Calm Leadership in a Loud AI Moment

A 26-Day Experiment in Lean Systems Thinking

On January 23, 2026, I started an experiment.

Not because I’m fascinated with AI.
Not because I felt behind.
Not because I wanted to chase the latest tool.

I started because I feel responsible.

For 30 years, I’ve led construction teams through Lean transformations, cultural resistance, and the very real pressure of deadlines and budgets. I’ve seen tools come and go. I’ve seen fads promise salvation.

AI feels different.

It’s louder.
Faster.
More urgent.

And most leaders I talk to are feeling one of two things:

Overwhelmed.
Or rushed.

Neither leads to disciplined decisions.

So I asked a different question:

What happens if you treat AI like a production system instead of a novelty?

The Experiment

For 26 days, I didn’t “play with AI.”

I operationalized it.

We built:

• Daily memory logs to preserve context and learning
• A defined workflow instead of random prompting
• Role clarity between human and AI
• A Trello board to track work
• A simple velocity metric to measure output
• Clear Definitions of Done across every role

In the early days, writing a solid first draft took about 45 minutes — outline, think, write, revise.

By Day 26, that dropped to about 8 minutes.

Not because the thinking got shallower.

Because the system got clearer.

Time from idea to publish used to take 3–4 hours.

By the end of the experiment, it averaged 30–45 minutes.

Iteration cycles dropped from 3–5 rounds down to 1–2.

That’s not speed for the sake of speed.

That’s friction removed through structure.

What Changed Under the Surface

Daily output in the first five days averaged 9 tasks per day.

By Days 21–26, that averaged 26.7 tasks per day — a 196% increase.

Velocity nearly tripled.

Rework dropped dramatically once every role had a clear Definition of Done — from 0 formalized workflows to 9 out of 9 standardized.

And one number stood out:

By Day 29, the system produced more output in a single day than the first five days combined — while requiring less direct intervention from me.

The constraint moved.

Early on, I was doing most of the work.

Later, I was reviewing and guiding.

That’s not automation.

That’s leadership shifting from execution to orchestration.

AI Is Not a Technology Problem

It’s a leadership discipline problem.

Technology accelerates whatever system it enters.

If your thinking is unclear, AI produces faster confusion.
If your standards are vague, AI amplifies mediocrity.
If your culture avoids accountability, AI scales avoidance.

But if your leadership habits are strong — clarity, reflection, measurement, ownership — AI becomes leverage.

That’s the mirror.

The AI Leadership Mirror

After 26 days, here’s what became obvious:

1. AI mirrors your clarity.
   If you can’t define what “good” looks like, neither can it.
2. AI scales your habits.
   Disciplined input produces disciplined output. Sloppy thinking produces noise.
3. AI exposes the discipline gap.
   Where systems are weak, the weakness becomes visible faster.
4. AI amplifies culture.
   In learning cultures, it accelerates insight.
   In reactive cultures, it accelerates chaos.

This wasn’t about prompts.

It was about standards.

What Actually Changed (The Evidence)

In 26 days:

• Time to first draft dropped 82% (45 min → 8 min).
• Idea to publish dropped 81% (3–4 hours → 30–45 minutes).
• Iteration cycles reduced ~60% (3–5 rounds → 1–2).
• Daily output increased 196% (9 tasks → 26.7 tasks).
• Velocity nearly tripled.
• Rework dropped ~70% once linting, workflows, and role clarity stabilized.
• Definitions of Done moved from 0 formalized workflows to 100% adoption across 9 roles.

The surprising shift wasn’t volume.

It was maturity.

The system stabilized.
Friction decreased.
Decision turnaround dropped from minutes-to-hours to seconds-to-minutes.

The work didn’t feel frantic.

It felt structured.

The Noise Problem

Right now, the loudest voices in AI are talking about what the tools can do.

Few are talking about what leaders must do.

Adopt this.
Automate that.
Move fast.
Don’t get left behind.

I’ve seen this pattern before.

When Lean first entered construction, many treated it as a tool instead of a system. They installed boards but didn’t change behaviors. They scheduled meetings but didn’t change promises.

We risk doing the same thing with AI.

Deploying technology without raising leadership discipline creates digital theater.

And theater doesn’t produce flow.

A Different Way to Experiment

If you’re leading right now and feeling pressure to “do something” with AI, here’s my suggestion:

Don’t roll it out company-wide.
Don’t chase every tool.
Don’t panic.

Run one disciplined 30-day experiment.

1. Pick one real workflow.
2. Define clearly what “good” looks like.
3. Assign ownership.
4. Measure learning weekly.
5. Reflect publicly.

Not speed.
Not hype.
Discipline.

Leadership isn’t about reacting to noise.

It’s about creating clarity inside it.

Calm Leadership Wins

If you feel behind, you’re not.

If you feel uncertain, you’re not alone.

This moment doesn’t require panic.

It requires steadiness.

Technology will continue accelerating.

The question is whether leadership maturity accelerates with it.

Because in moments of technological acceleration,
leadership discipline becomes the competitive advantage.

And calm leadership — especially now — is not passive.

It is deliberate.