Chapter 004　How Does the Executive’s Role Change in the Age of AI?

In Vol. I, Ch. 001 we defined management as the activity of continuously creating Future Value. In Vol. I, Ch. 002 we concluded that AI cannot become a CEO. This chapter asks what lies past those two conclusions: the practice. When the definition changes, what concretely changes in the executive’s job? What is let go, what is held on to, and what begins that did not exist before? The shape of a day, the agenda of a meeting, the priorities of personal development. This is a chapter about the office rather than a restatement of the theory. It stays with the individual officeholder; leadership as a capability distributed across an organization belongs to Vol. II, Ch. 012.

1 The question — why it arises now

What is the executive’s job? Remarkably few companies hold a clear job description for it. The president’s job is said to be deciding. The chair’s job is said to be watching. But what is decided, in what order, and how far the watching extends have been handed down as custom. The office was not defined. It accumulated. The fastest way to see its outline is to measure it in time. Most executives’ days are filled with hearing reports, approving proposals, and answering requests from outside. The three share one form. A question has already been framed, and an answer is returned to it. That form worked because the capability to produce answers was scarce. Read the structure of a market. Spot the anomaly in a number. Compare options and rank them. Each of these required long experience. So the people who had accumulated that experience were placed at the top of the organization. The upper part of the org chart doubled as a map of where judgment was distributed. AI rewrites that map. Analysis, comparison, and ranking are now supplied at a high standard. The information asymmetry the executive alone held thins as well. Company-wide numbers once visible only to the executive now appear on many managers’ screens at the same resolution. Two misreadings arise easily here. One is that executives become unnecessary. The other is that nothing changes. Both come from measuring the role by level of capability. The correct vantage point is not capability. It is supply and scarcity. What AI increased is the supply of answers. What has not increased is the supply of good questions. When answers get cheap, questions get expensive. That is all that has happened. So this chapter’s question takes this form. In a world where answers are supplied in abundance, what should the executive’s day be spent on? It is not a question about the name of the role. It is a question about the allocation of time.

2 Conventional answers and their limits

Three answers about how the executive’s role changes are in circulation. All three are practical. All three are insufficient. The first answer: “Executives should raise their AI literacy” This is the most widely repeated answer. Touch AI yourself. Actually use it. Know the limits of the technology. The claim is correct as far as it goes. No executive can judge the possibilities of a tool they have never handled. But this is a condition of entry, not a description of the job. Executives once learned spreadsheets. They learned electronic mail. The role did not change because of it. Fluency with a tool is a precondition of the office, not its content. Make literacy the answer, and the executive’s learning stops at operation. Which prompt to write. Which system to install. Which vendor to select. What is actually needed is the judgment of what AI should be made to do and what it should not. That judgment does not come out of operational knowledge, however deep the operational knowledge runs. The second answer: “Executives become the final approver” The second answer comes from the structure of responsibility. AI produces the proposal; a human approves it. The executive therefore becomes the last gate. The answer has a real strength. It locates responsibility clearly, and it is easy to explain to a board. In practice it breaks down quickly. The reason is volume. The number of proposals AI produces rises independently of the human capacity to approve them. Decisions that numbered a hundred a year become a thousand. If the executive reads each one, the check becomes a formality. A hollow approval looks like the discharge of responsibility and secures nothing. Human-on-the-Loop Management points to a different answer. Human beings stand above the loop and design the whole system. Rather than adjudicating item by item, they design which range passes automatically and which range a human must decide. A design, once built, keeps working. An approval wears down every time it is given. The third answer: “The executive’s job does not change. Only the tools change” The third answer is the one experienced executives give. Move people. Make up your mind. Take responsibility. These were the same a century ago and they are the same now. At the core, this is correct. The essence of assuming responsibility does not change, and no technology has yet changed it. But the answer misses the composition of the job. The core can hold while the allocation of time moves, and the office becomes a different office. A twentieth-century plant manager and a present-day plant manager are both doing the work of protecting the floor. The contents of a day are nothing alike. A change of role usually appears not as a change of name but as a change of proportion. What the three conventional answers share is that all three cast the executive as a recipient. Receiving AI. Receiving proposals. Receiving change. But the executive is not a recipient. The executive is the first source of the questions that flow through the organization.

3 Redefinition — the executive is the person who

designs the questions Future Value Theory recasts the executive in the Age of AI as follows. The executive is not the person who produces the right answers. The executive is the person who designs the right questions. First Principle 9 states it in one line. Leadership Means Designing the Future. Leadership means designing the future — the right questions and systems rather than the right answers. To design is not to select an answer. It is to decide where answers come into being.

3.1 What moves from answers to questions

Start with the relation between questions and answers inside an organization. Every answer has a question standing in front of it. The answer to “should we continue this business” exists only after that question has been raised. Where the question is never raised, the answer is never even considered. So the range of options an organization can handle never exceeds the range of questions raised. This is where the executive’s largest leverage sits. Change one answer and one decision changes. Change one question and every decision hanging beneath it changes. Until now that leverage was not conspicuous. The work of producing answers was so heavy that the work of raising questions looked small beside it. When AI takes over the work of answers, the proportion inverts. Question Design surfaces at the center of the executive’s job. Question Design is the name for this work. It has three layers. The first layer is which question to raise. “How do we lower unit cost” and “will this cost structure still exist in ten years” carry an organization into entirely different futures. The first prolongs the existing business. The second unsettles the definition of the business itself. Which of the two is chosen is not a matter of analytical skill. It is the executive’s choice. The second layer is in what order to raise them. Ask “who will do it” first, and the answer shrinks to the range of people currently on hand. Ask “what should be done” first, and the range of people itself becomes a subject of examination. Order quietly decides the conclusion before anyone has argued about it. The third layer is to whom to raise them. The same question put to corporate planning, to the front line, or to AI returns answers of different kinds. Deciding the addressee of a question is organizational design itself.

3.2 From Manager to Leader, from Leader to Architect

The change in the executive’s role can be organized as three stages. Manager. The Manager achieves given targets with given resources. The question arrives from outside. The job lies in the accuracy of the answer and the reliability of execution. Most twentieth-century management education was designed to build this capability. Leader. The Leader sets the target itself. Choosing where to go, and moving people in that direction. The question is no longer given; it is chosen. Management writing since Drucker has mainly described this stage. Architect. The Architect stands one level above again. Rather than setting individual targets, the Architect designs the mechanism from which targets and questions keep emerging. Deciding the relationship among people, AI, and capital. Drawing the boundary of who may decide what. Building the circuit in which learning happens. One misreading must be avoided. The three are not replacements for one another. They are nested. Becoming an Architect does not remove the executive’s responsibility for management. What is removed is the need for the executive to perform that management personally. AI absorbs the Manager’s job fastest. Standardized judgment, routine allocation, progress monitoring. These are territory AI can carry. The Leader’s job is supported in part. The choice of direction stays with a human. The Architect’s job grows heavier as AI grows, because the number of elements requiring design increases. The executive’s movement in the Age of AI can therefore be written in one line. Leave management behind, hold leadership, and shift the center of gravity toward design.

3.3 AI still cannot raise a question

There is an objection. AI can generate questions too. As a matter of fact it can enumerate questions without limit, and the quality of the enumeration improves year by year. But enumerating questions and raising a question are different acts. To raise a question is to select one from countless candidates, and to decide that the organization’s time and capital will be spent there. That selection carries responsibility. First Principle 4 states it. AI Optimizes. Humans Define. AI optimizes; humans define value, purpose, and direction. Deciding what to ask sits on the side of definition.

4 Structure — the job broken into five

With Question Design at the center, how is the executive’s job composed? Future Value Theory expresses it in one equation. Leadership = Purpose × Question Design × Capital Allocation × System Architecture × Trust This is the Leadership Formula. What deserves attention is that it is multiplication, not addition. Under addition, a weak term can be covered by a strong one. Under multiplication, the moment one term goes to zero, the whole product goes to zero. Purpose at zero, and the most skillful capital allocation has no direction. Question Design at zero, and the answers AI produces cannot be tested. System Architecture at zero, and an excellent question never reaches the organization. Trust at zero, and the design is never executed. The executive cannot retreat into the one term they happen to be good at. No term compensates for another. Five roles follow from the equation. Purpose Designer — defines what future to create. Capital Allocator — allocates capital toward that future. System Architect — designs the mechanism through which humans and AI amplify each other. Culture Builder — builds a culture that learns and keeps changing. Future Creator — turns possibility into reality. The five are not parallel. Purpose sets the direction. Question Design distributes the questions. Capital Allocation moves the resources. System Architecture turns all of it into a circuit, and Trust puts current through the circuit. Rearrange the order and the job does not function.

4.1 Future Time — where the returned time goes

One more equation shapes the structure of the office. Future Value = Future Time × Future Capability Future Time is time intentionally invested in creating the future. Future Capability is the capability that converts that time into value. This equation is multiplication as well. With time at zero, no Future Value appears however high the capability. The effect of adopting AI is usually described in terms of cost reduction. For the executive the largest effect is different. Time comes back. The problem is that the destination of returned time decides itself if left alone. Existing work flows into an empty slot. Meetings increase. Reports increase. External obligations increase. Time for the present always arrives wearing the face of urgency. Time for the future does not wear that face. So it is deferred, and it is deferred again. Future Time therefore does not arise naturally. It is a resource that does not exist unless it is reserved in advance. Designing the executive’s job comes down, in the end, to designing how much of one’s own time is assigned to the future.

5 What it looks like in practice — how the executive’s

day changes From here the picture is concrete. For each of the five roles, we describe how the executive’s day looks before AI enters and after. Purpose Designer Before. Discussion of purpose concentrated in the annual policy address and in the season when the medium-term plan was drafted. On every other day, purpose was a wall poster rather than an agenda item. The executive was the person who spoke the purpose, not the person who inspected it. After. Re-examining purpose becomes ordinary work. AI presents changes in societal challenges, advances in technology, and shifts in the language customers use, continuously. Into the executive’s day comes time to confirm whether our reason for existing is still effective against this change. What was an annual event becomes a routine inspection. Even so, the writing of the purpose belongs to the executive. AI can detect the signs that a purpose has begun to drift from reality. It cannot decide how the drift should be corrected. Capital Allocator Before. Capital allocation concentrated in the budgeting season. Divisions raised their requests, the executive assessed them, and an annual allocation table was fixed. Much of the basis for allocation was last year’s results and the negotiating power of the division. After. Allocation becomes continuous. Results by business, changes in the market, and the progress of investments are visible at all times, so there is less necessity to settle everything once a year. Into the executive’s day come short, repeated intervals asking where resources should move now. What changes in quality is the object of allocation. To people, to research, to data, to AI, and to societal challenges. It becomes the work of allocating diverse capital as an integrated whole, not financial capital alone. A budget table is a record of which future the enterprise chose. System Architect Before. Designing mechanisms was the work of the information systems function and the human resources function. The executive approved finished mechanisms rather than designing them. Changing the organization was handled as one part of personnel rotation. After. This becomes one of the executive’s central tasks. The division of roles between humans and AI. The boundaries of decision rights. The range that passes automatically, and the range a human must decide. All of these are designs that fix the speed and the safety of the enterprise at the same time. There is no one to whom they can be delegated. Into the executive’s day comes time to design who should make a given judgment, and on what basis. Not the individual judgment itself, but the circuit through which judgments are made. Culture Builder Before. Culture-building was spoken of in values training, in the company newsletter, and at anniversary events. Its effects were hard to measure, and it was an area whose priority slipped easily. After. Culture becomes a more urgent management problem. In an organization where AI issues proposals, thought stops quietly unless there is a culture that tests those proposals, argues against them, and overturns them. Without a culture of continuous learning, people cannot keep pace with the advance of the technology. Into the executive’s day comes time to check whether this is a setting in which dissent comes out easily. Do not speak first in the meeting. Ask for the counter-evidence to AI’s proposal before anything else. Culture forms as the accumulation of the executive’s small behaviors. Future Creator Before. New business was the responsibility of a dedicated unit, and the executive sat on the receiving end of its reports. In many companies new business occupied the position of what gets done with spare capacity. After. The executive participates personally in the process that turns possibility into reality. Because AI widens the set of options, the number of branch points in decision-making increases. Which possibility to bet on cannot be delegated downward. Into the executive’s day comes time spent facing what has not yet become a number. Unmet customer need. Unsolved societal challenge. Unapplied technology. Moving only once the numbers are complete is not a bet. It is following.

5.1 What to let go, and what must never be let go

Designing a day starts from subtraction. Work that should be let go has three properties. First, the criterion of judgment can be written down. Second, there are enough comparable precedents. Third, a mistake can be recovered from. Routine approvals, progress checks, summarizing documents, analyzing performance. These can go. Work that must never be let go has three properties of its own. First, it is work that sets the criterion itself. Second, it is work for which no precedent exists. Third, it is work in which a mistake cannot be recovered from. Concretely there are five. Setting the purpose. Deciding long-horizon capital allocation. Drawing the boundary of authority between humans and AI. Deciding whether to trust a person. And deciding to exit. Exit is on the list for a reason. AI can compute the rationality of an exit precisely. But an exit is not only a matter of numbers. It is the breaking of a promise, the ending of a relationship, and the negation of somebody’s effort. Only a person who can assume responsibility can hand it down. Getting this distinction wrong produces two kinds of failure. An executive who holds on to what should be let go slows the organization down. An executive who delegates what must never be let go loses the organization’s direction. The second failure is far more dangerous than the first.

5.2 How to rebuild the agenda of the executive meeting and the

board A change of role finally shows up as a change of agenda. If the agenda has not changed, the role has not changed. In most companies, reports occupy the majority of the executive meeting. Sales progress, deal status, issues by division. All of it is necessary information. But reports can be generated by AI. Time spent with human beings reading generated reports aloud is a waste of Future Time. The rebuild we propose is simple. Finish the reports before the meeting, and begin the meeting with a question. Concretely, split the agenda into three layers. The first layer is the quantitative reporting AI presents automatically. It is not handled in the meeting; it is circulated in advance. The second layer is the part of the reporting that covers anomalies and exceptions only. This is processed quickly. The third layer is the layer of questions. Only questions with no prepared answer are placed there. Third-layer items are written in this form. “Which of our assumptions is becoming obsolete right now?” “Will the definition of this business still hold in five years?” “If we were to move capital now, from where to where?” None of them can be answered by a report. The board can be rebuilt on the same structure. A board has two functions: oversight of the past, and design of the future. Traditionally most of its time went to the first. Once AI can supply the material for oversight precisely, the time the first function requires can be compressed. Move the time that is freed to the second. The benchmark we put forward is to give half the board’s discussion time to future agenda items. The appropriate ratio changes with the company’s circumstances. But a board that has not consciously decided the allocation will almost certainly tilt toward the past. Agenda design is the executive’s most direct practice of Question Design. Who decides the agenda? That is the answer to the question of who is managing the enterprise.

6 Questions for the executive

Before the closing questions, one word on the executive’s own development. When the role changes, the capabilities to be trained change with it. Executive education so far has trained the capability to produce answers. Financial analysis, strategy formulation, case study. Training for producing good answers quickly. The value of these does not disappear. Their priority falls. Four capabilities, in our view, should take priority now. First, the capability to raise questions. Training in doubting assumptions and rearranging the order of questions. It can be done daily. Before giving an answer in a meeting, check once whether the question being handled is the right one. That alone is training. Second, the capability to design. The power to draw boundaries between human and AI, between authority and responsibility, and between automatic and manual. This grows only out of experience. Design within a small scope, observe the result, and redraw the line. That repetition is the only route. Third, the capability to build trust. The last term of the Leadership Formula. The more AI there is, the harder the process of decision-making becomes to see from outside. Whether people accept what they cannot see depends on trust. First Principle 8: Trust Compounds Faster Than Capital. Trust compounds faster than capital and becomes the last durable advantage. Fourth, the capability to keep learning. First Principle 5: Learning Is the Ultimate Competitive Advantage. Learning is the ultimate competitive advantage, because knowledge and technology depreciate. A culture of learning does not grow inside a company whose executive does not learn. Some capabilities may fall in priority. Holding information in memory. Assembling documents. Covering the specialist knowledge of a particular domain. AI carries these. An executive spending time here is misallocating a resource. The argument, in one line. The executive in the Age of AI is an office that moves from producing answers to designing questions and designing the future. Three questions to close, each connected to tomorrow’s decisions. Question 1 — Of your own day, how many minutes went to the future? Open the calendar and count; the answer arrives immediately. For most executives that time is close to zero. Future Time does not exist unless it is reserved. Only time reserved first becomes Future Time. The time left over between meetings never does. Question 2 — In your most recent executive meeting, how many questions had no prepared answer? If there were none, that meeting was a reporting session. There is no longer any need for the executive to attend a reporting session. The value of a meeting is measured by the number of questions that could only be raised there. Question 3 — Of the work you currently hold, what could you let go without the organization breaking? An executive who cannot answer this has not designed the office. An executive who can answer and has not let go has a design that does not match execution. Letting go is not abandonment. It is the act of creating a resource for the future. What the three questions share is time. The change in the executive’s role appears not as a change in capability but as a change in what time is spent on. AI does not take work away from the executive. It returns time to the executive. What the returned time is used for is where one company’s future separates from another’s. Leadership Means Designing the Future. An executive who holds no time in which to design the future is not managing, however capable that executive is.

In brief

The executive in the Age of AI is an office that moves from producing answers to designing questions and designing the future.
The range of options an organization can handle never exceeds the range of questions raised. That is where the executive’s largest leverage sits.
The change of role appears not as a change of capability but as a change in where returned time is assigned.
From Manager to Leader, from Leader to Architect. The three are nested, not substituted.

Key concepts

Question Design / Leadership Formula / Future Time / Capital Allocation / Human-on-the-Loop Management

The chain of ideas

Future Time → Question Design → Leadership → Future Value → Enterprise Value

Related first principles

Principle 9 — Leadership Means Designing the Future. Principle 4 — AI Optimizes. Humans Define. Principle 3 — Capital Exists to Create Possibility. Principle 8 — Trust Compounds Faster Than Capital.

Related chapters

Vol. II, Ch. 012 “What Is Leadership in the Age of AI?” — digs into the executive as designer at the level of organizational capability
Vol. I, Ch. 005 “What Is Decision-Making in the Age of AI?” — the structure of the decisions that sit beneath the questions
Vol. II, Ch. 018 “What Should a CEO Learn in the Age of AI?” — turns the capabilities to be trained into concrete practice
Vol. VI, Ch. 052 “What Does It Mean to Redefine the Executive?” — rewrites the definition of the executive itself

Papers and companion volumes

Kadowaki, N. (2026a). Future Value Theory: A Management Framework for Enterprise, Capital, and Society in the Age of AI. VURA Working Paper Series. SSRN: https://ssrn.com/abstract=7120980 / Zenodo: https://doi.org/10.5281/zenodo. 21255662
Kadowaki, N. (2026b). Enterprise Redefinition: Toward an Enterprise Evolution Theory for the Age of AI. VURA Working Paper Series. (Published on Zenodo; under review at SSRN)
100 Questions on Management in the Age of AI, #017 “How Does AI Change the Executive’s Job?” / #019 “Why Do Executive Meetings Become Nothing but Reports?”

Chapter 004 How Does the Executive’s Role Change in the Age of AI?