Beyond The Gold Rush.

There is a familiar story about technological revolutions when everyone rushes to find the gold, the safest business is sometimes not to become the best miner, but to sell the picks and shovels.

The metaphor has survived because it captures an enduring strategic truth. When the outcome is uncertain, it can be wiser to build the infrastructure that enables many possible winners than to bet everything on identifying one.

In the age of AI, the metaphor can be taken even a step further. The most interesting position may not be to sell the shovels. It may be to build the rails.

The miner is exposed to the question: „Where is the gold?” The railway is exposed to a different and potentially more powerful question: „How much economic activity will the search for gold generate?” The first requires prediction. The second is about positioning. The difference may define a new approach to transformation, resilience and growth.

The corporate playbook of the last several decades was shaped by a world in which globalization, technological progress and relatively stable institutions reinforced one another.

The objective was to remove redundancy:

• Why hold excess inventory if a supplier can deliver tomorrow?
• Why maintain several suppliers if one global supplier can produce at lower cost?
• Why duplicate capabilities across geographies if one center of excellence can serve the entire organization?
• Why maintain multiple technologies if one standard can simplify the architecture?
• Why keep capital available when it can be put to work?

These questions made perfect economic sense. Until the assumptions underneath them began to change.

The pandemic demonstrated how quickly a local disruption could become a global economic shock. Wars and geopolitical tensions have subsequently exposed the fragility of energy, logistics and critical-material dependencies. Extreme weather increasingly affects physical infrastructure and agricultural systems. Cyber ​​threats have become systemic rather than merely operational. At the same time, AI is introducing a technological discontinuity whose consequences are difficult to forecast even for the companies investing most aggressively in it.

The important point is not that the world has suddenly become risky. The world has always been risky. The difference is that the risks are increasingly correlated.

Geopolitical tensions, trade fragmentation, energy insecurity, climate events, technological discontinuities, cyber threats, demographic shifts and financial volatility are no longer independent risks sitting in separate columns of an enterprise risk register. They interact.

A geopolitical conflict can disrupt shipping, which affects energy prices, which feeds inflation, which changes monetary policy, which alters investment decisions. A climate event can interrupt a semiconductor supply chain. A technological breakthrough can make an established business model obsolete faster than a traditional investment cycle can respond.

This changes the strategic question. It is no longer enough to ask: „What is the probability of this risk?” The more important question is: „What happens to the system if several risks occur at once?” That is where conventional risk management starts to run out of road.

The world has become a system in which shocks travel. And if shocks travel through systems, resilience cannot simply mean protecting individual assets. It must mean designing the system so that it can absorb disruption, adapt and continue creating value.

That is a very different ambition. It moves resilience from the margins of risk management into the center of company building. It also creates a strategic challenge for companies optimized for a single expected future.

What happens when the future refuses to cooperate?

The railway offers a different answer. It was not simply a more efficient way of transporting gold, but an infrastructure for uncertainty:

• It connected places whose economic potential was not yet fully known.
• It allowed people, goods, capital and information to move.
• It enabled multiple businesses to emerge.
• It created economic activity that would have been impossible without the network.

Most importantly, the railway did not need to know which miner would succeed. Its economics improved as the ecosystem grew.

A resilient company in the twenty-first century should aspire to something similar. It should not necessarily build its strategy around predicting precisely which technology, supplier, market or business model will win. It should build capabilities that allow it to benefit from several possible outcomes.

Every technological discontinuity creates a strategic hierarchy. At the visible end are the companies building the breakthrough products: the miners, in the old metaphor. They take the biggest bets, attract the biggest attention and potentially capture extraordinary returns. But underneath them sits another economy: the infrastructure that allows the entire ecosystem to function.

During the gold rush, the most interesting question was not necessarily which miner would strike gold. Someone had to sell the picks, shovels and transport. The infrastructure provider did not need geological clairvoyance. It needed to be positioned where economic activity was likely to increase.

AI makes this logic considerably more powerful. The modern equivalent of the railway is not a single piece of infrastructure. It is an interconnected system of models, data, compute, software, applications, identity, payments, workflows and transactions. As these layers proliferate, value does not simply accumulate at the top. It travels through the system. And that creates a different kind of strategic opportunity.

A company does not necessarily have to own the best model. It can benefit from the fact that there are many models. It does not necessarily have to invent the winning application. It can benefit from the fact that there are many more applications. It does not necessarily have to predict which AI agent becomes dominant. It can benefit from the fact that more agents are interacting, transacting and generating economic activity.

The strategic advantage therefore shifts from prediction to participation. Better still, it shifts from participation to positioning. This is why resilience, growth, and transformation should increasingly be considered together, not as separate ambitions, but as parts of the same strategic equation.

Transformation is often presented as the pursuit of a better future: better performance, new capabilities, new markets, new ways of creating value. Growth is the tangible expression of that ambition, the reason we transform in the first place. But growth without resilience can be surprisingly fragile, while resilience without growth can become little more than a sophisticated way of standing still.

Resilience asks a more demanding question: Can we remain viable and continue to grow, if the future turns out differently from the one we expected?

The strongest transformation strategy should therefore do more than optimize the business for the future we can see. It should strengthen the organization’s ability to perform under the expected future while increasing its capacity to adapt, absorb shocks, seize opportunities, and create value in the unexpected one.

In that sense, transformation is not simply about getting from A to B faster. It is about building an organization capable of changing the direction when the map changes, and still finding a way to grow.

That is the beginning of what might be called a resilience economy.And that is precisely what rails provide. A railway does not prevent every disruption. Tracks can be blocked. Routes can fail. Stations can close. But a well-designed network can provide alternative paths. The resilience comes not from eliminating uncertainty, but from creating options within it.

There is a moment in the life of almost every ambitious company when growth changes character. At the beginning, growth is something the organization must actively manufacture. It requires persistence, persuasion, capital, long hours, constant problem-solving and a remarkable tolerance for friction.

Customers have to be won. Processes have to be invented. Talent has to be attracted. Suppliers have to be convinced. Reputation has to be earned. Every additional unit of growth seems to require another unit of effort.

The young company pushes. The transformed company pulls. And the truly resilient company does something more, it keeps moving when the environment tries to stop it. That distinction captures something deeper than the difference between a start-up and an established business. It describes a transformation in the underlying economics of growth.

At some point, if the organization has built the right system, its customers, reputation, data, capital, talent, distribution networks and ecosystem begin to reinforce one another. Growth becomes less dependent on heroic effort and more dependent on the momentum of the system itself. The business becomes a train. And the most interesting consequence of that transformation is not simply that the train moves faster. It is that the train becomes harder to stop. That is where growth and resilience meet.

There is, of course, a temptation to view resilience as a defensive concept: insurance policies, contingency plans, emergency inventories, backup suppliers, and crisis committees. These are important, but they are not resilience in its full strategic sense.

Resilience is not a bunker. A resilient company is not simply one that survives bad things. It is one that retains the ability to act when bad things happen.

A fragile business is forced by circumstances. A resilient business retains choices. It can wait. It can invest. It can experiment. It can say no. It can absorb a mistake. It can change direction. It can survive a temporary decline without abandoning its long-term strategy.

In this sense, resilience is not merely the ability to withstand disruption, but the ability to preserve strategic freedom under pressure. And choice may be one of the most valuable assets a company can possess in an uncertain world.

Transformation, therefore, should not be understood simply as the process of making a company bigger, more digital or more efficient. Its deeper purpose is to increase the organization’s capacity to respond, adapt and continue moving.

The destination is not immunity. No serious company can promise immunity from disruption. Nor is it invincibility. The more useful objective is shock absorption.

Growth is possible but often labor-intensive. Sales are powered by individual relationships rather than a repeatable engine. Operations lean heavily on a handful of indispensable people. Growth itself often depends on the next round of fundraising. Institutional knowledge sits in employees’ heads instead of in systems. Distribution remains expensive, brand recognition is still a work in progress, and suppliers have little incentive to move the company to the front of the queue. Meanwhile, customers can be remarkably non-committal (they arrive with relative ease and can leave just as quickly).

In other words, the business can grow, but much of its growth is still being carried by people, capital, and goodwill rather than by durable infrastructure. The opportunity is real; the challenge is turning effort into leverage.

Transformation changes the physics. A transformed business develops assets and systems that reinforce one another. A strong brand lowers customer-acquisition costs. A larger customer base generates better data. Better data improves products. Better products strengthen reputation. Reputation attracts talent. Talent improves execution. Scale strengthens bargaining power. Better distribution expands reach. Greater reach attracts partners. Partners strengthen the ecosystem. The ecosystem creates additional opportunities for growth.

The individual components matter, but the real strategic asset is the connection between them. This is when growth begins to acquire momentum. The company no longer has to manufacture every increment of progress from scratch. It has built a system that helps produce the next increment. That is transformation.

And yet size is not automatically resilience. Sometimes scale creates the very vulnerabilities it claims to protect against.  A company with €1 billion in revenue concentrated in one product, one customer segment and one geography may be more fragile than a €200 million company with diversified customers, strong cash generation, multiple suppliers and highly adaptable operations.

So perhaps we should stop asking: „How big is the company?” and start asking: „How many independent ways does this company have to survive?”

That is a much better measure of strategic resilience:

• Can it survive losing its largest customer?
• Can it survive a 30% decline in one market?
• Can it replace its most important supplier?
• Can it finance itself during a credit contraction?
• Can it retain its best people during a downturn?
• Can it change its technology stack?
• Can it continue investing when competitors stop?
• Can it absorb a regulatory shock?
• Can it survive six bad quarters?

The answers reveal something revenue figures cannot. They reveal the company’s survival architecture.

A resilient large organization should be capable of decomposing problems, shifting resources and isolating failures without bringing down the entire system. Scale creates power. Architecture determines whether that power is robust. This is where the architecture of the company becomes as important as its size.

This changes the way we should think about diversification.

Diversification is often criticized because it can sacrifice efficiency. Maintaining multiple suppliers, production sites or technologies can cost more than concentrating activity around the most efficient option.

But the railway metaphor reveals the strategic value of alternatives. The railway does not need every route to be used every day. The value of a network partly comes from the fact that routes exist when they are needed.

The same is true of a resilient company:

• A second supplier may appear redundant until the first becomes unavailable.
• A second production location may appear inefficient until a geopolitical event closes the primary one.
• A modular technology architecture may appear unnecessarily complex until a new technology makes the existing system obsolete.
• A multi-model AI architecture may appear less elegant than committing to a single provider, until that provider becomes too expensive, technically inferior or strategically misaligned.

The company is buying the ability to move. And in an increasingly uncertain world, this has economic value.

The mistake is to interpret diversification as simply adding duplication everywhere. That would produce an expensive and bureaucratic organization rather than a resilient one. More is not automatically better. An organization can diversify itself into mediocrity. Too many markets, products, suppliers, technologies and priorities can dilute managerial attention and destroy the advantages that scale was supposed to create.

The real objective is selective diversification that leads to intelligent optionality. The company should know which dependencies are strategically dangerous, which redundancies are economically justified and which capabilities must remain under direct control.

The principle is simple: Do not diversify everything. Diversify where dependency creates existential fragility.

AI makes the railway metaphor more powerful. The old railway moved physical goods. The AI-era railway can move something more powerful: intelligence, decisions, transactions and economic activity.

Imagine an ecosystem in which different AI models specialize in different forms of reasoning. Applications are built on top of them. Agents interact with those applications. Businesses connect their workflows. Customers initiated transactions.

Trying to predict which model will win is one strategy. Building a business that benefits from all models becoming better, cheaper and more widely used is another.

The second strategy has a peculiar elegance. It converts technological uncertainty from a risk into an input.

• If Model A wins, benefit.
• If Model B wins, benefit.
• If five models specialize in different tasks, benefit.
• If a new architecture emerges, integrate it.
• If inference becomes cheaper, benefit from increased usage.
• If AI applications proliferate, benefit from the increased activity.
• If agents start executing transactions, benefit from the expansion of the transaction layer.

The company is no longer betting on one technological future. It is building rails across many possible futures. It grows regardless of the scenario.

Of course, resilience costs money. So does building a railway: laying tracks, maintaining capacity, and keeping routes open even when demand does not justify every mile in purely economic terms. The tension is real. There is, however, a genuine trade-off between efficiency and resilience. A system optimized entirely for efficiency tends to eliminate redundancy, spare capacity, and alternative routes. A resilient system deliberately preserves some of them.

This is where orchestration becomes strategically important. A company does not need to own every piece of the system to make it resilient. It needs to shape the network around itself so that multiple applications, models, developers, and transactions can expand through it, and so that the company can capture value from that expansion.

Ownership creates control over individual assets, orchestration creates leverage across the system. That is a powerful form of strategic resilience. The distinction can be expressed simply:

Gold-rush logic| AI-era logical resilience:

• Find the winning miner| Enable many winners
• Predict where the gold is I Benefit from more gold being discovered
• Own the best mine I Own critical infrastructure
• Optimize one route I Create multiple routes
• Maximize efficiency I Optimize efficiency and optionality
• Defend against disruption I Design for adaptation
• Growth comes from the company I Growth increasingly comes from the ecosystem
• Scale the business I Scale the network

Maintaining optionality often comes at the expense of short-term efficiency. Using multiple suppliers can reduce purchasing power. Holding additional inventory ties up capital. Diversifying production across geographies can sacrifice economies of scale, while designing modular systems can increase upfront costs.

But these trade-offs raise a broader question: Are traditional measures of efficiency capturing the full economic cost of fragility?

• A supply chain that saves two percent in normal conditions but loses forty percent of revenue during a major disruption is not necessarily more efficient.
• A technology architecture that is cheap until it becomes impossible to change may have hidden switching costs.
• A highly specialized workforce may be productive in stable conditions but vulnerable when demand shifts.
• A company with minimal financial buffers may maximize returns in the short term while losing strategic freedom when conditions deteriorate.

The economics of resilience are therefore about the value of options. A spare route has value. A second supplier has value. A modular technology stack has value. A workforce capable of redeployment has value. A balance sheet capable of acting during a crisis has value.

The challenge is to price those options intelligently. It is not just about minimizing costs, but about optimizing the economic aspects of the entire system across various possible scenarios. And once we think this way, resilience becomes measurable in a different language.

• Not merely: „How much does this redundancy cost?” But: „What is the value of the options it creates?”
• Note: „How much inventory can we remove?” But: „What is the economic value of continuity?”
• Note: „Why maintain multiple suppliers?” But: „What is the value of being able to switch?”

This marks the beginning of a much more sophisticated economics of resilience, and it is precisely where AI can become particularly important. It can analyze more information, monitor more dependencies, simulate more scenarios, personalize interactions at scale, automate parts of decision-making, translate information into action. The most interesting contribution of AI to transformation may not ultimately be automation. It may be making organizational flexibility economically viable.

The old corporate choice was often: efficiency or flexibility. AI creates the possibility of: efficiency through flexibility.

That is a profound shift. If intelligence becomes scalable, companies can afford to maintain more optionality because the cost of managing that optionality falls. And once the cost falls, resilience stops looking like an insurance premium. It begins to look like a growth capability.

A resilient company must be capable of changing without destroying itself. That requires flexibility at several levels.

• Financial flexibility means having enough liquidity and balance-sheet capacity to make decisions when others are forced to sell, cut or retreat.
• Operational flexibility means being able to shift production, suppliers, distribution or capacity.
• Technological flexibility means avoiding architectures that make innovation prohibitively expensive.
• Organizational flexibility means moving talent and decision-making towards the problems that matter most.
• Strategic flexibility means being able to change direction without losing the identity and capabilities that make the company valuable.

These capabilities rarely appear spontaneously during a crisis. They have to be built in advance. That is why resilience belongs inside transformation.

When the train becomes harder to stop. At the beginning, transformation is painful precisely because the organization has to push the boulder.

Processes must be redesigned. Technology must be replaced. Incentives must change. People must learn new ways of working. Capital must be redirected. Old assumptions must be challenged. Some successful practices must be abandoned because they belong to a business model that is no longer sufficient.

Transformation requires enormous effort because the organization is trying to change its own physics. But if it succeeds, the economics change: customers begin to pull, reputation begins to pull, data begins to pull, talent begins to pull, capital begins to pull, distribution begins to pull. And so the enire ecosystem begins to pull…eventually, the company acquires momentum.

A moving train is harder to stop than a stationary one. This is the real achievement of transformation. Not simply acceleration. Resistance to interruption.

A company with momentum has accumulated capabilities that allow it to continue moving even when individual components fail. It has buffers, alternatives, relationships, information and trust. It can absorb a shock without losing its strategic direction.

It can bend without breaking. This is why the distinction between growth and resilience is becoming increasingly artificial. Growth and resilience are becoming the same conversation.

For a long time, companies treated growth as offensive and resilience as defensive. Growth meant entering markets, increasing revenues, gaining customers and investing for the future. Resilience means reducing exposure, controlling costs and preparing for bad scenarios. But the strongest forms of growth increasingly accomplish both objectives at once.

• A better distribution network can increase sales and reduce concentration risk.
• A stronger balance sheet can fund expansion and provide shock absorption.
• Better data can improve customer experience and make demand shifts visible earlier.
• A broader talent pipeline can support growth and reduce dependence on a few critical individuals.
• A diversified supplier base can increase flexibility and create new commercial opportunities.
• A stronger ecosystem can expand the addressable market while creating additional routes around disruption.

The same investment can therefore serve two strategic purposes. It can make the company bigger and harder to break. That is a far more interesting definition of growth.

From risk management to strategic capability. The most mature organizations will therefore move resilience out of the defensive corner of the enterprise. It should influence strategy, capital allocation, technology architecture, product design and operating models.

A resilient company asks: Where are we overly dependent? Where are we unable to switch? Where are our critical bottlenecks? Which capabilities are difficult to rebuild? Which suppliers could become systemic dependencies? Which technologies could lock us into a particular future? Where could a local failure propagate through the organization?

And, crucially: Where could greater optionality create new growth?

This last question matters because resilience can create opportunities precisely when competitors are constrained:

• A company with flexible capacity can serve demand after a competitor’s disruption.
• A business with diversified sourcing can continue operating while others face shortages.
• An organization with modular technology can adopt a breakthrough faster than incumbents locked into legacy architecture.
• A company able to integrate multiple AI models can benefit from technological progress without having to make a single irreversible bet.

Resilience therefore has an offensive dimension. The ability to adapt can become a competitive weapon.

The transformation paradox: stop trying to control the future. This leads to a deeper definition of transformation. Traditional transformation often seeks greater control. AI-era transformation may require something more subtle: greater adaptability.

The objective is not to build an organization that has perfectly predicted what the market will look like in five years. That is increasingly impossible. The objective is to build an organization that can change direction without losing momentum. This means designing systems that are modular rather than monolithic:

• Platforms rather than isolated products.
• Interfaces rather than proprietary silos.
• Multiple suppliers rather than single dependencies.
• Multiple technology options rather than irreversible commitments.
• Data that can travel rather than data trapped in applications.
• AI architectures that can incorporate new models rather than architectures designed around yesterday’s winner.

The strategic lesson is therefore larger than the familiar gold-rush metaphor. The objective is not merely to sell picks and shovels. It is to build rails that remain valuable whichever miners succeed.

That means designing companies that can benefit from technological progress without needing to predict exactly where it will come from. It means supply chains that can reroute. Technology architectures that can switch. Organizations that can redeploy. Platforms that can incorporate new participants. Business models that benefit from ecosystem growth. And AI systems that make complexity manageable rather than merely making existing processes faster.

The railway does not know where the next mine will be. It doesn’t need to. Its strategic advantage comes from connecting the places where economic activity might emerge.

That is the deeper lesson for transformation. In a world of interconnected geopolitical, economic, technological and environmental risks, the winning company may not be the one with the most accurate forecast. It may be the one with the best architecture for being wrong. And that is what makes the metaphor of the rails so powerful.

The miner needs to be right about the future. The railway needs the future to keep happening. The miner is betting on the destination. The railway is betting on movement. And in the age of AI, where models will evolve, applications will proliferate, transactions will multiply and entire industries will be rewired, movement may be the more durable source of value.

The strategic imperative is therefore not simply to transform. It is to transform in a way that creates optionality, resilience and compounding.

Better still, build a system in which more trains, more passengers and more destinations make the rails increasingly valuable. That is when transformation stops being a program, it becomes a property of the business. And that is when growth stops being something the company has to chase and starts becoming something the company has designed to happen.

***

There is one final reason resilience matters: transformation does not guarantee wisdom.

A resilient company can still make bad decisions. It can still misread markets, invest badly, hire the wrong people or pursue the wrong strategy. Resilience does not eliminate errors. What it changes is the consequence of error.

A fragile company may make one bad decision and be forced into another simply because it has no room left. A resilient company has something far more valuable: time. Time to learn. Time to reconsider. Time to experiment. Time to change course without tearing the whole ship apart.

Transformation creates that room. It builds the financial, operational, and organizational capacity in which judgment can actually operate. You can wait instead of react. Experiment instead of betting everything. Invest through a downturn instead of simply trying to survive it. Walk away from an attractive opportunity because, strategically, it isn’t the right one. Change direction without having to change your entire identity.

In uncertainty, these are not minor advantages. They are forms of power.

If you feel this is a conversation worth having together, let me know. Either way, keep it handy!