Chapter 9: The Great Infrastructure Squeeze

When Trillions of Dollars Chase a Finite World

Imagine you’re the CEO of a major technology company. You’ve just convinced your board of directors to approve a $20 billion investment in one of the largest artificial intelligence data centers ever constructed. Wall Street loves the announcement. Investors celebrate your company’s ambition. Analysts begin calculating the extraordinary revenues that might eventually flow from the enormous computing capacity you’re about to build.

You’ve purchased hundreds of acres of land, negotiated agreements for thousands of advanced AI processors and hired some of the brightest engineers in the world. The architectural plans are complete, the construction contracts are being negotiated, and everyone involved understands the strategic importance of finishing the project as quickly as possible.

Then your construction manager delivers some rather unpleasant news.

The local utility cannot supply the full electrical capacity you need on your preferred schedule. Several critical transformers have long manufacturing lead times. Your construction contractor is struggling to assemble enough qualified electricians and pipefitters. The cooling equipment supplier is juggling orders from several other technology giants. And the engineering firm responsible for coordinating the project warns that a delay in any one of these systems could push back the entire schedule.

You have $20 billion available to spend.

But there’s something your $20 billion cannot buy.

Time that doesn’t exist.

Welcome to what I believe will become one of the defining economic challenges of the artificial intelligence revolution.

The Great Infrastructure Squeeze.

And if you’ve followed our MoneyQuake series, you know why this matters. We’re witnessing a collision between the extraordinary financial resources of the digital economy and the physical limitations of the industrial economy.

Technology companies can announce new projects in an afternoon. They can raise billions of dollars in capital, sign contracts for semiconductors and unveil architectural renderings of enormous AI campuses.

But the physical world doesn’t operate at the speed of a press release.

Copper has to be mined. Steel has to be manufactured. Concrete has to be poured. Transformers have to be assembled. Turbines have to be delivered. Transmission lines have to be constructed. Electricians, welders, plumbers and technicians have to perform the work.

And every one of those activities requires something the technology industry increasingly seems unwilling to wait for.

Time.

The Most Important Word in Economics

Scarcity.

It’s one of the first concepts students encounter in an introductory economics course. Human wants are unlimited, but the resources available to satisfy those wants are limited. The entire discipline of economics is built around understanding how people, businesses and governments allocate scarce resources.

For generations, investors have associated scarcity primarily with commodities. When oil supplies tighten, prices rise. When copper demand outpaces mine production, copper prices can increase. When shipping capacity becomes constrained, freight rates can surge.

But something more complicated is happening today.

Scarcity is beginning to emerge across multiple layers of the same industrial supply chain at precisely the moment when technology companies, utilities and governments are planning enormous investments in infrastructure.

Consider what happens when a technology company decides to build a giant AI data center. Before the first server arrives, the developer needs land, roads, foundations, structural steel, concrete, electrical equipment, cooling systems, fiber-optic connections and specialized construction workers.

Then the facility needs electricity. That may require new generation capacity, substations, transformers and transmission infrastructure. Those projects create demand for many of the same materials and workers already needed at the data center itself.

And if the electricity comes from a newly constructed natural gas power plant, that project may require turbines, pipelines, compressors, specialized steel and additional construction crews.

One investment creates demand for another investment, which creates demand for another.

The result is an extraordinary industrial chain reaction.

The bottlenecks multiply throughout the economy.

That’s what makes the Age of Scarcity so important to investors. The constraint may not be a single commodity. It may involve several interdependent materials, components, manufacturing processes and categories of skilled labor simultaneously.

Copper: The Metal Behind the Machines

Let’s begin with one of the most important materials in the entire electrical economy.

Copper.

Copper is essential to modern electrical infrastructure because of its conductivity, durability and versatility. It’s used in wiring, motors, generators, transformers, electrical distribution systems and countless industrial applications.

As the world electrifies more of its economy, copper demand grows. Electric vehicles need copper. Renewable energy systems need copper. Transmission networks need copper. Industrial machinery needs copper. Data centers and the power infrastructure supporting them need copper.

But copper has a problem.

You cannot manufacture it out of thin air.

You have to find it, mine it, process it and refine it.

And that takes years.

The International Energy Agency’s 2025 Global Critical Minerals Outlook warned that announced copper mining projects could leave the world facing a significant supply gap by 2035 under its stated-policies scenario. The agency estimated that expected mined supply could fall approximately 30% short of projected primary copper requirements.

That isn’t a guarantee that the world will run out of copper. Higher prices can encourage additional investment, recycling can expand supply, and technology can reduce the amount of copper required in certain applications.

But the projection illustrates the enormous challenge facing the industry.

Major copper deposits aren’t discovered every afternoon. Exploration takes years. Environmental studies take years. Permitting can take years. Financing, engineering and construction require enormous capital commitments.

Even after a mine begins operating, production may take additional time to reach full capacity.

This creates one of the central contradictions of the AI economy.

Technology companies can announce billions of dollars in new infrastructure spending overnight. But they cannot announce a new copper mine into production.

That difference between the speed of capital commitments and the speed of physical supply is where the MoneyQuake begins to generate some of its most interesting opportunities.

Concrete: The Uncelebrated Foundation of Artificial Intelligence

Now let’s talk about something considerably less glamorous.

Concrete.

Nobody attending an artificial intelligence conference gets particularly excited about concrete. Investors would rather discuss neural networks, advanced semiconductors, artificial general intelligence and the extraordinary capabilities of the latest AI models.

But before those models can run inside a data center, somebody has to build the building.

And that usually begins with concrete.

Enormous data centers require foundations, slabs, equipment pads, roads, loading areas and other civil infrastructure. Depending on the design, they may also require substantial concrete walls and structural elements.

The power plants supplying those facilities need concrete. So do substations, transmission structures, cooling installations and other infrastructure.

But concrete introduces an interesting complication.

Unlike semiconductors, which can be manufactured in one country and shipped around the world, ready-mixed concrete is fundamentally a local product. It has a limited workable delivery window, making the location of batching plants, aggregate supplies, trucking fleets and construction crews extremely important.

A project might have access to abundant cement nationally and still encounter local limitations in concrete production or delivery capacity.

And when several enormous construction projects are underway in the same region, those limitations can become increasingly important.

That’s a distinction investors need to understand.

Scarcity doesn’t always mean the world is running out of a particular material.

Sometimes it means the material isn’t available in the right location, in the right form, in sufficient quantities, at the right time.

The same applies to structural steel, aggregates, specialized piping, electrical components and heavy construction equipment.

The world’s largest technology companies may increasingly find themselves competing not merely for GPUs and engineering talent, but for the same concrete trucks, construction materials and industrial suppliers.

The Electrician May Be More Important Than the Algorithm

Here’s a question I want you to consider.

What good is a $20 billion artificial intelligence facility if you cannot find enough qualified people to build and connect its electrical systems?

For decades, America encouraged young people to pursue careers in software, finance, law, medicine and other professional occupations. Those fields created enormous wealth and attracted generations of talented workers.

But the AI infrastructure boom is revealing the importance of another group of professionals.

Electricians. Plumbers. Pipefitters. Welders. HVAC technicians. Crane operators. Heavy-equipment operators. Construction supervisors. Industrial maintenance technicians.

These are the people who transform enormous capital budgets into functioning physical assets.

A data center cannot be completed by software developers alone. Someone has to install electrical distribution systems, connect equipment, construct cooling infrastructure, assemble piping and ensure that everything operates safely.

And these skills cannot be acquired overnight.

The U.S. Bureau of Labor Statistics projects that employment for electricians will grow approximately 9% between 2025 and 2035, considerably faster than the average for all occupations. The agency projects tens of thousands of electrician openings annually, including positions created by retirements and workers leaving the occupation.

Those projections describe the broader electrical workforce rather than an AI-specific labor shortage, but they illustrate why recruiting and training skilled workers will become increasingly important as industrial construction expands.

Now imagine several multibillion-dollar data center campuses, semiconductor factories, power plants and transmission projects competing for experienced electricians in the same regional labor market.

You can see the potential problem.

Even if enough workers exist nationally, they may not be located where the projects need them. And even if companies offer higher wages, specialized skills still require training, experience and certifications.

A shortage of qualified labor can delay construction just as effectively as a shortage of copper or transformers.

In the AI economy, the electrician may become just as strategically important as the software engineer.

The Great Transformer Bottleneck

If I had to choose one piece of equipment that symbolizes the challenges facing the AI infrastructure boom, it might be the electrical transformer.

Transformers are essential because electricity has to move between different voltage levels as it travels from power plants through transmission networks and eventually into data centers, factories, businesses and homes.

Without transformers, the modern electrical grid cannot function.

Yet large transformers are not ordinary products that can be manufactured overnight. They require specialized engineering, electrical steel, copper, insulation systems, manufacturing facilities and extensive testing.

Many are custom-built for particular applications.

That makes production capacity extremely important.

When utilities, industrial developers and technology companies all increase their orders simultaneously, manufacturers cannot necessarily expand production quickly enough to satisfy everyone.

A delayed transformer can become a delayed substation. A delayed substation can become a delayed data center. And a delayed data center can mean billions of dollars of computing equipment sitting idle or operating below its intended capacity.

Think about the economic implications.

The most advanced processors in the world may be waiting for a piece of electrical equipment whose basic operating principles date back more than a century.

That’s the paradox of the AI revolution.

Technology is moving toward the future at breathtaking speed, but its progress increasingly depends upon the capacity of some of our oldest industrial industries.

When the Bottleneck Becomes the Business

There’s a fascinating economic principle operating underneath this entire story.

When an industry expands rapidly, the most valuable business isn’t always the one selling the final product.

Sometimes it’s the company controlling the scarce resource everyone else needs.

During the California Gold Rush, thousands of miners traveled west hoping to discover fortunes beneath the ground. Some succeeded. Many failed. But businesses supplying tools, transportation, clothing, food and other necessities could profit from the enormous influx of people regardless of which individual miners struck gold.

The same principle appeared during the railroad boom, the electrification of America and the development of the Internet.

And it may be appearing again today.

Artificial intelligence companies need semiconductors, but they also need electricity. Electricity requires generation equipment and transmission infrastructure. That infrastructure requires metals, industrial components and skilled workers.

Companies supplying scarce equipment or specialized services may enjoy stronger order books, greater negotiating leverage and opportunities to expand production.

But there’s an important distinction investors must understand.

Scarcity does not automatically guarantee profitability.

Higher material prices can hurt companies that consume those materials. Labor shortages can raise construction costs. Equipment manufacturers may struggle to expand capacity. Long-term contracts can limit how quickly higher market prices translate into earnings.

And when an industry eventually builds too much capacity, today’s shortages can become tomorrow’s surpluses.

That’s why identifying a bottleneck is only the beginning of the investment process.

The next question is who can actually turn that bottleneck into sustainable cash flow.

The New Industrial Hierarchy

For much of the past 20 years, investors were taught to think about the economy in terms of digital platforms.

Software companies enjoyed enormous margins. Internet businesses scaled rapidly. Technology giants could serve billions of customers without building a factory for every new user.

That model helped produce some of the largest corporations in history.

But artificial intelligence is introducing a different economic reality.

The next generation of computing infrastructure requires enormous upfront capital expenditures. These projects depend upon industrial supply chains, power availability, construction schedules and physical resources.

And that creates an entirely new set of strategic relationships.

A technology company may have enormous financial resources, but the utility controls the available electrical capacity. The turbine manufacturer controls production slots. The transformer manufacturer controls delivery schedules. The construction contractor controls access to specialized crews.

The mining company controls access to mineral production.

And in certain markets, the owner of strategically located land with access to power and fiber may possess an asset that cannot easily be replicated.

That doesn’t mean industrial suppliers suddenly become more valuable than the technology giants themselves.

But it does mean that bargaining power can shift toward businesses controlling essential capacity.

And that’s precisely the kind of economic change we look for in New World Assets and our broader MoneyQuake research.

The Companies Behind the Machines

Consider GE Vernova, one of the clearest examples of how the AI boom reaches beyond traditional technology companies.

GE Vernova manufactures gas turbines and electrical grid equipment used throughout the energy industry. As utilities and industrial customers expand generation and transmission capacity, the company participates directly in the infrastructure required to supply that electricity.

We originally identified GE Vernova for our New World Assets portfolio in April 2025 because we believed the world’s growing demand for reliable power would create a powerful investment opportunity.

Artificial intelligence has reinforced that thesis.

But GE Vernova is only one example of a much larger industrial ecosystem.

Companies such as Caterpillar manufacture equipment used in construction, mining, power generation and industrial projects. Electrical equipment manufacturers supply transformers, switchgear and power management systems. Engineering and construction firms design and build complex facilities.

Mining companies provide copper, silver and other essential materials. Energy producers supply the fuels needed to generate electricity. Water infrastructure companies help industrial customers manage another critical resource.

The opportunities span far beyond Silicon Valley.

And here’s what I find particularly fascinating.

Many of the businesses supplying the AI revolution belong to industries that investors spent decades describing as mature, cyclical or unexciting.

Heavy equipment. Mining. Power generation. Industrial construction. Electrical distribution. Water infrastructure.

Yet these are precisely the industries that may become increasingly important as the world attempts to turn trillions of dollars of digital ambition into functioning physical infrastructure.

The World Has Plenty of Money. Does It Have Enough Capacity?

This is the question I believe investors should be asking.

Governments can authorize spending programs. Technology companies can raise capital. Banks can provide financing. Investors can purchase shares in companies promising extraordinary technological breakthroughs.

But financial capital and physical productive capacity are not the same thing.

A billion dollars cannot instantly train a thousand experienced electricians. It cannot manufacture transformers that require specialized production equipment and long lead times. It cannot discover a copper deposit, complete environmental reviews and build a mine in a matter of months.

And it cannot guarantee that every planned AI project will receive sufficient electricity on its preferred schedule.

This is why I believe the Age of Scarcity is becoming one of the most important investment themes of the coming decade.

Not because the world is literally running out of everything.

It isn’t.

But because demand for certain physical resources and industrial capabilities may grow faster than supply can respond.

That creates bottlenecks. Bottlenecks create delays. Delays can increase costs. And in certain markets, those higher costs can translate into pricing power for companies that already control the scarce capacity.

The result may be a profound revaluation of businesses and assets that were previously overlooked.

The MoneyQuake Comes Full Circle

Think about where our journey began.

In 1946, ENIAC occupied an entire room and required a team of technicians to operate it. During the 1960s, IBM helped bring mainframe computing into corporate America. During the 1980s and 1990s, personal computers and the Internet connected millions of machines.

Cloud computing transformed the data center into an enormous industrial enterprise. Artificial intelligence then pushed computing requirements into an entirely different category, creating demand for gigawatt-scale campuses and extraordinary quantities of electricity.

And now we’re discovering that the biggest obstacle to building the AI economy may not be the intelligence of the machines.

It may be the limitations of the physical world surrounding them.

Copper deposits take years to develop. Power plants require enormous investments. Transformers must be manufactured. Transmission lines must be permitted and constructed. Concrete has to be delivered. Electricians, plumbers, welders and technicians must be trained and employed.

All of this takes time.

And time is something the technology industry increasingly seems unwilling to wait for.

That’s the tension at the center of the MoneyQuake.

The digital economy wants to expand at exponential speed, while the physical economy remains constrained by geology, manufacturing capacity, skilled labor and the time required to construct infrastructure.

And when those two worlds collide, something extraordinary can happen.

The value of the physical world begins to rise.

For decades, Wall Street rewarded companies that could grow without owning enormous amounts of physical infrastructure.

Now the next great investment cycle may reward some of the companies that own, produce, manufacture and construct the physical infrastructure everyone else desperately needs.

The mines. The power plants. The equipment factories. The transmission systems. The water infrastructure. The industrial capacity.

Because in an economy where everyone wants to build the future at once, the most valuable asset may not be the next revolutionary idea.

It may be the ability to turn that idea into something real.

And that brings us to the next chapter of our story.

Chapter 9: The Great Repricing of America’s Natural Resources

For generations, investors have valued America largely through its corporations, financial markets, technology and economic output.

But what happens when the world’s most valuable assets increasingly include the copper beneath our feet, the uranium in our ground, the energy we can produce, the water we control and the industrial infrastructure we can build?

What happens when the value of America’s physical resources begins to catch up with the extraordinary demand being placed upon them?

That’s where the MoneyQuake gets really interesting.

Get to the good, green grass first…

The Prophet of Profit,