AI Alignment: learning that argues back

A course modelled on the Oxford tutorial, the Harvard case, the MIT problem set, the Yale seminar and the Princeton precept — every lesson makes you commit a decision before it shows you the expert's, and defends the honest limits of what we know.