🤖 Human-AI Coordination
Best practices for coordinating work between humans and AI agents.
Roles and boundaries
| Actor | Best for | Not suited for |
|---|---|---|
| Human | Strategy, judgment, creativity, relationships | Repetitive tasks, data processing |
| AI Agent | Research, drafting, analysis, automation | Final decisions, sensitive communication |
Coordination patterns
Delegation model
- Human defines the task clearly (scope, format, constraints)
- AI executes and produces a draft
- Human reviews, validates, and refines
- AI incorporates feedback and finalizes
Collaborative model
- Human and AI brainstorm together
- AI structures and organizes the output
- Human adds judgment and context
- Iterate until quality bar is met
Effective AI tasking
- Be specific about expected output format
- Provide context — background, audience, constraints
- Set quality criteria — what does "good" look like?
- Break complex requests into sequential steps
- Review AI output critically — don't auto-accept
Anti-patterns
- Giving vague instructions and expecting perfection
- Over-relying on AI without human review
- Using AI for tasks requiring empathy or ethical judgment
- Not providing feedback to improve future interactions