The AI-Native Shift · Book Resources
Citations & Sources
Every external citation in the book, with the caveat each one is held to. Six have landed — verified and carried in prose with the caveat honored. Three are still open, and named here so nothing ships unverified.
Landed — the six
G8Chapter 1 — the Kurzweil concession
AGI = matching the best person in each field, across "the thousands of fields"; his date 2029.
Caveat honoredNever "5,000 domains"; one placement; concessive foil only (Moonshots EP #223, 2026-01-20).
G4Chapter 10 — the moat
“the real differentiator is not the data or even the models, but the ‘tacit knowledge’ embedded in the judgment of their people”.
Caveat honoredCited as CMR-published practitioner analysis (Accenture’s Tung & Roussiere), never academic research; their deliberately-designed-in hedge kept in prose.
G5Chapter 14 — the oversight illusion
Automation bias / rubber-stamping catalogued systematically.
Caveat honoredEntropy (MDPI) 2026 systematic review — one of the few entries citable as genuine peer-reviewed research.
G6Chapter 14 — the governor
The EU AI Act puts enforceable human-oversight obligations on high-risk AI systems as of August 2026.
Caveat honoredThe source never names "Article 14" — the prose cites the high-risk bucket, not the article number.
G1Chapter 15 — the inversion
95% of GenAI pilots, no P&L impact; the 5% built the operating model.
Caveat honoredThe ~70%-stall sub-claim is never printed; the what-separates-them read is stated as the book’s answer, not MIT’s.
G3Chapter 16 — the governing role
CAIO-driven centralized / hub-and-spoke models: up to 36% higher ROI.
Caveat honored"Up to" as a maximum; 600+ CAIOs (never "2,300 orgs"); never blended with any adoption-growth statistic.
Still open — three markers
Chapter 4 — the chapter thesis
The HBR annual study — "the ranking reflects appetite, not readiness," the #6 ranking, not-in-top-100 the prior year.
StateHighest priority — no resolution entry yet; register as G9 and verify print-grade first among the remaining items. Fallback: reframe on the first-party 2026-06-10 transcript, which carries the line.
Chapter 5 — the L1 waypoint
The G2 designated landing site — verified Gartner claims (63% data-management gap; 60% of AI projects abandoned through 2026 without AI-ready data).
StateVerified and staged in the marker; land-now vs. keep-as-designated-site is an open decision. Old 60%/40%/2.5x figures retired — never print.
Chapter 11 — the digital-native myth
The Prensky framing and its two decades of refutation.
StateCurrently rendered without attribution; verify sourcing before naming the term in print.
One further citation (the N-layer architecture consensus, G7) is verified in the outline but has no landing site in Book 3 — a deliberate no-land, or one to assign, as a production item.