Chapter 008　What Is Management Strategy in the Age of AI?

What is strategy? Twentieth-century management had a clear answer ready. Strategy is deciding where to concentrate limited resources. Which market to select, which position to hold, and what not to do. The definition worked for half a century. Its assumptions are now coming apart. What this chapter deals with is not the content of any particular strategy. It is how the work of strategy itself has to be redesigned.

1 The question — why it arises now

Strategy is the most worn-out word in the vocabulary of management. It is also the word most often used while its definition stays vague. Open what is called “strategy” at most companies. Inside you find an analysis of the market environment, a survey of competitor behavior, an inventory of internal strengths and weaknesses, and a set of target numbers three years out. Analysis, then options, then a choice. The format barely varies across industries. There was a reason for the format. Information was scarce, and analytical capability was scarcer. Few companies could read a market accurately. Fewer could read a competitor’s behavior structurally. Being good at analysis was itself a source of advantage. Strategy was the craft of translating that scarce analytical capability into a judgment about resource allocation. AI dismantles the core of that assumption. AI can analyze a market. AI can organize competitor behavior. It can inventory strengths and weaknesses, enumerate scenarios, and quantify each option to a high standard. It can do all of it in days, and in dozens of variants. Work that used to occupy a corporate planning department for months is becoming work of a few hours. What is happening here is not a gain in efficiency. The part of the strategy process regarded as the most specialized has turned into the cheapest part. The question therefore changes shape. It is not “how do we produce an excellent strategy.” It is “in a world where anyone can produce excellent analysis, what does the word strategy refer to?” When analysis is democratized, the choices based on analysis converge as well. Feed the same data to AI of the same capability with similar questions, and similar answers come back. The strategy decks of three competitors start to look remarkably alike. This is not a hypothesis. It is already observable across many industries. If the purpose of strategy is to create difference, then a procedure that no longer creates difference is no longer strategy.

2 Conventional answers and their limits

Three answers about strategy circulate today. Each was once correct. Each is losing its assumptions. The first answer: “Strategy is deciding what not to do” This is the most refined of the classical definitions. The essence of strategy is trade-off. A company that tries to do everything becomes nothing. So decide what you will not do. The claim rested on the hard fact that resources are finite. The problem is that the content of the finiteness has changed. What used to be scarce was hands, time, and analytical capability. Examining a new business meant assigning people, commissioning research, and spending months. Narrowing the field of examination was therefore strategy in itself. AI relaxes that kind of scarcity substantially. The cost of examining ten possibilities at once falls dramatically. So does the cost of running more experiments. The urgency of deciding what not to do is weaker than it was. Has the finite thing disappeared, then? It has not. It has moved. What remains finite is the attention of the executive, the trust of the organization, and the time staked on the future. There is a limit to the number of promises human beings can genuinely take on. There is a limit to the number of stories an organization can believe at the same time. The classical definition therefore gets rewritten. Strategy is not deciding what not to do. It is deciding what to keep taking on. From the craft of discarding options to the craft of sustaining a commitment. The object of the trade-off has moved from the set of activities to the use of time. This rewriting is not a rejection of the classical definition. It is a move in its range of application. Concentration is still required. What must be concentrated has shifted from narrowing the businesses to sustaining the intent. Run ten experiments in parallel while holding to one purpose for ten years. That is what concentration looks like in the Age of AI. The second answer: “Strategy is analyzing the environment, laying out options, and choosing the best” This is strategy as a procedure. Environmental analysis, option design, evaluation, selection. The sequence became the standard of management education. Its weakness has been pointed out for a long time. Analysis depends on past data. Data can speak only about what has already happened. Analysis-led strategy therefore converges, structurally, on an extension of what exists. AI does not remove that weakness. It strengthens it. AI evaluates a given set of options far more accurately than a human being does. But what generates the set of options is past cases and existing discourse. AI can increase the number of options. It cannot doubt the frame the options sit inside. Doubting the frame requires a subject willing to take on a sentence: this assumption may no longer hold. The result is that the more accurate the analysis, the shorter the distance between one company’s strategy and another’s. Precise analysis produces precise sameness. That is where the obsolescence of the strategy process sits. What has become obsolete is not analysis. It is the ordering that puts analysis at the start. The third answer: “Strategy is decided in the executive meeting and written into a document” The third answer concerns the form of strategy. The medium-term plan. The strategy deck. The annual review. At many companies, strategy refers to a document, a physical object. Documents have advantages. They fix an agreement and can be distributed across an organization. But a document begins aging the moment it is made. The problem is that the speed of aging has changed. In a world where assumptions hold for three years, a three-year plan was rational. In a world where assumptions move in six months, a threeyear plan is a ceremony. In practice, at most companies, a plan is either achieved or missed, and the errors in its assumptions are rarely examined at all. The three conventional answers share one view of the world. The environment is given, and the enterprise selects an optimal position inside it. Strategy in the Age of AI asks the question one level above. Who brings the environment itself into existence, and how?

3 Redefinition — strategy is conceiving a future and

redefining the enterprise, continuously Future Value Theory and Enterprise Redefinition define strategy again, as follows. Strategy is the process of conceiving what future should exist, and continuously redefining the enterprise itself by working backward from that future. The definition contains three shifts. A shift in the starting point, a shift in the object, and a shift in the form. We take them in order. Shift 1 — the starting point is not analysis but Future Vision Begin by separating forecasting from Future Vision, strictly. This distinction sits at the center of the chapter. Forecasting goes after the future to get it right. It extrapolates from current data, attaches probabilities, and draws the most likely picture. A good forecast is an accurate forecast. The quality of a forecast can therefore be checked afterward. Future Vision goes after the future to create it. It declares that a state which does not yet exist ought to exist, and carries the enterprise toward it. A good Future Vision is not an accurate one. It is a vision that moved people, moved capital, and as a result became real. The canon states it in one line. Forecasting predicts the future. Future Vision creates it. The difference is decisive in the Age of AI. AI forecasts better than people do. As data accumulates and models improve, the gap widens. A company that puts forecasting at the start of its strategy is therefore committed to calling a process strategy while AI replaces it. Future Vision, by contrast, is not something AI can carry. AI can answer what future is likely to occur. It cannot answer what future should exist. The second is not a question of fact but a question of value. To answer a question of value is to select a meaning and to take on responsibility for it. First Principle 4 says it. AI Optimizes. Humans Define. “AI optimizes; humans define value, purpose, and direction.” Future Vision is defined as the answer to three questions. What future should exist. Why that future is desirable. What role the enterprise should play within it. A document that cannot answer those three is not a strategy, however thick it is. It is a report on the environment. Shift 2 — the object of strategy is not the market but the enterprise itself In classical strategy, the object of strategy was the market. Decide which market to enter and where to stand within it. The enterprise itself was treated as a given bundle of resources. Enterprise Redefinition inverts that assumption. The primary object of strategy is not the market. It is the enterprise itself. The reason is simple. In the Age of AI, market positioning is copied quickly. Since analysis has been democratized, a good position is easy to find and a found position fills up fast. The advantage of position is thin. The capability to remake the enterprise itself, by contrast, is hard to copy. It is not an asset. It is a way of recombining capabilities. What you sell can be imitated. The habit of continuously rewriting your own definition cannot be imitated easily. First Principle 6 declares the position. Enterprise Exists to Redefine Itself. “Enterprise exists to redefine itself—continuous selfredefinition is its essence.” The sentence may look strange. Does an enterprise not exist for its customers? It does. What the principle states is the condition under which an enterprise continues to exist. When society changes, the value it needs changes. No enterprise can go on contributing to society while keeping the same shape. The purpose of continuing to contribute therefore requires continuing to change. Enterprise Redefinition is neither improvement nor transformation. Improvement takes the present self as given and makes it better. Enterprise Redefinition doubts the definition of the present self. And an improvement program has an end, while Enterprise Redefinition has none. Shift 3 — strategy is not a document but a cycle The third shift concerns form. Strategy is not a deliverable. It is a structure that keeps turning. The Enterprise Redefinition Cycle expresses it. Future Vision → Enterprise Redefinition → Execution → Learning → Future Value → (return to Future Vision) Conceive, redefine the enterprise, execute, learn, and Future Value emerges — and that Future Value makes a more distant Future Vision possible. Then back to Future Vision. There is no terminal point. The cycle is recursive rather than linear. Strategy as a document is a snapshot of one moment in that cycle. A medium-term plan is a photograph taken partway around. Because the photograph is called the strategy, the strategy is felt to be lost when the photograph ages. Call the cycle the strategy, and the account changes. What ages is the assumption, not the practice. Assumptions aging is evidence that the cycle is turning correctly. One point about the shelf life of a strategy belongs here. Where assumptions move fast, the content of a strategy ages fast. But not everything ages at the same rate. Of the five dimensions of Enterprise Redefinition — Purpose, Business, Organization, Capital, and Leadership — Purpose alone is different in kind. Business, organization, the allocation of capital, and the leadership structure all get rewritten substantially as the era moves. Core Purpose, however, usually keeps its core, even as its expression and the means of its realization evolve. The five dimensions do not change at the same frequency. The shelf-life argument therefore has to be split into two layers. The shelf life of means gets shorter. The shelf life of purpose gets longer. The faster the change, the more valuable a standard that does not change, because only Core Purpose can determine what may be discarded and what may not. An enterprise that changes at high speed without a core is not changing. It is drifting. The two-layer structure reaches into how strategy documents are written. At most companies, purpose and means sit in the same deck at the same weight. So every review puts both up for discussion at once. They should be managed separately. Doubt the means every quarter. Re-examine the purpose once every several years, and then deeply. An organization that doubts its purpose frequently is unsettled. An organization that never doubts its means goes rigid.

4 Structure — how strategy is assembled

Three structures carry the redefinition into practice.

4.1 The Future Value Chain — the order decides the strategy

Purpose → Learning → Redefinition → Creation → Enterprise Value This is the Future Value Chain. Enterprise Value comes last. Strategy work in practice often runs the chain backward. Set a target for enterprise value or profit first, divide out the revenue required, assign it to businesses, and line up the initiatives. Working backward is not the problem in itself. The problem is that the starting point of the working-back is a result. Put Enterprise Value at the start, and the moves that get selected skew toward extensions of the existing business. Only things that will certainly turn into numbers can be placed along the path of the calculation. Strategy becomes another name for the budget allocation table. What does keeping the order mean in practice? It means starting the strategy discussion from a confirmation of Purpose. Then confirming what was learned in the past year. Without learning, there is no basis for redefinition. Redefinition without a basis is a whim.

4.2 Future Back Planning and Future Horizon

The procedure that connects Future Vision to practice is Future Back Planning. Build-up planning starts from the present. Present capability, present customers, present earnings structure. It draws a future within reach of an extension of those. Under this method, the future that gets drawn is a function of the present. Future Back Planning runs the other way. First, place the future that should exist. Next, describe what kind of enterprise you have become in that future. Then trace back toward the present and decide what has to be started now. The two look similar and produce different conclusions. Build-up asks what we are able to do. Future Back asks what we must discard and what we must acquire in order to become the enterprise that future requires. Only the second question puts capability acquisition and capital reallocation on the agenda. How far out the future is placed is decided by the Future Horizon. The Future Horizon is the length of time an executive can genuinely take responsibility for. An executive who looks three years ahead has a strategy with a three-year reach. What matters is that the horizon is not a personal trait. It is an object of design. Tenure, evaluation metrics, the duration of capital, and the design of disclosure all govern it. The meaning of the horizon is concentrated in one equation. Future Value = Future Time × Future Capability This is the Future Time Equation. Future Value is the product of time directed at the future and the capability to create the future. Multiplication is the point here as well. Not addition. Under addition, high capability could compensate for short time. Under multiplication it cannot. An enterprise whose horizon is near zero produces Future Value near zero, however high its capability. And an enterprise that proclaims a long horizon without accumulating capability produces zero just the same. AI raises the second term substantially. It speeds learning, speeds design, and speeds experimentation. The first term — how many years ahead you are willing to take on — is not something AI can decide. In the Age of AI, therefore, the gap between companies opens widest on the Future Time side.

4.3 The apparatus that keeps the cycle turning

Turning the Enterprise Redefinition Cycle requires machinery that guarantees the turning. Strategy as a document has an approval process. A cycle has none. Three things are the minimum. First, write the assumptions down. Not the conclusion of the strategy, but the assumptions holding the conclusion up. Second, check periodically whether those assumptions have broken. Third, decide in advance who starts the redefinition when they do. A company with all three can rewrite itself without a crisis. A company without them rewrites itself only after performance deteriorates. The difference between the two corresponds closely to the difference between Level 2 and Level 4 of the Enterprise Redefinition Maturity Model (ERMM). At Level 2, the Improvement Enterprise, the company improves continuously and rarely redefines. It becomes increasingly efficient while remaining fundamentally unchanged. AI is a powerful accelerator of that efficiency. So the further AI deployment goes, the more comfortable the Level 2 enterprise becomes, and the less it moves. The gain in efficiency conceals the need for redefinition. At Level 4, the Continuous Redefinition Enterprise, the company redesigns itself before external disruption requires it. Strategy here is not a response to an incident. It is ordinary practice in calm conditions. Three cautions travel with the model. Progression is not linear; an organization may hold Level 4 capability in one dimension while remaining at Level 2 in another. Maturity is read across all five dimensions in balance, since strong purpose without adaptive organizational systems remains insufficient. And reaching Level 5 as rapidly as possible is not the objective; different industries may require different levels of organizational adaptability. What to ask is not how high the level is. It is whether the five dimensions are coherent with one another.

5 What it looks like in practice — how the work of

strategy changes on the ground Now lay the abstraction over the working life of an enterprise. Picture the corporate planning department of a manufacturer. Its year used to be fixed. Market research in the first half. Drafting the direction in the second half. Confirming the plan at the year’s end. Most of the staff time went into producing documents and securing internal agreement. After AI was deployed in earnest, the picture changed. Organizing the market, comparing competitors, enumerating scenarios — all finished within days. What to do with the freed time became, in itself, the new question. Here the road forks in two. One company used the freed time to produce more documents. Finer-grained analysis, more scenarios, thicker meeting packs. The conclusion is almost identical to last year’s. It arrived back at the same place, faster and more precisely. The other company used the freed time differently. It replaced the agenda of the executive meeting, swapping reporting for Future Vision. Analysis produced by AI was taken as the starting premise, and the human hours went to arguing whether the worldview the analysis assumes is correct. And once a year, the company set aside an occasion to put into words what future it wants to bring into existence. The difference is not AI performance. It is what each company did with the time AI returned. Within a few years the fork becomes a gap that cannot be closed. The first company has a strategy too. But its strategy can be reproduced by AI in a few days. What the second company is accumulating is not documents. It is an organizational habit of handling Future Vision. Habits cannot be bought. Patterns of Future Vision in publicly known cases The same shape can be read in real companies, within the limits of public information. A semiconductor company was long known as a maker of processors for graphics. What can be read from public information is that it built a general-purpose parallel computing platform early, and opened the development environment to outside engineers. That is not a move derivable from an analysis of its main market at the time. A future in which parallel computation would be needed broadly was placed first, and the move was worked backward from it. A software company moved its definition from selling its products outright to providing the platform on which customers’ operations keep running. In the early phase of the shift, what can be read from public information is that the decision cut into its own existing earnings structure. That is the kind of choice analysisled decision-making does not select easily. A diversified electronics company kept exchanging its businesses over a long period while holding to a core of delivering experiences that move people. What it sold changed many times. Why it sold did not. As far as public information allows, all three chose their position as the consequence of a Future Vision rather than as the conclusion of an analysis. And because the Future Vision came first, each could make the decision to cut into an existing business. What can be read from public information extends that far and no further. The patterns of a strategy breaking down The inverse picture is worth setting out. Companies whose strategy stops working share a small number of patterns. The first pattern is the absence of Future Vision. The analysis is precise, the options are exhaustive, the evaluation is sound. But nobody can answer what future the company wants to bring into existence. Such a company’s strategy becomes a collection of reactions to its environment. The second pattern is the stopping of the cycle. A Future Vision exists. Once declared, it is never inspected. The assumptions break and the document lives on, while the front line silently absorbs the gap between the document and reality. The third pattern is the loss of the core. In a rush to respond to change, the company swaps out its Purpose along with its businesses. Employees are told a different story every year and stop believing any of them. Trust compounds faster than capital, and it is lost faster too. The three patterns correspond to a missing starting point, a missing cycle, and a missing core. Designing a strategy means holding all three in place at once.

6 Questions for the executive

The argument, in one line. Strategy in the Age of AI is an endless cycle: conceiving the future that should exist, and continuously redefining the enterprise itself by working backward from that future. Not analysis. Not a comparison of options. Not a document. All of those are components of the cycle, nothing more. First Principle 9 states the position briefly. Leadership Means Designing the Future. “Leadership means designing the future— the right questions and systems rather than the right answers.” Designing does not mean optimizing within given conditions. It means setting the conditions. Three questions to close. Each can be answered at your next executive meeting. Question 1 — Which part of your strategy deck could AI not reproduce in a few days? The reproducible part will stop creating any difference before long. If there is a part that cannot be reproduced, it is the Future Vision and the responsibility that was taken on. If the whole deck is reproducible, the company does not yet have a strategy. It has an analysis. Question 2 — What future do we believe should exist? Can your executive team answer that in the same words? If the answers differ, the enterprise is heading toward several futures at once. If the answer is “in line with the growth of the market,” that is not a Future Vision. It is a forecast. Question 3 — When the assumptions break, who starts the rewriting of the strategy? At most companies the role is assigned to no one. Deteriorating performance is the only trigger in place. Turning the cycle without waiting for deterioration requires the authority and the procedure for starting it to be designed in advance. None of the three questions asks which strategy is correct. All three ask whether the practice of strategy has been designed correctly. In a world where AI carries the analysis, the value of strategy moves away from analysis. What remains is the Future Vision of what future should exist, and the work of continuously remaking the enterprise into one that matches that future. The content of a strategy will go on aging, again and again. That is as it should be. To ask for a strategy that does not age is to ask for a world that does not change. What should be asked for is the capability to notice the aging, rewrite, and declare again. And at the center of that cycle, exactly one thing has to stay fixed. The answer to why this enterprise exists. Strategy is the craft of realizing an unchanging purpose in continuously changing form.

In brief

Strategy is the cycle of conceiving the future that should exist and continuously redefining the enterprise itself by working backward from it.
AI can carry forecasting, but only human beings can carry the Future Vision of what future should exist.
The primary object of strategy is not the market but the enterprise itself. Advantage in positioning is imitated quickly.
The shelf life of means gets shorter. The shelf life of Core Purpose, if anything, gets longer.

Key concepts

Future Vision / Enterprise Redefinition / Future Back Planning / Future Horizon / Purpose

The chain of ideas

Purpose → Future Vision → Future Back Planning → Enterprise Redefinition → Future Value

Related first principles

Principle 6 — Enterprise Exists to Redefine Itself. Principle 9 — Leadership Means Designing the Future. Principle 4 — AI Optimizes. Humans Define.

Related chapters

Vol. V, Ch. 045 “What Is Enterprise Transformation in the Age of AI?” — separates transformation from redefinition explicitly
Vol. VI, Ch. 057 “How to Carry Out Enterprise Redefinition” — the seven stages that keep the cycle turning
Vol. I, Ch. 005 “What Is Decision-Making in the Age of AI?” — the implementation of working backward and of the Future Horizon
Vol. IV, Ch. 036 “What Is Management That Creates Future Value?” — the route from management practice Future Vision to daily

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, #014 “Should We Let AI Write Our Strategy?” / #054 “In the Age of AI, Is the Future Predicted or Created?”

Chapter 008 What Is Management Strategy in the Age of AI?