Failing grades soar with AI usage, dwindling math skills in Berkeley CS classes

Article excerpt

Failing grades in UC Berkeley's computer science program have surged as professors report increased student reliance on AI tools and a decline in fundamental math skills. Faculty members attribute the grade spike partly to students using ChatGPT and similar language models to complete assignments without genuine understanding, while also noting that incoming students struggle with calculus and discrete math prerequisites. The trend raises concerns about whether AI shortcuts undermine learning outcomes in technical fields that traditionally demand rigorous problem-solving abilities. Berkeley instructors are weighing how to adapt grading and pedagogy as AI becomes ubiquitous in student workflows.