Using Pensive AI for Chemistry Learning and Assessment
Julia Chamberlain
Description
Grading tools that use Artificial Intelligence (AI) to support or automate the evaluation of student work have arrived. In Spring 2026 I used a new tool called Pensive, which provided AI support for student work on in-class problem solving, an AI tutor for students to consult outside of class, and AI grading of paper exams administered in person. While similar in many ways to the beloved and widely adopted Gradescope tool, Pensive adds AI grading to our assessment toolbox. This talk shares my experiences using Pensive AI in a 100-student general chemistry course for grading exams and in-class problem solving, including instructor, TA, and student perspectives. I will discuss time investment, workflow, AI accuracy, and the new questions about assessment design and grader expertise that emerged from this pilot. Finally, I will answer the question, “Is it still authentic teaching if artificial intelligence grades my exams?”
Submit a question for Dr. Chamberlain to answer during the live discussion panels at SITT.
About the Presenter
Dr. Julia Chamberlain is an Associate Professor of Teaching in the Chemistry Department. An active participant in the Scholarship of Teaching and Learning, her research considers why learning chemistry can be hard, and how to make it easier to understand. Her projects at UC Davis focus on understanding students’ study practices in General Chemistry, the contributions of undergraduate Learning Assistants as facilitation partners in discussion sections, and the role of technology in physical and online learning spaces. When not at the computer, she enjoys gardening, hiking, and spending time in the zone of proximal development.
