Skip to main content

You're a shape,
not a score.

AI skill is a vector across six domains. The shape of that vector — where you're strong, where you lag — is the whole story a single number hides.

Find your shape — free →
FoundationsEthicsJudgmentInteractionEvaluationOrchestration

The certification contour — high and even, with Orchestration introduced.

Five depths, six domains

Every domain runs the same five-rung ladder — so a “3” means the same depth everywhere. What changes is what that depth looks like in each domain.

RUNG 1
Latent
RUNG 2
Emerging
RUNG 3
Practicing
RUNG 4
Proficient
RUNG 5
Expert
L1Foundations
L2Ethics
L3Judgment
F1Interaction
F2Evaluation
F3Orchestration

Hover any cell to read what that depth looks like in each domain.

Which of these is you?

Archetypes are generated from the shape of a profile, not hand-picked. Pick one and watch it move against the contour (dashed), colored by its capability–safety balance.

Flat — the generalist

Tilt — tier-led

Gap — guardrails vs throughput

Peak — the specialist

Valley — the blind spot

Dyad — the role pair

The Practitioner

Flat — the generalist · established altitude

FoundationsEthicsJudgmentInteractionEvaluationOrchestration
Literacy: InformedFluency: CapableGap 0 · matched

Gap = how far doing-skills (interaction, orchestration) run ahead of judgment (ethics, evaluation, judgment). Above zero, capability is outrunning safety.

Even generalist — solid and balanced across all six domains, with room to deepen everywhere

24 base archetypes, past 35 with altitude. Three-plus-feature profiles are described compositionally — never given a bespoke name.

When capability outruns judgment

In plain terms: are your AI doing-skills running out ahead of your judgment? We measure it as throughput (what you can produce) minus the guardrails (ethics, judgment, evaluation) — and the same skill total can tip three very different ways.

The Operator

Capability ahead of safetygap +2.83

High capability over low understanding. Capability races ahead — the flagship danger.

The Certified Contour

Matchedgap -0.5

Capability is high but never allowed to outrun judgment. The balance the credential certifies.

The Skeptic

Safety ahead of capabilitygap -3.17

Maxed judgment and ethics, deliberately light hands. Safe — but value is left on the table.

Every domain has a failure mode

When a domain lags, a specific, named risk surfaces. The growth spectrum and our six-risk framework are two readings of the same axis.

L1 · Foundationsliteracy
When it lags ↓
The Wizard Problem

Oscillating between magical over-trust and reflexive dismissal.

Surfaces in: The Operator, The Recipe-Follower
L2 · Ethicsliteracy
When it lags ↓
Moral Offloading

"The AI recommended it" — so whose fault is it?

Surfaces in: The Loose Cannon, The Watchdog
L3 · Judgmentliteracy
When it lags ↓
The Deference Reflex

The judgment skill of working without AI atrophies through under-exercise.

Surfaces in: The Credulous Expert
F1 · Interactionfluency
When it lags ↓
The Mirror Trap

An AI trained to agree is not a collaborator — it's a mirror.

Surfaces in: The Whisperer
F2 · Evaluationfluency
When it lags ↓
The Fluency Trap

Polished prose feels true. AI is always polished.

Surfaces in: The Unchecked
F3 · Orchestrationfluency
When it lags ↓
Brain Fry

Managing AI tools is cognitive work. That work has a cost.

Surfaces in: The Conductor, The Builder

These dangerous profiles explain why the contour stops where it does — never labels pointed at a person.

A certification that maps to these six domains

The Applied AI Certification, offered by the nonprofit Center for Practical AI, certifies confident, capable independence across this exact framework — a contour, not a summit.

Learn about it at cp-ai.org →

See your own shape

The free assessment maps your six-domain profile, names your archetype, and shows exactly where you sit against the contour.

Take the free assessment