---
id: "claim-translation-difficulty"
type: "claim"
source_timestamps: ["§ How AI Might Take Jobs—and How It Probably Won't"]
tags: ["productivity", "process-engineering"]
related: ["concept-individual-vs-process-productivity", "action-redesign-business-processes"]
confidence: "high"
testable: true
speakers: ["Thomas H. Davenport", "Laks Srinivasan"]
sources: ["execution"]
sourceVaultSlug: "hbr-seg-execution"
originDay: 8
articleStem: "hbr-foci-62-layoffs-ai-potential-not-performance"
sourceUrl: "https://hbr.org/2026/01/companies-are-laying-off-workers-because-of-ais-potential-not-its-performance"
sourceTitle: "Companies Are Laying Off Workers Because of AI’s Potential—Not Its Performance"
---
# Translating Individual AI Gains into Process-Level Efficiency Is Highly Difficult

**Claim (confidence: high · testable: true):** While isolated tasks show measurable improvement (e.g., a **10–15% boost in programming performance**), organizations struggle to convert these individual gains into systemic business value.

Determining exactly how many people and what specific AI capabilities are required to execute an optimally structured, end-to-end business process requires disciplined experimentation that most companies have not undertaken (see [[quote-process-difficulty]]). Furthermore, employees report that *actual* AI-based productivity gains are much smaller than the C-suite expects — a gap that feeds anticipatory action.

This is the mechanism behind [[concept-individual-vs-process-productivity]], and it motivates [[action-redesign-business-processes]]. The methodology gap is the subject of [[question-translating-productivity]].

**Enrichment corroboration:** BCG says AI is changing jobs *faster than companies are redesigning operations* and recommends end-to-end workflow reshaping plus A/B testing. McKinsey identifies leadership, talent, operating model, and workflow redesign — not tool adoption — as the main barriers to capturing AI value. EY finds trained, role-tailored users (81+ hours annual training) realize markedly higher gains, reinforcing that value capture is organizational, not automatic.


## Related across articles
- [[claim-verification-negates-productivity]]
- [[claim-process-redesign-required]]
- [[concept-productivity-paradox]]