AI-Native
Built from scratch with AI in every stage. A blank canvas: the rare luxury of having no legacy to absorb.
You don't have an AI problem.
You have a structural bottleneck.
Most firms are doing AI. Licenses bought, pilots launched, decks printed. Nothing structural moves. This book reframes AI from a tooling conversation into an organizational design diagnosis.
You’re in.
Chapter 1 is yours below: the reframe you can carry into a boardroom. We’ll email you once more, the day the book goes on sale in late 2026.
Download Chapter 1Where the book is
- Completed: Written
- In progress: In review
- Upcoming: Production
- Upcoming: On sale
Now: First six chapters rewritten from early-reader notes. Proofreaders are reading along; a full manuscript revision comes next.
Why AI stalls
The bottleneck is not in the machine. It is in the gap between what AI can produce and what your organization can absorb.
That gap has a name. Design debt. It lives in the org chart, the incentive plan, the governance model, and the career ladder. It is where billions in AI investment go to die. Once AI enters the system, it stops hiding.
80%+
Of AI projects fail, twice the rate of non-AI technology projects
RAND, 2024
95%
Of companies see little to no measurable P&L impact from generative AI
MIT NANDA, 2025
71% → 21%
Use generative AI regularly, but only 21% have redesigned a single workflow
McKinsey, 2025
5 symptoms of design debt
- Decision latency stays high despite the tooling.
- Incentives stay misaligned despite the dashboards.
- Governance stays theatrical despite the compliance automation.
- Power structures stay intact despite the transformation.
- The organization was never designed for the feedback AI provides.
What designing for AI looks like
- Decision rights redrawn for the speed AI enables.
- Incentives that reward AI-aligned behavior.
- Governance that is traceable, not theatrical.
- One learning loop, from signal to action in hours.
- Structure that amplifies AI instead of walling it off.
Where this book stands
Three ways to meet AI. This book takes the hardest one.
An AI-equipped organization deploys AI tools. An AI Augmented Organization redesigns how decisions flow. Harvard named the first two positions and stopped. This book names the third, where almost every real enterprise lives.
AI-First
AI made the core capability that enhances operations. Powerful, but it still assumes you can put the technology at the center.
AI-Augmented
Human capability amplified where judgment and expertise still matter, with workflows designed so humans and AI deliberately share the thinking. Where most real enterprises live, and what this book is about.
Most of the world’s value is created by organizations that already exist. They don’t need a blank canvas. They need a redesign.
And augmentation is none of the usual theatre:
- A Chief AI Officer without redesigned governance is a job title.
- A Center of Excellence without changed decision rights is a knowledge silo.
- An AI roadmap without restructured delivery cycles is a slide deck.
Three properties, resting on one information architecture.
Not three labels. Three structural tests. An organization that merely uses AI fails all of them; an AI Augmented Organization is built to pass them.
Structural Cognition
AI is wired into how the organization thinks, not a tool people open and close. Turn it off tomorrow and decision-making actually changes.
Designed Co-Decision
An explicit decision-rights matrix: what AI recommends, what AI decides, what humans own. Unassigned authority defaults to whoever acts fastest, and AI always acts faster.
Feedback at Operating Speed
Loops that close in hours, not quarters. The elapsed time between “we learned something” and “we changed something” measured in days, not months.
One information architecture. Signals from any function can inform decisions in any other, without being walled off by the org chart. None of the three properties can run without it.
5 parts. One argument that compounds.
From the problem, to the model, to its application across the enterprise, to the human layer, to the scoreboard. ~250 pages, 13 chapters, 2 diagnostic appendices.
I
The Diagnosis
AI doesn’t fail because of technology. It fails because of design debt. Once AI enters the system, the debt stops hiding.
The Structural Bottleneck · What an AI Augmented Organization Is (and Isn’t)
II
The Core Model
The operating heartbeat and the structural preconditions: the Unified Value Loop, and the organizational design that decides whether AI amplifies or stalls.
The Unified Value Loop · Organizational Design for AI
III
The Enterprise Pillars
AI across functions. Not isolated tooling, but interconnected parts of one intelligent system: strategy, governance, finance, HR, operations.
Strategy · Governance · Financial Controlling · HR · Operations
IV
The Human Layer
Augmentation, not automation. Clear decision rights, escalation rules, and the culture that lets a system actually learn.
Human–AI Collaboration Patterns · The Cultural Dimension
V
The Scoreboard & Synthesis
How to know it’s working: decision cycle time, insight-to-action latency, cost of delay. Then the compounding edge of designing for AI instead of merely using it.
Metrics of an AI Augmented Organization · Structural Leverage
Written for the room where transformation is actually sold.
No over-explained basics. The argument assumes you already know the tools work and the programs still stall, and it starts there.
Senior partners & MDs
You sell transformation, not training, tooling, or theatre. The frameworks on offer rarely survive contact with a CEO.
Transformation practice leads
You carry a portfolio and a reputation into rooms where “we bought AI” is treated as a strategy.
Internal change leaders
You’re moving a real organization while copilots get bought and the incentives that break them stay untouched.
About the author
Jurgen De Smet
Chief Simplification Officer and keynote speaker with 25+ years in organizational design and product development. A software engineer since 1996, he helps enterprises de-scale complexity instead of adding to it.
Creator of the H³ (Harvest · Hack · Harness) system, he works with leadership teams on the structural moves that make AI compound rather than stall.
“AI capability moves too fast for plans. Only rhythm survives.”
Barco · KPN · Danske Bank · Heineken · Fluvius · Base Company · SoftNet · AccentJobs · Česká spořitelna
Get Chapter 1 free
Walk in with the instrument, not the slogan.
Leave your email. Get Chapter 1, The Structural Bottleneck, free right now. No chapter behind a second form, no drip sequence.
- The 5 symptoms of design debt: how to spot the bottleneck in any organization.
- The absorption gap, and the one ratio that measures it in your own operation.
- The Executive Conversation: what to say in the boardroom, and what to ask for.
You’re in.
Chapter 1 is yours below: the reframe you can carry into a boardroom. We’ll email you once more, the day the book goes on sale in late 2026.
Download Chapter 1Why now: the gap isn’t closing. Every model release makes AI’s output bigger and the absorption gap wider. The debt compounds: by 3, by 5, by 10. The reframe is only an edge while most of the market still calls it a tooling problem.