Teaching with AI

What has changed

Writing an essay once required synthesizing the material, and producing an argument required constructing one. Assessment traditions were built on the link between producing work and doing the thinking behind it, and that link no longer holds. Producing a finished piece, once the hardest step, is now often where student work begins.

SFBU is a young university, and its courses are built for the conditions students live in now. CEPI's job is to design courses that belong to those conditions.

The Pyramid After AI

AI produces fluent, plausible, well-structured output in almost any field. The human work begins with deciding what to do with that output. CEPI's framework for this work is the Pyramid After AI, and it complements Bloom's Taxonomy. The Pyramid sits inverted above Bloom's, and six capacities rise from the point where AI output becomes the starting point for human work. Students use them in cycles, and a student often starts to judge before finishing decoding.

A psychology student reads an AI summary of trauma research and notices that it has merged acute and chronic trauma. The distinction matters for clinical work. The summary reads as authoritative, and catching the error takes subject knowledge the tool lacks.

Course design after AI

Students orient to what earns the grade, so assessment is where a course shows its priorities. SFBU designs assessments that ask for the student's process along with the product. Many include a brief oral defense, a presentation or live problem-solving. Every major assignment states what AI does and what the student must supply. Assessment develops this in detail, and Relational Intelligence takes up the human relationships that matter more once AI is part of the work.