---
type: "synthesis"
articles: ["a032", "a033", "a050", "a051", "a086"]
tags: ["learning-and-development", "capability", "measurement"]
id: "cross-completion-not-capability"
sources: ["reskilling"]
sourceVaultSlug: "hbr-seg-reskilling"
originDay: 10
articleStem: "hbr-seg-reskilling"
sourceUrl: "(unified vault: 13 sources)"
sourceTitle: "HBR — People Ⅲ-B · Reskilling / L&D / talent / restructuring"
---
Across the L&D-heavy articles a single insight recurs: **an organization can log training, certificates, and tool access and still hold zero new capability.**

A033 names the illusion directly — the [[concept-capability-mirage]] driven by the [[concept-forgetting-curve]]: "[[quote-textbooks-surgery|using slide decks to master AI is like using textbooks to master surgery]]" ([[contrarian-training-vs-capability]]). A051's [[concept-capability-debt-d10]] is its longitudinal twin — the invisible liability that accrues while capability quietly erodes. A032 makes the same point about tools: [[contrarian-fluency-is-not-enough]] — fluency is necessary but nowhere near sufficient. A086 quantifies the enterprise gap ([[claim-ai-competence-gap]]) and argues generic "Gen AI 101" fails ([[action-shift-ai-training-focus]]). A050 shows *why* even good intentions fail operationally: learning time is instantly swallowed by delivery, so [[claim-infrastructure-scales-adoption]] (a [[concept-centralized-internal-hub]]) — not tool access — is the differentiator.

The corpus's shared metric complaint: legacy L&D counts *inputs* ([[prereq-traditional-ld-metrics]], completion rates, cost-per-learner) when the real question is *acquired judgment*. A033's ROI framing ([[claim-ai-roi-failure]]) closes the loop: unused capability is why AI investments miss returns. Remedies diverge in [[cross-reinventing-ld]].