Purpose, the Foundational Principle and Creating Possibilities

Quick Summary

Purpose is our choice. Assumptions about reality need testing. A foundational principle must change investments and priorities. Capabilities and partnerships allow direction to become action.

In an age of uncertainty, we do not know what will happen. We must decide what we want to create and build the capacity to act. Purpose and the foundational principle give the creation of possibilities a binding direction.

A company must decide what kind of company it wants to be. What does it want to achieve? Why does it want to exist? What does it want to be when it grows up?

I begin strategic thinking with these questions. The answers define the purpose the company chooses for itself. From that purpose, we clarify the foundational principle that will guide it. Only then do we create possibilities for putting it into practice.

This inquiry requires us to shape the organization we want to build. It will determine what we invest in, which capabilities we develop, with whom we collaborate and what we give up. When it does not change decisions, it remains a statement.

My experience with Dani Ziv at Plasan Sasa, and the thinking presented by Jensen Huang, NVIDIA’s founder and CEO, allow us to examine how this sequence works in practice. In both cases, capabilities were built that expanded what was possible. Partnerships enabled those capabilities to become economic activity.

Looking inward and choosing a purpose

A description of products and services explains what the company does today. Purpose defines what we want to achieve through it. Current activity is the result of earlier decisions. It cannot, on its own, determine what we want to become.

Even an established company must ask what it wants to be when it grows up. A company can increase sales while moving away from its purpose. Growth therefore requires a clear answer. What kind of company are we building through this growth?

The inquiry begins with those responsible for the company’s direction. What motivates us to devote our time and resources to it? What contribution do we want to make through it? What will change in customers’ lives if we succeed? And what will we want to look back on with pride?

We will distinguish what matters to us from the form we have imagined. If the company becomes large but the contribution we sought does not materialize, will we consider that a success? If the contribution can be made in a different way, will we choose it even when it requires profound change?

One owner may seek independence over many years. Another may want to build a company to sell. The difference will affect investments and decisions. It must be clarified, and a decision made about how to manage it. We will listen to employees and customers and examine actual outcomes. The responsibility for choosing the purpose remains ours.

Profit and returns sustain the company and enable investment in its future. We must clarify how our chosen contribution will create an economic foundation. Company size alone does not answer the question of purpose. Nor does a capability we currently lack end the inquiry. It defines work we will need to do to put our choice into practice.

Dani Ziv, Plasan and the capability that preceded demand

In the early 2000s, I was asked to lead a Management by Objectives process at Plasan. The strategy existed mainly between Dani Ziv’s ears. My role was to help translate the direction into a work plan that people in the company could understand and implement.

As I remember it, there were apple crates made of composite materials at the beginning. Later, a capability was built to apply composites to armor protection. Dani extended the use of the knowledge beyond the existing product. The material became a basis for protecting people and vehicles.

Capabilities in vehicle design and armor integration were then built. I remember the vehicle workshop and the people recruited to it. The challenge was to preserve the vehicle’s performance and stability while adding the weight required for protection. Meeting that challenge required understanding the vehicle as a system.

The following formulation is my interpretation today of the process I observed. The purpose was to protect people required to operate in dangerous environments. The principle guiding development was to combine their protection with their ability to perform the mission. This led to work on the armor materials and the vehicle together.

This principle has clear managerial consequences. It requires knowledge of materials, engineering, design and testing. It requires asking what added armor does to the entire vehicle. And it requires investing in a capability that is not yet a mature product.

When the wars in Iraq and Afghanistan sharply increased demand, capabilities already built at Plasan had an opportunity to develop on a large scale. I do not attribute a forecast of these wars to Dani. The lesson I take is that preparing the capability expanded the possibilities the company could realize when demand changed.

Modular solutions were later developed for bolted assembly, integration into vehicle manufacturers’ production lines, and field maintenance and component replacement. [8] The solution expanded the possibilities for manufacturing and supplying armor. A partner without all of Plasan’s knowledge could also participate in realizing it.

A range of possibilities was built at Plasan. Many were never realized. Some generated knowledge that contributed to the next initiative. Others remained directions for which sufficient justification was not found. To understand the success, we must remember those possibilities too.

Jensen Huang and the contribution that justifies a company’s existence

Huang describes NVIDIA’s original purpose as building computers to solve problems that ordinary computers struggle to solve. He also proposes a test for choosing tasks. When others can do the work, they can be allowed to do it. The company should choose important, difficult work that it has reason to believe will not be done without it. [1]

I distill from this a conception of purpose. The company seeks to expand the limits of what the world can do. To fulfill its purpose, it must identify the bottleneck and build the capability that makes it possible to overcome it.

At NVIDIA, the bottleneck is the computing capability available, including the time, cost and energy needed to solve problems. Huang presents accelerated computing as a way to expand the use of computing within the constraints of the existing approach. [4]

I distinguish two things here. Purpose is our choice about what deserves to be created. The assumption about the bottleneck is a claim about reality. That claim must be tested. Commitment to a purpose does not exempt us from testing assumptions.

The foundational principle therefore requires looking inward and testing outside. What do we want to make possible? What prevents it today? What can we change? And why is our particular contribution needed?

The foundational principle must change decisions

From our purpose, we formulate the foundational conception on which we will build the company. It determines what will be central to decisions, and what we must justify whenever we act differently.

At Plasan, connecting protection and performance required broader knowledge. Expertise in materials alone was insufficient for the chosen mission. An understanding of the vehicle and an ability to integrate the fields were also needed. The principle influenced the people recruited and the work that was built.

At NVIDIA, expanding the limits of computing also required software. CUDA is a programming environment that enables developers to use graphics processors for many computing tasks. Huang describes how the investment initially added costs to chips while applications that would justify those costs for customers were still lacking. [2]

The test of a principle is an actual decision. Which investment will we advance because of it? What will we prioritize when resources are limited? What revenue will we forgo to build the capability we need? And which existing activity will we need to change?

If every decision fits the principle, the principle needs greater precision. If it does not change a choice, it remains a statement. It gains managerial meaning when we act on it, especially when doing so comes at a cost.

Creating possibilities and building the conditions for their realization

Purpose defines the direction. The principle defines the commitment and the bottleneck to be examined. Possibilities are the courses of action we create to put them into practice. They can include products, new capabilities and relationships with partners.

Huang talks about markets worth zero billion dollars today. He seeks to create economic activity that does not yet exist. The absence of revenue today does not prove there is no need. Sometimes, the capability to meet it is missing. [2]

In his Caltech address, he said: “If we don’t build it, they can’t come.” [3] I read this sentence as an obligation to act. If future customers depend on a capability that does not yet exist, we must clarify what needs to be built so that they can act.

However, the absence of a market is not proof of an opportunity. The need may be weak, or the cost of meeting it too high. We must therefore explain what will have value for the user, what must change for the activity to become possible, and how we will test the explanation before expanding the investment.

This is how I understand the process at Plasan. Knowledge of materials enabled the development of armor panels. Knowledge of vehicles enabled exploration of the integration of protection and performance. Modular solutions also enabled different methods of assembly and maintenance. Each capability added courses of action the company could examine.

A new possibility also emerges when we change the unit of design. Instead of examining the armor panel alone, we examine the vehicle. Instead of examining the chip alone, we examine the computing system. This reveals bottlenecks that improving one component does not resolve. And the connections between components can be redesigned.

The GTC keynotes show a consistent expansion of the system being designed. In 2022, hardware, software, simulation and AI factories were presented together. In 2023, broader access through cloud partnerships was added. In 2024, the connection between computing power and running models was emphasized. [4] In 2025, software for managing the work of an AI factory was also presented. [5] The technological capability was connected to the conditions required to operate it.

An ecosystem that enables others to succeed

Realizing a possibility sometimes requires an entire organizational network around a capability. One participant develops applications, another supplies infrastructure, and another brings professional knowledge or access to customers. The ecosystem is the system of relationships among them. It must give every participant a reason to keep investing.

Huang describes NVIDIA as a company that combines vertical integration internally with horizontal integration that opens it to partners at every layer. How?

It designs chips, systems and software together, and allows its capabilities to be integrated into others’ platforms. [6]

The connection between developers and cloud providers illustrates the logic. Developers create applications that require computing infrastructure. Cloud providers can supply it and earn revenue from its use. Huang notes that CUDA developers are also customers of the cloud providers. The developer community is therefore part of the value he brings to a partner. [5]

In 2025, NVIDIA introduced NVLink Fusion. The initiative allows partners to combine their own processors and custom chips with NVIDIA infrastructure. [7] I see this as an important strategic choice. The company allows others to participate in some of the components, while seeking to expand its role in the system that connects them.

At Plasan, too, partnerships with vehicle manufacturers such as Oshkosh connected armor capabilities to platforms and large scale production. Kits that can be assembled on the partner’s production line made collaboration part of the design of the solution itself. [8] A strong engineering capability became easier to integrate into another company’s system.

The two companies differ greatly in the scale and structure of their activities. Plasan also develops its own vehicles. NVIDIA operates in some of its partners’ fields. I therefore do not claim that they always avoid competing with partners. In both, I identify the power of a complementary capability that enables others to deliver a better outcome to their customers.

From this, I derive a test for building an ecosystem. What new possibility are we opening for the partner? How will the partner profit from realizing it? Which part of the work must the partner perform, and which part must we perform? And what will make the customer prefer the joint outcome?

An ecosystem can deepen competitive advantage. A competitor must contend with a technological capability, as well as knowledge, applications and partnerships already working together. But collaboration does not eliminate competition. NVIDIA itself warns that customers and partners may develop their own solutions or buy from competitors. [9]

In my view, the aim is to make collaboration with us worthwhile and our contribution difficult to replace. To preserve that position, we will need to keep creating value. Dependence created in the past is not a commitment by partners to the future.

Invest in capability and know when to change direction

I remember a request for a separate profit and loss statement for the vehicle workshop. As an economist, I thought the request missed its stage of development. Its cost needed to be examined, as did its learning and the capability it was building.

When building a new capability, revenue in the current period is not enough to determine whether the investment is progressing. We must define which assumption we tested and what we learned, what new performance we achieved, and what commitment from a user or partner would justify continuing.

However, purpose is not a license for unlimited investment. Before each stage, we will establish what we want to learn, what it will cost and what would justify expansion, a change or a stop. We will also examine the value of the knowledge created and the possibility of using it in another direction.

Plasan Through the Possibility Card

We can return to the Plasan story through the card. This is my retrospective analysis of the case I described. The desired outcome is protection that allows the mission to continue. The approach is to integrate armor into the vehicle while preserving its performance. The condition to be tested is the performance of the whole system after integration.

The required capabilities include materials, vehicle design, integration and testing. People and partners connect the fields of knowledge and link development with production. Investment in the vehicle workshop must be assessed against performance and learning, and against the resources that further work will require. The demand that emerged later illustrates how the capability was put to use.

The card also raises questions about a similar decision today. What new performance has been achieved, and what is still missing? What will the next test cost, and what would justify it? What evidence would require us to change the design or stop? These are questions I propose for a manager, not a record of a procedure used at Plasan.

Huang says that he repeatedly reexamines the principles behind an investment. Have conditions changed? Has someone already solved the problem? And is the bottleneck still as important as we thought? [2] I adopt this examination as a managerial obligation.

When the principle remains valid, we can change the way we put it into practice. When an assumption it rests on has been disproved, we must decide afresh. Our commitment requires us to learn even from evidence we find uncomfortable. This is how we distinguish justified persistence from costly stubbornness.

When to Narrow Possibilities and Commit

Creating possibilities does not require keeping all of them open. Management time, people and capital are limited. Sometimes a partnership, development or market entry requires a commitment, even at the cost of closing other directions.

We will examine what a possibility enables and what it costs to keep it available. What does it prevent us from doing now? What would it take to rebuild the capability if we gave it up? And when must we decide to avoid losing the path to realization?

We will preserve a capability when the cost of maintaining it is justified. We will commit when progress requires concentrating resources. We will define what might change the decision and when we will review it. Even a justified commitment provides no knowledge of what will happen.

Turn the inquiry into strategic work

Management must emerge from the inquiry with explicit choices. First, it will define what it wants to make possible in the world and to whom that contribution matters. It will then formulate the principle that will guide its decisions and clarify which assumptions about reality require testing.

From this, it will create different courses of action. Each possibility will include the desired outcome, the capability that must be built and the partners required. We will examine who receives value, who pays, and what will make the activity sustainable for those involved.

We will now allocate resources to an experiment that can change a decision. We will examine the bottleneck that could prevent realization. Test an engineering capability, user behavior or a partner’s willingness. Choose the test according to the assumption decisive to the possibility. The experiment will provide a basis for continuing, changing or stopping. It does not provide knowledge of the future.

There are differences among the kinds of thinking needed here. Optimization improves performance within a defined framework. Discovery reveals knowledge or a connection we did not know. Creating possibilities builds new directions for action that can be developed. Management needs to know which kind of work is required to move forward.

AI can be a partner in this thinking. We will present it with the purpose and principle we have chosen. Ask for different ways to put them into practice and the assumptions those approaches rest on. Require it to examine the weaknesses of our preferred approach too. We will take the proposals back to reality and test them with the people we want to serve.

Learning together also requires sharing the reasoning. Employees need to understand why an investment was chosen, what it is meant to make possible and what could change the decision. They can then contribute to the inquiry and create further possibilities.

Learning by Thinking Together

The future belongs not only to those who know how to solve problems, but to those who know how to create possibilities we have not thought of before. The strategic responsibility is to decide which of them we will build, and how we will enable others to create additional value through them.

In an age of uncertainty, a foundational principle must open different courses of action. Dependence on a single path leaves the company exposed to change it cannot address. Management is therefore committed to creating possibilities from the principle and building the conditions that will enable their realization.

A question for the reader

If your chosen purpose were already binding on you today, what capability would you start building, and for whom would it open a new possibility?

One Step into Practice

Choose an investment decision still to be made and examine it against the foundational principle. Explain what choice the principle requires and what you will give up as a result. Check with the team whether resource allocation reflects that choice.

Sources for further reading

1 Stanford 2024 A conversation with Jensen Huang on purpose and thinking from first principles Source

2 Stripe Sessions 2024 A conversation with Patrick Collison on creating markets and CUDA Source

3 Caltech 2024 Huang’s address on building capability before a market emerges Source

4 GTC keynotes from 2022 to 2024 The official record of the expansion of the computing system 2022 2023 2024

5 GTC 2025 Huang’s keynote on AI factories and the developer and cloud provider community Keynote transcript Official record

6 GTC 2026 Huang’s keynote on vertical integration and openness to partners Keynote transcript Official record

7 NVIDIA 2025 The introduction of NVLink Fusion for integrating partners’ chips and processors Source

8 A description of modular armor solutions and their integration into vehicle manufacturers’ platforms Plasan Plasan Plasan North America

9 NVIDIA 2026 The annual report includes the risk of customers and partners developing their own solutions Source