Week 12: Capstone studio

Work time and peer defence rehearsal for the redesign brief, and a look back at how the semester's opening claim held up

What a defence that survives looks like#

Every rehearsal defence this week takes the hardest question the room can find, not the easiest one a friend might lob to help you look good. The pattern that separates a defence that survives from one that doesn’t is almost always the same: whether the tradeoff was named upfront, in your own words, before anyone had to press for it, or only admitted once someone else found it. A design that touches the ETA model from week 2, the rating threshold from week 4, or the pay-per-delivery line item from week 6 should be able to say, specifically, which of those three it changes and which it leaves alone — a defence that claims to fix everything at once is usually a sign the tradeoffs haven’t actually been worked through.

Closing the loop#

Week 1 opened with a claim: “the algorithm decided” is sometimes an accurate description and sometimes a place where a paper trail conveniently stops. Twelve weeks of dispatch mechanics, one detailed case study, and three regulatory regimes later, that claim gets tested one more time — against your own capstone design, where you’re the one who has to say which decisions in it you made, and defend having made them. The three regimes compared in weeks 7–10 each answered a version of the same question — China’s 2021 Guiding Opinions, Australia’s employee-like worker category, and the EU’s rebuttable presumption — and your capstone defence is, in effect, a fourth answer to that same question, argued from the position of someone who built the mechanism rather than someone regulating it after the fact.

Before the session#

Bring a near-final draft of the capstone specification and defence; this is the last structured feedback before submission.