#  Assessment in The Age of AI 

 



 ##  

  expand\_more  

 
  

 

As faculty grow concerned about inappropriate AI use in asynchronous, take-home assignments, the Bok Center works with them to design and support other modes of assessment. These typically include an in-person component or work completed without AI.





 

##  Problem sets paired with in-class follow-ups (COMPSCI 20) 

COMPSCI 20: Discrete Mathematics kept its take-home problem sets and paired each one with a short, paper-based follow-up in class. Shortly after submitting a problem set, students answer two questions on paper, with no laptops: one asks them to explain the key conceptual move in their own submitted work, and one asks them to extend the result or transfer the technique to a new case. A student who did the work can answer; a student who submitted work they do not understand generally cannot. Rebecca Nesson, Lecturer on Computer Science and the course head, paired the follow-ups with a broader redesign: fewer problem sets, more in-class quizzes, daily group assignments, and AI practice bots students use to prepare for quizzes. Nesson also uses AI on the instructor side, to read the handwritten follow-ups and group them for faster review. "It was really nice to be able to do something that was so simple as this paper follow-up in class," Nesson said. "Anybody could do this."

The pattern can fit any course with staged take-home work: problem sets, response papers, coding assignments. To adapt it, contact the Bok Center.





 

##  Rethinking assessments for a 50-person course (LING 10) 

Professor Kathryn Davidson worked with the Bok Center on assessment design for LING 10 / HUM 16: Language, a medium to large-enrollment course in the humanities. The conversation produced assignment formats that privilege depth and process over output a chatbot could supply: incorporating examples from on-campus locations such as museums; distant-reading assignments in which students use AI to synthesize findings across a large set of linguistic studies and then evaluate the synthesis; and a puzzle hackathon in which students work in groups on puzzles they create for each other, with prizes awarded. None of the three formats depends on linguistics.

To talk through the assessments in your own course, request a consultation with the Bok Center.





 

##  Assignment design workshop for a department (Government) 

Faculty and teaching fellows in the Government Department took a Bok Center workshop on designing and grading assignments in the age of generative AI. Participants began with a norming activity using current AI tools to build shared intuitions about what the tools can and cannot do. They then worked through real essay prompts and problem sets, adapting them for clarity, engagement, and alignment with learning goals while reducing the room for uncritical AI use. The Honor Council closed the session with recent trends in AI-related cases and time for discussion of policy.

To bring a version of this workshop to your department, contact the Bok Center.





 

##  Redesigning assignments for the age of AI (ENGLISH 189VG) 

Professor Vidyan Ravinthiran’s ENGLISH 189VG: Video Game Storytelling asks students to think critically about the form of video games and about their personal connections to the games they play. The Learning Lab worked with Professor Ravinthiran to redesign the course’s assignments so that students keep thinking and feeling for themselves. One assignment became a map of ten minutes of gameplay, drawn by hand. Another became a personal essay about the student’s favorite game, bringing together the student’s emotional and analytic investments in the experience of play. The redesign aims each assignment at material the student alone holds: the student’s own hand, and the student’s own experience of a game they know well. Neither format is specific to the study of games.

To talk through redesigning your assignments for the age of AI, contact the Bok Center.