the final exam. This observation is supported by broader studies from China and UC Berkeley, which indicate that students who rely heavily on AI for assignments tend to perform poorly on supervised assessments. AI
IMPACT Suggests that over-reliance on AI tools for coursework may hinder students' ability to perform on traditional, supervised assessments.
RANK_REASON Article discusses an observation and cites studies regarding AI use in education, fitting the commentary bucket.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
