a framework for Aligned AI

The Mission Matrix

How audacious leaders root AI in their organization’s mission — and audition AI for one of four roles, by job to be done. Stop asking “How should we use AI?” Start asking “What can we finally become?”

Explore the framework →
High meaningMeaningLow meaning
Low unique expertiseUnique expertiseHigh unique expertise
Your growth edgeHigh meaning · Low unique expertiseAI as capabilities builderCoach
Your core craftHigh meaning · High unique expertiseAI as forcefieldBodyguard
Routine tasksLow meaning · Low unique expertiseAI as hand-offApprentice
Skilled but drainingLow meaning · High unique expertiseAI as focused aidePartner
Unlocking organizational ambition

Four roles. One clear map.

Plot any activity across two axes — the depth of meaning required in the work, and how uniquely your expertise is needed to do it. The quadrant tells you which AI role belongs there.

← Low unique expertiseUnique ExpertiseHigh unique expertise →
High meaning ↑Depth of Meaning↓ Low meaning
Quadrant IIEnable doing the previously impossible

Coach

Challenge & push

What could I not do before but have always wanted to?

Building your own prototypePublic speaking practiceNew strategic capabilities
AI as capabilities builderCoach
Quadrant IFilter noise + create a forcefield

Bodyguard

Protect & make space

What is deeply meaningful and I find purpose in being present and in the friction?

Client debriefsWeekly team lunchesHigh-stakes decisions
AI as forcefieldBodyguard
Quadrant IIIAutomate with transparency

Apprentice

Automate, delegate or eliminate

What routine work could I fully hand off if I trusted the output?

Weekly metrics pullsSprint retro summariesMeeting notes
AI as hand-offApprentice
Quadrant IVAssist and learn

Partner

Lighten the load

What do I have unique expertise in, but not meaning — and would be happy to teach and hand off?

Performance review draftsResearch synthesisDraft generation
AI as focused aidePartner
Going deeper

Quadrant details

What each role demands of your organization — the levers, artefacts, tools, risks, and skills to develop.

IBodyguardProtect & make space
IICoachChallenge & push
IIIApprenticeAutomate, delegate or eliminate
IVPartnerPartner & lighten
Key levers
Values and human capabilities
Responsible but aggressive experimentation on the frontier
Accurate workflow mapping
Effective context capture and relevant use
📄Artefacts
Culture, values, apprenticeship, rituals
Experimental play with safe sandboxes
Wikis and workflows
Specialized skills and memory combing
🔧Example tools
Agents designed to filter
Claude Code/Codex, ChatGPT/Claude/Gemini
Simple agents, automations
Skills, Projects, AI embedded in existing tools (Canva, Notion etc)
⚠️Risk
Not protecting this space aggressively enough
Not experimenting boldly or quickly enough
Not fully handing off
Not transferring context effectively
Get good at
How to build effective filters the agent will respect and follow
How to manage memory effectively so it understands your goals and how you like to be coached
How to see all of the work and describe it at the right level of detail
Seeing the boundaries of the point work you want to hand-off
How to implement it

Three steps to putting the matrix to work

1

Map the work

Start with your organization's mission, expertise, and the jobs to be done. Plot them across the matrix — your map stays stable even as the AI tools keep changing.

Take the assessment →
2

Audition AI for the role

For each job to be done, audition AI candidates for the right role: Coach, Bodyguard, Apprentice, or Partner. Pick the model that fits the moment — don't use one AI for everything.

3

Lead with purpose

Design the rituals, norms, and guardrails that enable your organization to operate sustainably within the framework.