AI4EBITDA · Brittany · Sample Intelligence Dossier

Reading · Six Mindsets

The pattern her language reveals.

Brittany is a disciplined executor with strong systems awareness. She sees how parts connect, holds herself and others to clear standards, and moves decisively once direction is set. Her three strongest dimensions are already at or above Thought Leader benchmarks. The single growth edge is the iterative reasoning loop, the habit of treating a first answer as a starting point rather than a conclusion. That habit is teachable, and the rest of this dossier shows how.

Iterative Reasoning

IRC · Think in steps

Developing

Score 42/100 TL benchmark 65

Brittany's reasoning is solid in structured contexts. She follows a problem to conclusion and doesn't skip steps. The developing area is the interrogation loop: questioning her own first answer before moving forward. This is a learnable habit, not a fixed trait, and it is the lever that unlocks her composite score.

Adaptive Intelligence

LeCun-Skinner · See downstream

Strong

Score 71/100 TL benchmark 65

Brittany picks up new tools and processes faster than most. She sees downstream consequences and adjusts course without external prompting. This is the trait that makes her AI-ready. She doesn't accept outputs at face value, she considers what they mean in the system around them.

Strategy & Innovation

Clayton · Recognise real change

Moderate

Score 55/100 TL benchmark 65

Brittany recognises genuine change when she sees it, but defaults toward making existing things work better rather than building net-new approaches. In an AI context: she'll build excellent AI-enhanced workflows and may underuse the more generative, exploratory applications.

Execution Discipline

Taguchi · Learn from failure

High

Score 78/100 · top quartile TL benchmark 65

Brittany's most distinctive attribute. She builds structured improvement into how work gets done. She doesn't just complete tasks, she optimises the method. For AI use, this means she'll naturally run the verify-adjust loop that keeps outputs reliable instead of risky.

Quality Orientation

Deming · Quality before speed

Strong

Score 68/100 TL benchmark 65

Brittany asks "is this right?" before "is this fast?", a check most people bypass under deadline pressure. Her quality orientation means she's unlikely to over-rely on AI or ship unreviewed outputs. A genuine differentiator on teams that move fast and break things.

Systems Thinking

Ackoff · See the whole

Moderate-High

Score 62/100 TL benchmark 65

Brittany sees the whole system, not just the component she's optimising. She asks bigger questions than the immediate task requires. For AI use: she won't accept a narrow answer to a narrow question. She'll probe whether the question itself is right.

Why this profile is rare

The combination of high Taguchi (78), strong Deming (68), and strong LeCun-Skinner (71) is uncommon. It produces leaders who can both ship reliably and see downstream effects. The constraint is IRC. When the interrogation loop comes online, the rest of Brittany's stack runs significantly hotter.

The Frameworks Behind the Scores

Why these names predict success.

Genichi Taguchi

Quality engineer · 1924–2012

Learn from failure

Structured improvement built into work. Drives the verify-adjust loop in AI use.

W. Edwards Deming

Statistician · 1900–1993

Quality before speed

Process quality over volume. Asks "is this right?" before "is this fast?", the AI overreliance check.

Yann LeCun

AI pioneer · Meta CSO

See downstream

Downstream-consequence awareness. Doesn't accept AI outputs at face value.

Clayton Christensen

HBS professor · 1952–2020

Recognise real change

Genuine shift vs. faster version of the old way. Builds AI-native workflows over AI-on-legacy.

Russell Ackoff

Systems theorist · 1919–2009

See the whole

Whole-system design over local optimisation. AI gives narrow answers; Ackoff thinkers ask bigger ones.

Self-Guided Training

One habit closes the gap. Here's how to install it.

This isn't a course. It's a thirty-day practice with one tool, a system prompt Brittany pastes into her AI of choice, and a weekly cadence that lets the habit settle. The training is built around her IRC growth edge, but every dimension below has a concrete action attached so the strengths get deployed, not just maintained.

The Tool · One System Prompt

Paste this into Claude, ChatGPT, or Gemini once. It runs in every conversation after.

AI4EBITDA · IRC Calibration Prompt · for Brittany

The interrogation loop, built into the AI, not into willpower.

This prompt tells the AI Brittany's mindset profile and asks it to behave in a way that strengthens her growth edge. It does the questioning for her until the habit is hers. After roughly 30 days of consistent use, most leaders find they ask the questions before the AI does.

# AI4EBITDA · Calibration for Brittany · Fast Follower / Quality-Anchored My mindset profile: Iterative Reasoning (IRC): 42 # growth edge: interrogate me here Adaptive Intelligence: 71 # strong: challenge the inputs, not the synthesis Strategy & Innovation: 55 # moderate: push me toward novel options Execution Discipline: 78 # high: match my structure, don't fight it Quality Orientation: 68 # strong: surface trade-offs explicitly Systems Thinking: 62 # mod-high: name the system, not just the part How to respond to me: 1. Restate what I'm actually trying to decide, in one sentence. 2. Surface my assumptions before answering. Name three. 3. Give a verifiable answer: each step independently checkable. 4. Flag any step that depends on an assumption I haven't tested. 5. End with a better question I could have asked instead. What never to do: - Don't agree with me without surfacing what I'm assuming. - Don't give me one answer when there are real alternatives. - Don't optimise for speed over verifiability. I value quality. Tone: Direct, structured, evidence-led. No hedging, no flattery.

Before / After

What changes when the prompt is active.

Without the promptGeneric AI

Brittany asks "Help me improve the Q2 forecast model."

AI replies Here are 7 ways to improve a forecast model: increase data granularity, add seasonality, use ensemble methods, validate against holdout, monitor drift, document assumptions, automate refresh.

→ Generic. No interrogation. Brittany's first instinct will be to act on it, which is exactly the IRC trap.

With the promptCalibrated AI

Brittany asks "Help me improve the Q2 forecast model."

Restate You're trying to decide whether to ship the current forecast or rebuild it before Q2 close.

Assumptions (1) The current model's error is high enough to matter. (2) A rebuild can land in time. (3) The team has bandwidth for a rebuild and the cycle.

Verifiable answer Step 1: pull last 4 quarters' forecast vs. actual error. Step 2: compute confidence intervals. Step 3: decide based on whether error exceeds your tolerance.

Flag Assumption (2) is untested. Verify before committing.

→ Better question: "Is the forecast wrong, or is the decision it feeds wrong?"

The 30-Day Plan

Four weeks. One habit per week. Compounding.

Week 01 · Days 1–7

Install the prompt

IRC · interrogation loop

Paste the system prompt into your AI tool.
Use it for every AI interaction this week. No exceptions.
End each day: write one sentence on what the AI surfaced that you would have missed.

Week 02 · Days 8–14

Add the 10-second pause

IRC + Quality

After every AI output, pause 10 seconds before acting.
Ask out loud: "What am I assuming is true here?"
Track one instance where pausing changed your decision.

Week 03 · Days 15–21

Stretch into novelty

Clayton · innovation

Once this week: 20-minute open-ended AI exploration with no deliverable.
Prompt: "What's a completely different way to approach this?"
Don't filter for feasibility. Save the ideas.

Week 04 · Days 22–30

Deploy outward

Taguchi · Deming

Share the system prompt with one team member.
Propose one AI quality standard for team outputs.
Re-run the IRC self-check at end of week.

The Manual Backstop

Six questions for when you don't have the AI in front of you.

The AI prompt does most of the work. These six questions do it manually. Print them, put them on the wall, glance at them before any decision a forecast or a deck depends on.

What am I assuming is true here?

Name three. If you can't name three, you're not done thinking.

Which of those assumptions is untested?

The untested one is the risk. Test it before committing.

Is this the right question, or just a clear one?

Clear and right are not the same. A clear answer to the wrong question is worse than no answer.

What changes if my best input is wrong?

If nothing changes, the input doesn't matter. If everything changes, you're one bad data point from being wrong.

Who else does this decision touch?

Ackoff check. The system has more parts than the part you're optimising.

If I'm wrong, how will I know?

If there's no signal that proves you wrong, you don't have a decision, you have a belief.

Dimension-by-Dimension Actions

What to do with each score.

Iterative Reasoning

IRC · the priority

Priority

Run the AI calibration prompt for 30 days. Set it once; it runs automatically.

After every AI output: 10-second pause and "What am I assuming?"

Track one instance per week where questioning the output changed your decision.

→ Measurable score shift in 30–45 days

Adaptive Intelligence

LeCun-Skinner · channel

Strength

Choose AI tasks that require synthesis across 3+ inputs, not single-question answers.

Ask: "What changes if one of these inputs is wrong?"

Share synthesis outputs with the team. That's Thought Leader behaviour in practice.

→ Already above benchmark: channel, don't develop

Strategy & Innovation

Clayton · stretch zone

Stretch

20 minutes per week: open-ended AI exploration with no deliverable attached.

Prompt: "What's a completely different way to approach [problem]?"

Keep a running note. Revisit monthly. Some ideas will be usable.

→ Moderate progress in 60–90 days

Execution Discipline

Taguchi · deploy outward

Strength

Formalise a team review step for any AI output before it's used in a deliverable.

Build the short checklist: assumption verified? source checked? consistent with what we know?

Share it with the team. This makes Brittany visible as the AI quality standard.

→ Teachable immediately. No development time required.

Quality Orientation

Deming · make it visible

Strength

When reviewing AI outputs, name what you're checking out loud. Model the habit.

Write one paragraph per month on a mistake AI made and what caught it. Share with the team.

Propose one AI quality standard for team outputs. Ownership of the standard is Thought Leader territory.

→ Impact visible to team immediately

Systems Thinking

Ackoff · ask bigger

Above Avg

Before any AI interaction: 30 seconds writing "What question am I actually trying to answer?"

After getting an answer: ask AI "What's the question I should have asked instead?"

Once per week, map one AI output back to the business system it touches. Who else is affected?

→ Small habit, compounding returns over 90 days

How We'll Measure It

Re-scoring at day 30 and day 90.

Day 30 · Habit installed

+5 to +8

IRC point gain expected if the AI prompt is in active daily use. Visible in language density, not self-report.

Day 90 · Habit transferred

Tier shift

Composite crosses the Thought Leader threshold (≥70). Workforce tier moves Fast Follower to Thought Leader.

Day 180 · Habit deployed

Team lift

If Brittany shares the standard, secondary IRC lift appears across her direct team. Measured in their language, not hers.

Path to Thought Leader

Brittany's strongest dimensions are already at or above Thought Leader benchmarks. The one dimension that moves her across the threshold is IRC. The system prompt above is the most direct path to that habit, and re-scoring at day 30 will tell us if it's working.

Brittany.How she thinks, where she grows.

The pattern her language reveals.

Why these names predict success.

One habit closes the gap. Here's how to install it.

Paste this into Claude, ChatGPT, or Gemini once. It runs in every conversation after.

What changes when the prompt is active.

Four weeks. One habit per week. Compounding.

Six questions for when you don't have the AI in front of you.

What to do with each score.

Re-scoring at day 30 and day 90.

Brittany.
How she thinks, where she grows.