The gap between 96 and 48.6

A Brown economics professor gave his first-ever take-home midterm this spring. The class average was 96%. His midterms usually average between 65 and 80%, and this exam was harder than his past ones. He ran it through ChatGPT, and the answers mirrored what his students had submitted. The giveaway was the style. On one question, a direct argument was the obvious route. ChatGPT and many of his students used a convoluted proof by contradiction instead. Correct, but oddly contrived. Then he made the final exam in-person. The class average was 48.6%, the lowest in the history of the course. Eighteen students dropped, and nineteen failed. Inside Higher Ed tells the fascinating story of Prof Serrano’s attempts at pursuing this with the Brown Ethics committee and the university’s meek response. Putting aside pontification on what this means for education, the gap between the midterm and final performance reminded me of AI work slop. The graph hilariously differentiates between those who…

Commenting on Bubbles has been removed. Why? If you want to talk about this post, write to the author, use the comments on their blog if it has them, or reply with a "RE:" post on your own blog. Suggestions? feedback@bubbles.town