I design each course from what students should still be able to do after the semester: build an analytical frame, connect ideas across fields, judge evidence under uncertainty, and state recommendations with the tradeoffs named. Generative AI is part of that setting, so adaptive learning matters as much as systems thinking.
In Environmental System Science, students work through feedbacks, energy–material flows, and system-dynamics models. In Cities at a Crossroads, they diagnose real urban problems (energy, heat, water, land) and propose evidence-based policy, including through fieldwork. Students may use generative AI as a technical assistant; they must check outputs against observation, and they perform live and unassisted when the grade is for judgment. The aim is that they leave better at framing questions, choosing methods on purpose, and acting on the socio-environmental systems they will inhabit.