Chapter 031　How Does AI Change Future Value?

Does AI increase an enterprise’s Future Value? Most executives assume it does. AI does speed up analysis, widen the range of options, and lower the unit cost of execution. Yet the companies that have deployed AI most aggressively are not always the ones whose Future Value has risen most. The same tool widens the future at one company and narrows it at another. Where does that asymmetry come from? This chapter takes up one relationship only — AI and Future Value — and sets out its structure.

1 The question — why it arises now

Volume III dealt with Future Value itself. Its definition, its difference from present value, its relation to revenue. It fixed the indicator and the management practice. It worked through the relationship to Purpose, to Mission, and to the brand. Some readers will have noticed something. AI has barely appeared in any of it. That is not an oversight. The structure of Future Value holds without AI. The capability to create value that does not yet exist was present in enterprises long before AI arrived. But AI does not sit outside that structure either. The Future Capital Equation makes this plain. AI is written into it explicitly, as one term. The theory has treated AI as a constituent of Future Value from the beginning. The problem is that almost no one in practice understands what that term means. Ten years ago the management question about AI was whether to adopt it. Five years ago it was where to start. As of August 2026 both questions are already becoming historical. Most companies use AI. Most business processes touch it somewhere. Adoption itself is no longer the issue. A new divergence has opened in its place. Investment in AI is rising. Processing volume is rising. And still, many executives have no sense that their company’s future has widened. That divergence is not a matter of feeling. It is a matter of structure. AI is certainly changing something inside the enterprise. What it is changing is not necessarily Future Value. The question therefore has to be broken into three. Where in Future Value does AI act? Can AI be the origin of Future Value? And can AI reduce Future Value? The third question is rarely asked. It matters most. An executive team watching only the routes by which AI adds cannot notice that it is walking one of the routes by which AI subtracts. This chapter does not take up the division of roles between AI and people. It does not take up what makes an AI deployment succeed or fail. Volume I covered both — Vol. I, Ch. 009 and Vol. I, Ch. 010. One thing is at issue here. What position does AI, as a form of capital, occupy relative to Future Value, which is a capability?

2 Conventional answers and their limits

Three answers to this question circulate today. Each is partly right. None of them, as stated, can be used to make a decision. The first answer: “The company that has AI has the future” This is the most artless version. AI is the central technology of the era, and the companies that use it at a high level will win the next era. Its weakness is that it mistakes AI for a capability. AI is capital. The Future Capital Equation places it on the capital side, and so does the Capital dimension of Enterprise Redefinition. Capital creates no value by itself. Owning a plant is not the same as making a good product. Holding cash is not the same as making a good investment. Holding AI is not the same as creating the future. Capital works only once it has been allocated, combined, and pointed at a purpose. AI is no different. AI held as inventory adds nothing whatsoever to an enterprise’s Future Value. The second answer: “AI predicts the future, so the uncertainty in Future Value falls” This is a more refined answer. As predictive accuracy improves, an enterprise can see the future with higher confidence. If it can see, its investment decisions about the future become reliable. There is a fundamental confusion here. Prediction operates only along the extension of existing data. About an event that has never once occurred, AI has no material from which to predict. And Future Value is the capability to create value that does not yet exist. Because it does not yet exist, there is no data. What lies at the core of Future Value is precisely what cannot be an object of prediction. There is a further point. Even if predictive accuracy became perfect, it would not settle which future to choose. Knowing the probability attached to each of three futures does not tell you which of them your enterprise should take on. That is a separate judgment. Prediction guesses the future correctly. What an enterprise has to do is create it. We stated this distinction in 100 Questions on Management in the Age of AI, #007. The third answer: “As AI advances, AI will eventually create Future Value” The third answer treats the matter as a question of time. It cannot be done now, but as performance rises, AI will set purposes, redefine businesses, and create value. To reject this answer, we do not need to argue about the limits of the technology. We must not argue that way. An argument grounded in the limits of technology collapses every time the technology advances. What is needed is a ground that holds independently of performance. It lies in the structure of responsibility. We take it up directly in the next section. What the three conventional answers share is that they treat AI as a subject acting alone. AI has. AI predicts. AI creates. The theory shows a different picture. Alone, AI amounts to nothing.

3 Redefinition — AI Capital becomes Future Capital only

in combination with people Here is the reading at the center of this chapter. The Future Capital Equation, character for character. Future Capital = Financial × Human × Learning × Trust × AI × Knowledge × Ecosystem × Purpose Eight terms, multiplied. We call the AI term AI Capital in this chapter, and the Human term Human Capital. Neither is a new concept. We are naming terms that already sit inside the equation. What it means that this is multiplication That the equation multiplies is not decoration. It is the criterion of management judgment itself. Under addition, raising AI Capital raises the total even when the rest is weak. Under multiplication it does not. If Human Capital is zero, Future Capital is zero however large AI Capital grows. The same holds if Purpose is zero. The same holds if Trust is zero. An abundance of financial capital cannot compensate for absent purpose, and advanced AI cannot compensate for absent trust. So AI Capital is not Future Capital on its own. It becomes part of Future Capital only in combination with the other terms. This is not abstract. On the ground it appears like this. A company distributes a high-performance model across the whole organization. The people using it do not understand their own business deeply. Nobody can judge whether an output is sound. The result is a large volume of plausible documents and no change in any decision. AI Capital has increased. Future Capital has not. The reverse picture exists too. The models in use are ordinary. The people asking the questions are sharp. Somebody can convert an output into a business decision. In that company the same AI produces far more Future Capital. The difference does not lie on AI’s side. It lies in the terms AI is multiplied by. Three points of connection Where, concretely, do AI Capital and Human Capital join? There are three points of contact. First, at the point where the question is set. AI answers what it is asked. It is silent about what it is not asked. What to ask can be decided only by people who know the structure of the business and the challenges of society. The quality of the question fixes the ceiling on the output. Second, at the point where output becomes a decision. An AI output is, in itself, a proposal. A proposal becomes a decision at the moment someone adopts it. That conversion happens only on the human side. Third, at the point where the result is carried. Decisions have consequences. When something does not work, someone bears it. This point is the core of the argument that follows. Why AI cannot stand at the origin of the chain The Future Value Chain runs in this order. Purpose → Learning → Redefinition → Creation → Enterprise Value The origin is Purpose. Can AI set a Purpose? It cannot. But not because AI lacks capability. Purpose is the act of deciding, among countless possibilities, that one of them has meaning. That decision cannot be verified as true or false. Which of the other choices would have been right can never be established, even afterward. Nobody can know what would have happened had the enterprise taken a different road twenty years ago. What makes an unverifiable decision stand is not proof. It is the act of carrying it. Someone declares that the consequences of this choice are theirs to bear, and the choice becomes a decision. To set a purpose, therefore, is the same act as to take on responsibility. AI cannot take on responsibility. This is not a question of computing power. It is a question of attribution. AI absorbs no loss. It does not lose its job. It does not lose trust. It does not explain itself to shareholders. It does not bow to employees. It cannot resign. It cannot make amends. A subject with nothing to lose cannot make a wager. A subject that cannot wager cannot choose a future. What matters here is that raising performance does not change this structure. A model a hundred times more capable has no more to lose. Responsibility is not a function of capability. So this is not a problem that time will resolve. One objection is available. Build an institution, it says. Give AI judgments legal personality, set up a mechanism that pays damages, and responsibility is established. But even if such an institution existed, the damages would be borne by the people or the legal person behind it. The point of attribution simply moves outside AI. An institution is a device for making the location of responsibility explicit. It is not a device for moving the subject of responsibility into AI. A second objection is available too. Human executives do not carry responsibility adequately either. As an observation, that is correct. But carrying responsibility badly and standing outside the position of carrying it are different things. The first is a problem of improvement. The second is a structure. The fourth of the First Principles states exactly this structure. First Principle 4 — AI Optimizes. Humans Define. AI optimizes; humans define value, purpose, and direction. “Define” here does not mean only deciding. It includes carrying what has been decided. Definition and responsibility are two faces of one act. So AI acts powerfully from the second stage of the chain onward. It cannot touch the first. That asymmetry is the foundation of everything in the relationship between AI and Future Value.

4 Structure — what AI brings to each of the five stages

What, then, does AI actually do at each stage of the chain? We take them in order.

Figure IV-1　. The Future Value Cycle — value circulates

Purpose. AI cannot make a purpose. It does three things around one. First, it makes the distribution of societal challenges visible; it can organize where unsolved challenges are concentrated out of large volumes of primary material. Second, it detects the gap between the purpose an enterprise declares and the way it actually allocates resources. Any mismatch between word and deed shows up in budgets and headcount. Third, it translates the purpose into the language of each function. All of this, however, is work that begins after a purpose has been set. AI’s contribution is not to create the origin. It is to make the origin function. Learning. This is the stage where AI contributes most. Collection, summary, comparison, and hypothesis generation become faster by an order of magnitude. One equation should be recalled here. Future Value = Future Time × Future Capability AI recovers time from executives and from the organization. If the recovered time is spent as Future Time, Future Value rises. Hours once spent producing meeting materials, spent instead on re-examining assumptions, deepen learning. If the recovered time goes into doing more of the existing work, Future Time does not rise. AI returns time. It does not decide where the time goes. Redefinition. AI widens the set of redefinition options. It can lay out, as many candidate cases, what would happen if the business were defined along a different cut. That work is heavy for people and easy for AI. But the substance of redefinition is not choosing. It is letting go. It is the act of deciding which of the businesses that carried the company for decades, which of the people it raised, and which of the trading relationships it built will be released. AI can produce the arithmetic of letting go. It cannot carry the pain of it. Creation. The unit cost of prototyping and validation falls. The number of attempts rises. This reliably raises the probability of creation. There is an asymmetry here, though. AI is strong where data is plentiful and weak where there is none. And a market that does not yet exist has no data. The newer the object of creation, the smaller AI’s contribution. In the region of highest Future Value, AI is least dependable. Enterprise Value. AI helps make Future Value Creation Capability visible and explicable. Organizing non-financial information, preparing for dialogue with investors, keeping indicators continuously updated. Running an instrument like the VURA Future Index (VFI) becomes a realistic workload with AI. One caution applies. A rise in valuation and a rise in Future Value are different events. A company that has only become better at explaining itself is also rewarded by the market, for a while. AI Integration is one of seven terms Place the other equation. FVCC = Purpose × Learning × Redefinition × AI Integration × Ecosystem × Capital Allocation × Trust Of the seven terms that compose Future Value Creation Capability, AI Integration is one. This equation multiplies as well. The relationship is multiplicative: weakness in any single capability weakens the whole, so purpose, AI, and capital are each individually insufficient, and Future Value emerges only when all seven reinforce one another. Two consequences follow. One. However high AI Integration is driven, if Purpose is zero then FVCC is zero. A company that invests across the enterprise in AI while unable to say what it exists for has no Future Value Creation Capability. Two, running the other way. Improvement in AI Integration does not automatically produce improvement in the other six. It can take attention away from them. Executive attention is finite. While AI adoption occupies the corporate agenda, nobody is looking at Trust, at Ecosystem, or at Capital Allocation. That is what it means for AI Capital to be one term in a product. It is necessary. It is never sufficient. And concentrating on one term can starve the rest.

5 What it looks like in practice — AI can also reduce

Future Value So far we have followed the routes by which AI adds Future Value. Time recovered, options widened, attempts multiplied, visibility sharpened. These are real. Take the adding side first. There are four routes. The first is the reallocation of time. AI absorbs routine work and the released hours go into designing the future. The second is an extension of the Future Horizon. The effort required for long-range analysis falls, and the number of years management can hold in view grows. The third is the number of experiments. When prototyping and validation get cheaper, the same budget tests more hypotheses. The fourth is the extension of the ecosystem. Knowledge from other fields can be handled quickly, so partnerships that were previously out of reach become possible. All four share one property. None of them happens automatically. Time left alone is absorbed by other work. A horizon does not extend unless someone decides to extend it. The number of experiments does not rise without a design that permits failure. Every adding route presupposes human design. And where the design is absent, the routes by which AI reduces Future Value appear. There are three. Route 1 — the acceleration of short-term optimization AI is strong against a given objective function. The clearer, the more measurable, and the faster in feedback that function is, the better AI performs. Inside an enterprise, few indicators meet those conditions. Quarterly revenue. This month’s gross margin. Next week’s conversion rate. All measurable, all fast in feedback. The possibility of a market that will exist in ten years, by contrast, cannot be measured. Its feedback returns only in ten years. So when management runs AI without stating the objective function, AI moves toward the indicators that are easy to measure. The organization accepts this as rational optimization. The result is the acceleration of short-term optimization. If the direction is right, speed is good. If the direction is short-term, speed means traveling further in the wrong direction. Before AI, short-term optimization was constrained by the supply of human hands. When the constraint comes off, the bias is amplified. Route 2 — the tilt toward making the existing business efficient The second route is a distortion in capital allocation. Many companies judge AI investment by expected payback. It looks like sound discipline. But this criterion fixes the destination of AI investment in a single direction. Payback can be computed only for an existing business. An existing business has a current cost structure. It has transaction volume. It has a payroll. So a reduction can be estimated. A business that does not yet exist has nothing to compare against. Nothing can be estimated. Screen by payback, therefore, and the AI budget will always flow toward making the existing business more efficient. What follows from that? The Enterprise Redefinition Maturity Model (ERMM) describes this state clearly, as Level 2. The Improvement Enterprise actively pursues operational excellence; digital transformation becomes systematic; AI adoption expands. However, improvement remains incremental. Existing business models are rarely questioned. In the paper’s own words, organizations at this level “become increasingly efficient while remaining fundamentally unchanged” (Kadowaki, 2026b). Three things travel with the ERMM whenever it is used. First, progression is not linear. Organizations frequently display characteristics from multiple levels simultaneously; a firm may possess Level 4 AI capability while remaining Level 2 in leadership. The model evaluates organizational coherence rather than isolated excellence. Second, maturity is assessed across all five dimensions, in balance. Organizations with exceptional technological capability but weak leadership redesign cannot achieve higher maturity. Third, the objective is not reaching Level 5 as rapidly as possible. Different industries may require different levels of organizational adaptability. AI speeds the arrival at Level 2. Because the gains from efficiency show up as numbers, management reads them as success. While the sense of success lasts, no need for redefinition is felt. From the standpoint of Future Value this is a decline. Efficiency defends present enterprise value. Future Value does not come from there. Route 3 — convergence The third route advances most quietly. Many companies use the same foundation models. They feed them similar internal data. They evaluate against the same benchmarks. And they ask similar questions. The outputs converge. Conclusions about the market converge. Strategic options converge. Priorities converge. Competitive advantage is difference. When everyone reaches the same answer with the same tool, the difference disappears. Only one place is left where difference survives. The question side. What to ask, which assumption to doubt, which future to take on. This part is not standardized. It is not standardized because it is the territory of responsibility. Put the other way: a company that does not work on Question Design moves closer to the average the more it uses AI. Moving toward the average is, in Future Value terms, a retreat. Future Value is the capability to create value society does not yet have. It is not to be found among the answers everyone reaches. What separates the adding side from the subtracting side The three subtracting routes have something in common. None of them has anything to do with AI’s performance. Short-term optimization happens because people did not set the objective function. The tilt toward efficiency happens because people did not change the criterion for capital allocation. Convergence happens because people did not design the questions. The decline, in other words, was not caused by AI. AI amplified the absence of human design. The indicator an executive should watch is therefore neither the AI adoption rate nor the number of AI users. It is what the time AI returned was spent on. If that time vanished into more of the existing work, AI Capital has grown and Future Capital has not. If it went into re-examining assumptions and designing the future, both have grown. Two companies with identical adoption rates part company here.

6 Questions for the executive

We close on the assumption that AI will keep advancing. Judging by the trend as of August 2026, its capability is likely to continue improving. Neither the speed nor the end point can be predicted with confidence; we do not assert them. That is exactly why management should design in one particular way. Design so that nothing depends on where AI’s capability line sits. Three stances follow. First, invest in the terms AI is multiplied by. AI’s performance is supplied by the market. It rises whether or not you act. What does not rise on its own is Human Capital, Trust, Purpose, and Ecosystem. These can be grown only inside your own enterprise. When the AI budget goes up, check whether the budgets for these terms went up with it. If they did not, the product as a whole will not grow. Second, decide where responsibility sits before you extend the scope. When you widen AI’s field of application, decide who carries the results in that field before it goes live. This is not for the sake of control. A decision domain with no responsible person cannot hold a Purpose. Without a Purpose, no Future Value comes out of that domain. Third, treat questions as an asset. The only place where convergence can be avoided is the question. Who is asking which question, and has it changed since last year? Design the renewal of questions as a standing organizational practice. On that basis, three questions. Question 1 — What share of your AI budget goes to something other than making the existing business more efficient? If that share has stayed low for several years, the enterprise has designed itself to remain at Level 2. Establish whether that is deliberate or whether the investment criterion is producing it. The two call for different responses. Question 2 — Where is the time AI returned going right now? Check it function by function. Did throughput rise, or did thinking time rise? If only the first, Future Value has not increased. Question 3 — Are the questions you put to AI different from the questions your competitors put to theirs? If they are not, the outputs will not differ either. The only thing that can make a difference is your own view of your business and of society. None of the three questions asks about AI’s performance. All three ask what human beings have placed around it. AI changes Future Value a great deal. In both directions. What sets the direction is not AI. It is us. AI Capital is one term in a product. What turns it into Future Capital is the terms on the human side. And the only subject that can stand at the origin of the chain is one that can carry the result. However far AI advances, that position does not fall vacant. It is the executive who fills it.

In brief

AI both increases and decreases Future Value. What sets the direction is not AI but people.
AI Capital is only one term in a product. It becomes Future Capital only in combination with the others.
AI cannot take on responsibility. So it cannot stand at the origin of the chain, where Purpose sits.
When the AI budget rises, confirm that the budgets for people, trust, and purpose have risen with it.

Key concepts

Future Capital / Future Value / Future Value Chain / Human-on-theLoop Management / Question Design

The chain of ideas

Purpose → AI × Human → Future Capital → Future Value → Enterprise Value

Related first principles

Principle 4 — AI Optimizes. Humans Define. Principle 2 — Future Value Precedes Enterprise Value. Principle 3 — Capital Exists to Create Possibility.

Related chapters

Vol. III, Ch. 023 “What Is Future Value?” — the definition of Future Value this chapter assumes
Vol. I, Ch. 009 “How Should Work Be Divided Between AI and People?” — the prototype of the separation of roles
Vol. IV, Ch. 036 “What Is Management That Creates Future Value?” — why “creating” cannot be delegated to AI
Vol. IV, Ch. 033 “Are Intangible Assets Future Value?” — the eight forms of Future Capital and the boundary of accounting

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, #052 “What AI Cannot Create Will Become Your Company’s Value” / #007 “Can AI Predict the Future?”

Chapter 031 How Does AI Change Future Value?