---
id: "action-deconstruct-jobs"
type: "action-item"
source_timestamps: ["§ Where and When to Use Generative AI", "§ Applying the Framework"]
tags: ["task-analysis", "implementation"]
related: ["framework-gen-ai-deployment", "concept-cost-of-errors", "concept-knowledge-type-tacit-vs-explicit"]
action: "Break down broad jobs into component tasks to evaluate Gen AI suitability based on error cost and knowledge type."
outcome: "Clear identification of which specific workflow steps can be safely and effectively automated or augmented by AI."
speakers: ["Bharat N. Anand", "Andy Wu"]
source_url: "https://hbr.org/2025/11/the-gen-ai-playbook-for-organizations"
source_title: "The Gen AI Playbook for Organizations"
sources: ["agentic"]
sourceVaultSlug: "hbr-seg-agentic"
originDay: 6
articleStem: "hbr-cl-87-genai-playbook-orgs"
sourceUrl: "https://hbr.org/2025/11/the-gen-ai-playbook-for-organizations"
sourceTitle: "The Gen AI Playbook for Organizations"
---
# Deconstruct jobs into component tasks

**Action.** Break down broad jobs into component tasks to evaluate gen AI suitability based on error cost and knowledge type.

**Why.** Don't ask whether gen AI can perform an entire job (e.g., 'marketer' or 'lawyer') as well as a human. Instead, break each role into specific component tasks (e.g., 'generating taglines' or 'drafting boilerplate clauses'), then evaluate each task against the [[concept-cost-of-errors|cost of errors]] and the [[concept-knowledge-type-tacit-vs-explicit|type of knowledge]] required. This is **step 1 of the [[framework-gen-ai-deployment|deployment framework]]** and the prerequisite for routing any task to its zone.

**Outcome.** Clear identification of which specific workflow steps can be safely and effectively automated or augmented by AI — including tasks in high-stakes jobs (see [[concept-human-first-zone|Human-First Zone]]) that are individually safe for AI.