The Teacher Who Grades Questions

Segal recounts an unnamed teacher who, recognizing that AI tools could produce competent essays on any topic within minutes, redesigned her assessment to evaluate not the finished text but the questions a student would need to formulate before a meaningful essay could be written. Students were given a topic and an AI tool; the assignment was to produce the five questions they would ask—of the material, of themselves, of the AI—before generating the essay. The teacher reported that students' actual writing improved after the change, but the improvement was secondary. The primary effect was developing a cognitive habit—question-formulation, the identification of the boundary between knowledge and ignorance—that the old assessment culture (grading essay products) did not incentivize and that AI's commodification of answers made newly essential. The innovation is a textbook case of social invention: an individual perceiving the gap between material conditions (AI can answer) and institutional structure (assessment still rewards answering), and creating a new practice that realigns the institution's evaluative framework with