From Awareness to Agency
Where is the earliest weak transition between seeing an opportunity and sustaining intentional change?
A business can know that AI matters and still be unable to change how work happens.
Decision: Choose the next capability experiment before choosing a larger programme or another tool.
The Capability Agency Ladder is a Dreamineering synthesis: diagnose the earliest weak transition across Awareness, Imagination, Integration, Integrity, and Agency. It is a practical navigation model, not a universal scientific maturity scale and not a rank for people or organizations.
Model
The same sequence in semantic text:
| Move | Capability | Transition test |
|---|---|---|
| Awareness | Establish what is true about the journey, its current work, evidence, constraints, and affected people. | Can the team distinguish observation from assumption? |
| Imagination | Rehearse credible alternatives before committing resources. | Can the team describe more than one plausible way the work could improve? |
| Integration | Connect the chosen path into coherent roles, information, decisions, tools, and workflows. | Can the team show how the alternative would operate end to end? |
| Integrity | Test alignment between commitments, behaviour, and observed results. | Do actual decisions and outcomes honour the stated value and guardrails? |
| Agency | Initiate, observe, adapt, and sustain intentional change. | Can the team run the loop, learn from evidence, and choose the next move without waiting to be pushed? |
The moves are sequential for diagnosis, not necessarily for execution. Real work loops backward when new evidence changes what is known or which future remains credible.
Diagnose transitions
Do not assign a maturity score. Name one consequential journey, then inspect the stages in order. Stop at the first stage whose evidence is absent, contradictory, or too weak for the next transition.
- Weak Awareness makes later plans sophisticated guesses.
- Strong awareness with weak Imagination traps the team inside the current process.
- Strong alternatives with weak Integration produce pilots that never become coherent work.
- Integrated work with weak Integrity says one thing while incentives, behaviour, or outcomes say another.
- Aligned work with weak Agency depends on exceptional prompting, sponsorship, or rescue instead of a repeatable learning loop.
The earliest weak stage is the constraint. Later weakness may be real, but investing there first usually adds activity without repairing the missing transition.
Worked Business Example
A service-company leader begins with a vague ambition: “We should use AI across customer operations.”
| Stage | What the leader finds |
|---|---|
| Awareness | Enquiry triage delays useful first responses; sensitive promises already require accountable human judgment. |
| Imagination | The team can describe assisted triage, agent-prepared responses, and unchanged manual work as credible alternatives. |
| Integration | No option yet names the trigger, source context, decision owner, escalation rule, or evidence return. This is the earliest weak stage. |
| Integrity | Not diagnosed yet; the chosen flow does not exist to test against commitments. |
| Agency | Not diagnosed yet; there is no operating loop to sustain. |
The leader does not buy an “AI transformation.” The next experiment is narrower: map one product line’s enquiry flow from detection through accountable response, assign Human, Agent, and System responsibilities, and test it beside the current process for two weeks. Success means the team can run the mapped flow, explain exceptions, and decide whether to continue without outside rescue.
The AI-Native Transformation Journey provides the deeper method for turning that boundary into a decision-grade experiment.
Reachable support
This diagnostic uses the Education Principles distinction between independent capability and supported capability.
- Independent state: the leader sees an opportunity but cannot identify the capability constraint.
- Proximal move: locate the earliest weak stage in one named journey.
- Knowledgeable support: use this page, the canonical education teaching, a qualified person, or a bounded instrument that can support the diagnosed gap. Support does not receive authority beyond that gap.
- Contingent scaffold: use only the prompts for the weak stage; do not impose the whole model when one question is enough.
- Fading: remove prompts, facilitation, or instrumentation when the leader can make and justify the decision unaided.
- Transfer: apply the model independently to a second business journey with materially different work or stakeholders.
Stage prompts
| Weak stage | Proximal prompt | Useful bounded support |
|---|---|---|
| Awareness | What do we directly observe, what are we inferring, and what evidence would change our view? | A baseline, observation log, or domain expert who can verify the current flow. |
| Imagination | What are three credible alternatives, including keeping the current approach? | Scenario prompts or a facilitator who challenges hidden assumptions. |
| Integration | What must detect, decide, act, prove, and learn—and who or what owns each move? | A flow map or the transformation-journey method. |
| Integrity | Where do stated commitments, actual behaviour, incentives, and results diverge? | A guardrail review or accountable stakeholder who can surface consequences. |
| Agency | Which bounded loop can we initiate now, what will we observe, and when will we adapt or stop? | A dated experiment record with an owner, review point, and stop condition. |
On-Page Diagnostic
Complete this in one sitting with the people who own the journey and bear its effects. Evidence can be brief, but it must be inspectable.
CAPABILITY AGENCY DIAGNOSTIC
Named journey:
Intended change and beneficiary:
Protected constraints:
Evidence by stage
1. Awareness — what is true now, and how do we know?
2. Imagination — which credible alternatives have we rehearsed?
3. Integration — how does the chosen path connect roles, information, decisions, and tools?
4. Integrity — where do commitments, behaviour, and results align or diverge?
5. Agency — what intentional loop can we initiate, observe, adapt, and sustain?
Earliest weak stage:
Why this is the first unsupported transition:
One proximal move:
Bounded MKO or scaffold:
What authority remains with the leader and affected people:
Independent success test:
Fade condition:
Transfer journey:
Review date:
Decision on review: continue | adapt | redirect | stop
The output is a capability experiment, not a transformation verdict. Use Education Performance to distinguish assisted output from independent performance and transfer.
Failure modes
- Tool-first action: buying or deploying software before locating the weak transition.
- Person scoring: turning the ladder into a rank for people, teams, or organisations.
- Stage skipping: funding later activity while earlier evidence remains absent.
- Permanent support: keeping prompts or facilitation after independent judgment is possible.
- Assisted-output proof: claiming capability from work the learner cannot repeat or transfer unaided.
Limits and Evidence
This five-stage sequence is defined by Dreamineering. Integration, Integrity, and Agency are model elements here; their presence in the sequence is not established by a universal scientific law.
Decety and Grèzes review evidence that mental simulation contributes to imagining one's own and other people's behaviour. That narrower finding supports the role of simulation within Imagination; it does not prove the complete ladder or a fixed progression (primary article).
Use the model as a hypothesis when stages overlap, power is contested, consequences are irreversible, or evidence is costly. A qualified person may be necessary where legal, safety, clinical, cultural, or domain judgment exceeds the page's support. Value Integrity deepens the alignment test; Orchestration helps direct AI-supported work without surrendering judgment.
Changes my mind: repeated evidence that repairing the named earliest transition does not improve independent action or transfer should revise the sequence or replace it for that context.
Retrieval
Retrieve this page when an AI ambition is clear but the next organizational capability is not, when a pilot stalls between possibility and operating work, or when support may be creating dependency rather than agency.
Context
- depends-on Capabilities — capability investment serves a beneficiary and never ranks human worth.
- depends-on Education Principles — choose a reachable supported move, then fade support.
- pairs-with Value Integrity — inspect alignment between values, conduct, and consequences.
- pairs-with Orchestration — direct AI assistance while retaining human judgment and accountability.
- applies-to AI-Native Transformation Journey — turn the diagnosed capability constraint into one bounded operating experiment.
- proved-by Education Performance — test independence, fading, and transfer rather than assisted output alone.
- depends-on Purpose — keep the diagnosed capability in service of a worthwhile beneficiary and protected values.
Questions
- Which later-stage activity are you funding while an earlier transition remains unsupported?
- What is the smallest support that makes the next move reachable without creating dependency?
- What second journey would prove that the leader can use the model independently?
Next question: What evidence will show that support can now fade?