Summary
The episode works through four questions that students commonly find difficult in the final lesson of HSCI 488: whether data with names removed can enter an approved AI assistant, whether an approval covers features a vendor adds later, how much the energy cost of AI should change practice, and how to judge whether a new release is better for one's own work. Using the fictional Cedar Valley Health Authority, it separates the privacy question raised by small counts from the governance question raised by the First Nations principles of OCAP®, and it applies the B.C. framework's requirements for embedded AI features and updates. A harder worked example estimates the energy used by a year of generated campaign images, corrects a unit error, and compares figures measured with the same method before weighing cost against benefit. A personal test set example shows why equal scores can hide different failures. The hosts close with a discussion of deskilling, agree on a practical rule for keeping core skills in use, and encourage listeners to keep learning about and experimenting with AI tools.