72. Project 7 — Find Something Nobody Told You to Build
The final project. Not a tutorial. Not a course assignment. Not an implementation of somebody else's algorithm. Your own question.
Specification. There is none. The constraints are the book's whole argument: it must originate from your question — found in your friction journal (Ch. 55.3), your claims matrix (55.2), your surprises file (59.6), the empty cell only you noticed; it must be honestly sized (55.9's filter — laptop-tractable, or cloud-hardware-scaled); it must be executed with Ch. 54's discipline (hypothesis first, baselines strong, statistics real); and it must produce a public artifact (62's standards) — code, a write-up, a figure that shows something nobody had shown, even if the something is small.
Candidate shapes, to prime not prescribe: a measurement nobody has published (seed-variance of routed circuits; drift in cloud calibration data analyzed longitudinally — you have free longitudinal access nobody exploits); a tool that closes a friction you hit (your Ch. 46.8 benchmark harness, generalized; a Bell-test statistics playground); a negative result worth knowing (does error-mitigation technique X actually help workload class Y? the honest sweep); a small theory instantiation (measuring barren-plateau onset constants for an ansatz family); an educational artifact of unusual quality (the interactive explainer this book couldn't be in print). The test of a good Project 7: you can't find it assigned anywhere, and you can explain in three sentences why you cared.
Acceptance criteria. One: you asked it. Two: you answered it with evidence. Three: it's public. Whether it's "significant" is not yours to judge and is the wrong metric — the apprenticeship ends not when you produce significance but when you can originate: notice a question, size it, pursue it, and report honestly what you found. That capability, compounded over years, is the entire difference between someone who knows the field and someone in it.
What it proves. Chapter 59.8's checklist, final line: capable of independent research. The next part is logistics.