Training
Certification Leadership Frameworks Agentic for Business
Community Keynotes Retreat Blog Video Book A Consultation
Back to Blog

The Other 50%, the Four Offices, and the Number That Proves It Worked

We have taught these ideas one at a time. The leaders who get AI to pay off see them as a single system. Here is the system, plus the measurement layer almost nobody teaches.

We have published two of our core frameworks separately. First, the three skills that unlock the other 50% of leadership. Then the Four Offices of the Future, the org structure the AI era demands.

Taught apart, each one is useful. Taught together, they become a complete operating model for leading a company where AI is on the team. And there is a third piece we have not put in writing until now: a way to measure whether any of it is working. We call it infinite leverage. This post is the map that links the skills to the structure to the proof.

Leadership Did Not Get Harder. It Got a Second Half

Leadership used to be 100% people. You hired well, you built culture, you held people accountable. That half is not going anywhere.

The AI era adds a second half. Half of every workflow can now be directed to, delegated to, and executed by AI. Leading that half is a different discipline, and most leaders were never taught it. That is the gap behind the stat we keep citing: 95% of companies see no ROI from AI. The tools work. The leadership around them does not.

The other 50% is the skill set that lets a leader direct AI, not just manage people.

Three skills unlock it. None of them are technology skills. All three are leadership disciplines.

Three skills to direct AI The second half of leadership rests on three disciplines: workflow design, information architecture, and writing instructions for code and AI. THE OTHER 50% OF LEADERSHIP Three skills to direct AI A discipline, not a tool you buy 01 Workflow design The process that turns inputs into an outcome. Who does what, and when Some of the who are now AI agents 02 Information architecture Organize what you know so AI can use it. ABC: Always Be Cataloguing If AI can't find it, it does not exist 03 Writing instructions Translate vision into instructions a machine executes. Prompts are code You direct, AI writes

We unpacked all three in the three skills that unlock the other 50% of leadership. If you have not read that one, start there, then come back.

The non-technical person is done You don't write the code, you direct it. Prompts are instructions and AI writes. Directing AI is the new technical. THE EXCUSE JUST EXPIRED The non-technical person is done You don't write the code. You direct it. Prompts are instructions. You say what you want, and AI writes. Directing AI is the new technical

The Skills Need a Structure to Live In

A leader can master all three skills and still stall, because the company is built for a world where leadership was only about people. The Four Offices fix that.

Reorganize the business into four offices, and notice that every outcome a company cares about maps to one of them.

Underneath all four sits a central database. This is the part leaders skip, and it is the part that decides whether anything above it works. No central database, no organized information, no leverage. It is the foundation everything else rests on, and it is ABC made structural.

The central database: one identity core, four offices A shared identity core feeds four offices, Talent, Operations, Innovation and Revenue, each with its own tables. SHARED IDENTITY CORE · THE CENTRAL DATABASE people · companies · brands · legal_entities · interactions · integrations one identity spine every office reads from, a person relates to many brands Talent who powers the company team_members · departments · roles candidates · applications · interviews one_on_ones · performance · goals skills · coaching · compensation Operations how the company runs projects · tasks · milestones meetings · action items (Thoughtflow) processes · playbooks · vendors tools · expenses · ledger Innovation what the company creates products · apps · experiments content · courses · curriculum ip · frameworks · knowledge base prompt & agent library · roadmap Revenue how the company earns pipelines · deals · inquiries products · orders · subscriptions events · bookings · memberships affiliates · partners · campaigns Four offices · one identity spine · every table reads from the core

When we run this exercise live, the same blind spot shows up every time. In a room of 102 founders in Australia, only 5% chose Talent as their most important office. The office that builds the people and the AI who run everything else, treated as an afterthought.

Every outcome falls into one of four buckets. The database under them is the floor the whole building stands on.

The Four Offices of the Future Four offices, Revenue, Operations, Talent and Innovation, resting on one central database. THE OPERATING MODEL FOR THE AI ERA The Four Offices of the Future Revenue Everything that brings money in Operations Everything that delivers the work Talent Builds the team, human and AI Innovation Creates what comes next Central database the foundation everything rests on

Four Questions That Turn the Structure Into a Plan

The Four Offices are not a chart to admire. They are a template you run, one office at a time. For each office, ask the same four design questions.

  1. What is the process? The actual sequence of steps that produces the result.
  2. What outcome does it generate? The concrete thing the process exists to deliver.
  3. How many human tokens do you need to put in? The human attention, judgment, and hours the process truly requires.
  4. Where does AI fit on the team? Which steps AI executes, and where a human stays in the loop.

Run those four questions across Revenue, Operations, Talent, and Innovation, and the abstract idea of leading AI becomes a worklist. This is where the three skills get used. Workflow design answers question one. ABC feeds the central database that every answer depends on. Code is how you actually build the steps AI runs.

We laid out the full diagnostic in the Four Offices of the Future. The framework above is the short version. The post is the deep one.

The Layer Almost Nobody Teaches: Infinite Leverage

Here is the question the frameworks alone do not answer. How do you know it is working?

Most AI value is described in fog. Faster. Smarter. More efficient. None of it is measurable, which is exactly why 95% of spend shows no return.

Infinite leverage is the measurement layer. The premise is simple: a single human token can be infinitely multiplied by AI. You know it is real when concrete units of value ship into the digital world. Pull requests. Features. Published pages. A pull request is a single finished unit of change shipped into a real system. You either hit your goal or you did not. Value stops being a feeling and becomes a count.

Leverage is not how busy your team looks. It is the gap between how many people you have and how much you ship.

That gap is the whole point. Watch it widen, and you are seeing the other 50% at work.

Infinite leverage: four times the output per hour The old way took 100 hours, 20 to plan and 80 to execute. With AI it takes 25 hours, 20 to plan and 5 to execute. Four times the output per hour. INFINITE LEVERAGE · HUMAN TOKENS Same plan. A quarter of the time The old way · 100 hours Plan 20h Execute 80h With AI · 25 hours Plan 20h Execute 5h, AI does it output per hour The plan is the human token. It stays. AI collapses the execution.

A Real Company, Carried With Humility

Here is what that divergence looks like in practice, told honestly.

Edge8, the company behind this institute, is roughly 20% smaller in 2026 than it was a year earlier. That number is not a strategy we are proud of. It was mostly attrition. People left on their own. We actually let only two people go. The rest chose to move on, and we would have kept many of them.

Over the same window, revenue is up 60% since January.

Sit with the gap. A fifth fewer people, and 60% more revenue. We did not backfill every departure one for one. AI absorbed the work, the remaining team operated with the other 50%, and output kept climbing while headcount fell.

That is infinite leverage, and the honest version matters. This is not a story about cutting people to lift profit. We did not engineer the shrink, we are not celebrating the losses, and the lesson is not "do more with fewer." The lesson is that when leverage is real, output and headcount stop moving together. The 60% is the proof. The 20% is carried with humility.

The proof: 20% smaller team, 60% more revenue People down twenty percent, mostly attrition. Revenue up sixty percent over the same window. The gain is the proof; the loss is carried with humility. A REAL COMPANY · CARRIED WITH HUMILITY The gap is the leverage 20% smaller People mostly people who chose to leave 60% larger Revenue over the same window The 60% is the proof. The 20% is carried with humility

Why This Is Urgent Now

The common objection is that AI is not ready. That is the wrong frame. The point is not whether AI is ready. The point is that the speed of AI exposes bad process.

Knowledge living in one person's head. Reliance on a superstar nobody can replace. Tribal knowledge that was never written down. Under AI's speed, these are revealed for what they always were: simply poor business practice. Good practice means documented workflows and information that does not live in a single skull. That is ABC and workflow design again, now load-bearing.

Think of the old science-fiction starfighter. The craft moves faster than any pilot can track alone, so no human is fast enough to fly it without AI in the loop. You do not need AI implanted in your body. You need it on your team. The modern world moves too fast to fly clean otherwise. One clean hit lands only because the AI was flying with you.

Lead the Other 50%

Put the three pieces together and the system is complete. The three skills are how you lead AI. The Four Offices are where that leadership lives, with a central database underneath. Infinite leverage is how you prove it worked, in pull requests and features measured against goals.

The AI Officer Institute teaches founders and executives to run all three as one operating model. If you want to learn to lead the other 50%, apply for the next cohort. Your AI team is already waiting.

For the full business cut of this argument, including the deeper version of the Edge8 numbers and what they mean for how companies reorganize, read the complete argument on Edge8.

Frequently Asked Questions

How has leadership changed because of AI?
Leadership used to be 100% people: hire well, build culture, hold people accountable. That half is not going anywhere. The AI era adds a second half, because half of every workflow can now be directed to, delegated to, and executed by AI. The AI Officer Institute calls this the other 50%: the skill set that lets a leader direct AI, not just manage people. It rests on three leadership disciplines: information architecture (Always Be Cataloguing), workflow design, and writing instructions for code and AI.
What are the Four Offices of the Future?
The Four Offices are Revenue, Operations, Talent, and Innovation. Every outcome a company cares about maps to one of them. Underneath all four sits a central database. No central database, no organized information, no leverage. To turn the structure into a plan, ask the same four design questions of each office: What is the process? What outcome does it generate? How many human tokens do you need to put in? Where does AI fit on the team?
How do you measure the value of AI?
Most AI value is described in fog: faster, smarter, more efficient, none of it measurable. Infinite leverage is the measurement layer. A single human token can be infinitely multiplied by AI, and you know it is real when concrete units of value ship into the digital world: pull requests, features, and published pages measured against goals. A pull request is one finished unit of change shipped into a real system. You either hit the goal or you did not. Value stops being a feeling and becomes a count.
Why does AI expose bad business process?
The common objection is that AI is not ready. That is the wrong frame. The speed of AI exposes bad process. Knowledge living in one person's head, reliance on a superstar nobody can replace, and tribal knowledge that was never written down are revealed for what they always were: poor business practice. Good practice means documented workflows and information that does not live in a single skull. That is cataloguing and workflow design, now load-bearing.
Can you grow revenue with a smaller team using AI?
Yes, when leverage is real, output and headcount stop moving together. Edge8, the company behind the AI Officer Institute, is roughly 20% smaller in 2026 than a year earlier, mostly through attrition the team is not proud of (only two people were let go; the rest chose to leave). Over the same window revenue is up 60% since January. Departures were not backfilled one for one. AI absorbed the work and the remaining team operated with the other 50%. This is not a story about cutting people to lift profit. The 60% is the proof of leverage; the 20% is carried with humility.

Learn to Lead the Other 50%

The AI Officer Institute trains founders and executives to run the three skills, the Four Offices, and infinite leverage as one operating model. Your AI team is already waiting.

Apply for the Next Cohort