Pitt faculty split on AI tools for grading and lesson planning
Some professors use AI to save hours on grading while others avoid it for creative writing, citing reliability issues.
By Beatriz Soto ·
University of Pittsburgh faculty are integrating artificial intelligence into their daily academic routines, with usage patterns ranging from detailed exam feedback to high-level brainstorming. While some instructors rely on the technology to handle administrative burdens, others remain cautious about its accuracy and role in creative work.
The adoption of these tools follows a general trend in higher education. Research by Anthropic indicates that college professors are increasingly using AI for curriculum development, lesson planning, research writing, and grading. At Pitt, the implementation varies by department and personal teaching philosophy.
Anne Yust, a teaching associate professor of mathematics, has adopted the tools aggressively to manage her large student load. She teaches approximately 150 calculus students, a number that makes detailed individual feedback on exams difficult. Yust used Claude to generate varied problem sets, noting that the tool now rarely makes mistakes on calculus problems, a significant improvement from the previous year when it struggled with such tasks.
Yust also trained the software to provide short responses to student solutions on midterms. She reviewed a handful of comments and edited them before releasing the rest. Students reported that the AI-written feedback was useful, and attendance at her office hours increased as they sought to clarify their mistakes. Yust stated that she currently spends about 40 hours grading one midterm, a time sink she hopes AI will reduce to allow for more direct student interaction.
In the English department, usage is more selective. Maggie Jones, a teaching professor and contributing writer for The New York Times Magazine, uses Claude for finding primary sources and brainstorming. However, she finds the idea of using AI for actual writing "pretty horrifying" for journalism and personal essays. Jones noted that AI sources are not always reliable, so she verifies all material independently.
Weifeng Xu, an assistant professor of neuroscience, uses AI to create lecture notes based on her slides. She tells students the notes are AI-written but fact-checks them for accuracy. Xu explained that the tool provides students with an additional handle for review, itemizing knowledge points that would take her significantly longer to write by hand.
Ross Harrison, an adjunct associate professor of political science, uses AI sparingly for charts and graphics. He described the process of having AI suggest organizational changes for his research as feeling like having a colleague sitting next to him. He believes the tool is particularly useful for early-career faculty designing new courses.
Nick Canada, a graduate student teaching Digital Humanity, observed that faculty often use AI for organization and skimming large PDFs. He noted that stigma around AI use can lead professors to hide their reliance on the tools. Canada suggested that while AI can improve education, it risks becoming a crutch if not used carefully.
