The Data Center Gold Rush: Who Gets Rich and Who Gets the Bill? Part 1

by | Oct 20, 2026 | blog | 0 comments

The Data Center Gold Rush- Who Gets Rich and Who Gets the Bill Part 1

Watch this on Rumble: https://rumble.com/v7h44je-the-data-center-gold-rush-who-gets-rich-and-who-gets-the-bill-part-1.html

View my last shows on Data Centers:

What the Data Centers Are Really For: The Currency War Beneath the Cloud
What the Data Centers Are Really For: The Rise of the Information Empire

Opening — We Were Asking These Questions Years Ago

Welcome to Cause Before Symptom, where we don't chase symptoms—we test the cause against scripture.

Tonight, I want to take you back to something we investigated before. Long before this latest explosion of artificial intelligence infrastructure, I began asking questions about the enormous data centers appearing across America and around the world. Why were corporations investing billions of dollars into buildings filled with computers? Why were governments offering extraordinary incentives to attract them? And why did these developments seem to coincide with major changes taking place inside the international financial system?

Those questions led to earlier investigations, including *The Rise of the Information Empire* and *The Currency War Beneath the Cloud*. We examined the concentration of computing power, digital currencies, artificial intelligence, banking infrastructure and the growing importance of information in determining how economies operate. At the time, some of these connections were visible, while others remained questions we could not conclusively answer.

But something has happened since those investigations. The construction accelerated. The financial commitments grew into the hundreds of billions. Banks became involved in increasingly complicated financing arrangements. Technology companies began investing in one another, guaranteeing purchases and signing enormous long-term computing agreements. Governments continued offering incentives, while electricity providers confronted demands that could reshape regional power systems.

And I wanted to know something. How close were we to understanding what was happening?

So I returned to the investigation. What followed became more than 1,300 research checkpoints examining corporate financial statements, construction loans, investment funds, government agreements, energy infrastructure and research published by the Bank for International Settlements, the International Monetary Fund and the World Bank. I wanted to know who was providing the money, who expected to receive it, what guarantees were protecting those investments and what would happen if the projected demand never materialized.

What emerged was far more complicated than a collection of technology companies building computers. There were banks arranging multibillion-dollar loans, investment firms purchasing infrastructure, chip manufacturers supporting their customers' financing and local governments offering substantial tax incentives. In some cases, the same companies appeared on multiple sides of the financial arrangements.

But I also discovered something important about our earlier work. Some questions we raised now have substantial documentary support. Others remain unanswered. And certain possibilities that seemed connected cannot yet be demonstrated as parts of one coordinated system. I am not interested in pretending otherwise. If I discover something that contradicts an earlier conclusion, I want you to hear that too.

Scripture asks a question in Luke 14:28: Which of you, intending to build a tower, does not first sit down and count the cost, whether he has enough to finish it?

That question seems particularly appropriate tonight. Because the world is building something enormous, and the cost is not limited to the price of the buildings. There are electricity requirements, water demands, public incentives, private debts, contractual guarantees and expectations of future revenue that may take years to prove themselves.

So tonight, in Part One of *The Data Center Gold Rush: Who Gets Rich and Who Gets the Bill?*, I am going to follow the money.

I want to show you who finances these projects, who stands to profit, how the agreements work and where the risks ultimately fall. And along the way, we are going to revisit those earlier investigations to see what we understood, what we missed and what the evidence now reveals.

Because before I can tell you what these data centers are really for, I believe we must first understand who is paying to build them—and why.

Investigation 1 — We Saw the Buildings, but Did We Understand the Plan?

When I first began investigating the expansion of data centers, something did not sit right with me. I understood why companies needed servers. I have worked with servers and data centers for years. I understood cloud computing, storage, networking and the enormous amount of information moving across the internet. But what caught my attention was the scale of the expansion and the number of seemingly unrelated industries moving in the same direction.

In our earlier investigations, *The Rise of the Information Empire* and *The Currency War Beneath the Cloud*, I began asking whether these facilities were becoming something more significant than places to store information. Could they become the infrastructure upon which financial institutions, governments, corporations and entire economies would increasingly depend? And if they did, what would that mean for the people who relied upon those systems?

We examined digital currencies, tokenized assets, artificial intelligence, digital identity and the increasing ability of institutions to collect and analyze information. I wondered whether the real value of these enormous computing facilities would eventually come from the decisions they could help make, rather than simply the information they could store.

At the time, those were questions. Some developments were already documented, but the larger connections were difficult to establish. I could see the direction technology was moving without being able to demonstrate exactly how the financial arrangements behind the construction worked.

That is where our latest investigation changed things.

We discovered that the Bank for International Settlements has been researching artificial intelligence applications involving payment systems, transaction monitoring and financial supervision. Its Project Aurora explored how advanced analytics could identify suspicious financial activity across institutional boundaries. Other research has examined how AI agents might simulate decisions within wholesale payment systems. These are documented research projects, not proof that autonomous artificial intelligence has taken control of banking.

The International Monetary Fund has also examined how financial institutions increasingly depend on common technology providers and how interconnected AI investments could create financial instability. Meanwhile, the World Bank has investigated how governments and investors might finance the infrastructure necessary for artificial intelligence to expand internationally.

Think about what that means. We originally questioned whether computing infrastructure would become increasingly important to the financial system. Today, major international financial institutions openly publish research explaining why that infrastructure matters and how artificial intelligence could transform financial operations.

But there is a distinction I want to make. Finding that the BIS studies automated banking does not prove that a particular data center was constructed for the BIS. Finding that governments are researching digital currencies does not prove that every new computing campus is intended to operate a future monetary system. Those conclusions would require additional evidence.

What we can establish is that several transformations are occurring together. Computing capacity is expanding. Financial institutions are researching increasingly sophisticated automated systems. Governments are treating digital infrastructure as strategically important. And investors are committing enormous amounts of capital in anticipation of future demand.

That brings us back to our earlier question about control. Whoever owns a building does not necessarily control every application operating inside it. But companies providing essential computing services can gain considerable economic influence through pricing, contracts, technical standards and the difficulty customers face when changing providers.

So how close were we to the truth? We were asking questions about developments that are now supported by substantial institutional evidence. But the documents also tell us where to stop short of claiming more than we know.

And that is why tonight's investigation begins here. The buildings caught our attention. The financial relationships behind them may tell us far more about the future being constructed.

Investigation 2 — The Biggest Computing Construction Boom in History

To understand what is happening, I want you to imagine an industrial construction boom unlike anything the technology industry has experienced before. Across America and around the world, companies are acquiring enormous parcels of land, constructing buildings filled with specialized computing equipment, arranging new electrical generation and securing long-term access to power. These are not ordinary office buildings. Some proposed campuses require electricity comparable to the consumption of entire cities.

And the financial commitments are staggering. Microsoft, Amazon, Google, Meta and Oracle have announced enormous infrastructure investments. Nvidia supplies much of the specialized computing equipment. OpenAI and Anthropic develop artificial intelligence models that require tremendous computing resources. Companies such as CoreWeave specialize in providing access to the powerful processors these models need. Behind them stand construction companies, utilities, banks, private investment firms and governments competing to attract development.

But here is something I discovered during our investigation. When a company announces a hundred-billion-dollar investment, that does not necessarily mean it has already spent a hundred billion dollars. Announcements can include multiyear construction plans, financing arrangements, anticipated equipment purchases and contractual commitments dependent upon future conditions. The difference matters because the headlines often describe the full ambition of a project rather than the amount of money already invested.

Consider what it takes to build one of these facilities. The buildings require land, roads, foundations, cooling equipment, security systems and sophisticated networking. Inside are thousands of specialized processors, many manufactured by Nvidia. Those processors generate enormous amounts of heat and consume substantial electricity. Supporting them can require new substations, transmission lines, backup power systems and, increasingly, dedicated electricity generation.

And unlike traditional industrial facilities, the most valuable equipment inside these buildings can become technologically outdated in a relatively short period. A data-center building might remain useful for decades, while its processors could lose economic competitiveness much sooner. That creates an unusual financial challenge. Investors may be financing long-lived infrastructure containing equipment whose commercial value changes rapidly.

The construction boom also reaches far beyond the technology companies themselves. Utilities must anticipate electrical demand years in advance. Local governments must decide whether projected employment and tax revenue justify economic incentives. Communities must consider water availability, environmental effects and changes to their electricity systems. Financial institutions must determine whether future computing revenue will be sufficient to repay the enormous debts incurred today.

And that brings us to the most important question about this expansion. Who is going to purchase all this computing power?

Artificial intelligence already has genuine commercial applications. Businesses use it for software development, customer service, research, document analysis and automation. Scientists use advanced computing for simulations and discoveries. Financial institutions use machine learning to identify fraud and analyze risk. Governments and other organizations also require increasingly sophisticated computing systems.

But the existence of real demand does not establish that every proposed facility will become profitable. The industry is making extraordinary financial commitments based partly on expectations about applications, customers and revenues that have not yet fully materialized.

That is why I do not believe we should measure this gold rush only by the number of buildings under construction or the billions announced in press releases. We must examine who is putting actual money into the projects, who is borrowing it, who expects repayment and what happens if future customers do not arrive in sufficient numbers.

Because building the infrastructure is one thing. Building an economy capable of paying for it is something entirely different.

Investigation 3 — Follow the Money: Who Is Financing the Gold Rush?

When people see a new data center under construction, they usually recognize the technology company associated with it. They see the name Oracle, Microsoft, Amazon or another major corporation and assume that company is paying for everything. But one of the most revealing discoveries in this investigation is that the company using a data center is not necessarily the company financing it, constructing it or owning the property.

Behind these enormous projects is a financial network involving commercial banks, private investment funds, infrastructure developers and specialized real-estate companies. These institutions do not necessarily need artificial intelligence to become profitable themselves. Their business is providing capital, owning infrastructure, collecting interest and receiving payments under long-term contracts.

Consider the investment firm Blue Owl. Its real-estate investment structures have participated in data-center financing arrangements, including investments connected with the Red Chiles entities associated with Project Jupiter in New Mexico. These arrangements illustrate how institutional capital can be used to acquire or develop facilities that technology companies occupy under contractual agreements. The infrastructure owner expects to earn returns from the tenant's payments rather than directly selling artificial intelligence services.

Then there are the banks. Our investigation identified reported involvement by institutions including BNP Paribas, Goldman Sachs, MUFG and SMBC in financing associated with Project Jupiter. These banks can help arrange enormous construction loans, coordinate lending syndicates and distribute portions of financial exposure among participating lenders. In some transactions, a bank earns fees for arranging financing even if it does not retain the entire loan until maturity.

This is where the financial machinery becomes interesting. Imagine a developer borrowing billions to construct a data center. The developer pays contractors, purchases equipment and builds the facility. A technology company agrees to occupy or use it under a long-term contract. The developer expects those payments to support operating expenses, interest and repayment of borrowed money. Equity investors hope to receive whatever returns remain after the project's obligations are satisfied.

That structure can work when the project is completed, the tenant performs and the anticipated revenue arrives. But it also creates several different layers of risk. Construction costs can exceed expectations. Electricity connections can be delayed. Interest expenses can rise. A tenant's financial condition can deteriorate. And the value of the underlying property may depend heavily on whether another customer could realistically use the facility.

Another important distinction concerns who ultimately holds the debt. A bank arranging a loan may distribute portions to other financial institutions or investment vehicles. Private credit funds can also provide financing. Some investment structures receive capital from institutional and individual investors. But identifying a financing mechanism is not the same as proving that a particular pension fund or insurance company owns a specific project's obligations. That requires actual ownership records.

We also examined investment vehicles that offer periodic redemptions rather than unrestricted withdrawals. Their underlying assets may be difficult to sell quickly, even when investors receive regular statements showing estimated values. That does not mean the investments are fraudulent. It means their liquidity and valuation depend on rules that investors need to understand.

So who gets rich during the data-center gold rush? Equipment suppliers can earn revenue from sales. Banks can earn arrangement fees and interest. Developers can earn construction and development profits. Infrastructure owners can collect rent. Investment managers can collect management fees. Equity investors may receive substantial returns if everything performs as expected.

But projected returns are not guaranteed profits, and collecting fees does not eliminate every financial risk.

The deeper question is whether all these financial claims can ultimately be supported by the income generated from the computing services themselves. Because when the construction is finished and the lenders expect repayment, someone must produce enough revenue to keep the entire arrangement functioning.

# Investigation 4 — The Circular Money Machine: When Everyone Finances Everyone Else

One of the most fascinating discoveries in this investigation was that some of the companies selling the equipment needed for artificial intelligence are also helping finance the companies purchasing that equipment. At first glance, these arrangements can look like ordinary business partnerships. But when I began following the money, I discovered relationships that deserve a much closer examination.

Consider Nvidia, the company supplying many of the specialized processors powering today's artificial intelligence systems. Nvidia sells expensive computing equipment to cloud providers and infrastructure operators. Those companies use the equipment to offer computing services to artificial intelligence developers and other customers. But Nvidia has also invested in companies that purchase its technology and entered agreements to buy computing capacity from infrastructure providers.

CoreWeave provides a particularly revealing example. The company operates specialized computing infrastructure using Nvidia processors. Nvidia has invested in CoreWeave, while CoreWeave purchases Nvidia equipment. Nvidia has also agreed, subject to contractual terms, to purchase substantial amounts of computing capacity that might otherwise remain unsold. These are documented commercial relationships, but they create overlapping financial interests between the supplier and its customer.

Now consider what happens when similar relationships expand across the industry. An investor provides money to an artificial intelligence developer. That developer signs a computing agreement with a cloud provider. The cloud provider purchases processors from a chip manufacturer. The chip manufacturer may invest in the developer, support the infrastructure provider or guarantee certain financial obligations. Money moves through several companies, generating sales, contractual commitments and anticipated investment returns along the way.

That does not automatically mean anything improper has occurred. The equipment is real, the facilities are real, and computing services can have genuine commercial value. Strategic investments and purchase guarantees are established business practices. But there is an important distinction between revenue generated by independent customers and revenue supported by money supplied through closely connected business relationships.

Imagine a restaurant supplier investing in a restaurant, lending it money to expand and then reporting increased food sales because the restaurant used that financing to purchase more supplies. The sales are real. But the long-term success of the arrangement still depends on whether enough independent customers eventually purchase meals.

The same principle applies to artificial intelligence infrastructure, although the financial structures are considerably more complicated.

Our investigation also examined the enormous computing commitments announced by companies such as OpenAI and Anthropic. Some agreements extend many years into the future and involve extraordinary projected expenditures. But a multiyear commitment is not the same as cash already paid, and announced spending should not automatically be added together when contracts may contain overlapping capacity, options or conditional obligations.

The International Monetary Fund has raised concerns about interconnected financing relationships within the artificial intelligence industry. Its analysis identifies the possibility that companies acting as investors, customers and suppliers to one another could transmit financial stress across the broader system if expected demand or investment returns disappoint.

And this brings us to the question that matters most. Where does the money ultimately come from?

If businesses, governments and consumers purchase enough useful artificial intelligence services, the infrastructure can generate revenue capable of supporting its financial obligations. But if the expansion depends heavily on repeated investments, borrowing and guarantees, the industry becomes vulnerable when investors demand stronger evidence of profitability.

I am not claiming that artificial intelligence is an illusion or that every financing arrangement is circular. What I am saying is that the financial relationships deserve scrutiny because they can make the expansion appear stronger than independent customer demand alone would justify.

During a gold rush, selling shovels can be extremely profitable. Financing the miners who purchase those shovels can be profitable too. But eventually, somebody has to find enough gold to pay everyone back.

Investigation 5 — Project Jupiter: The $165 Billion Question

Now I want to take you to Doña Ana County, New Mexico, because one of the most revealing discoveries in our investigation involves a massive proposed data-center development called Project Jupiter. This project provides an extraordinary example of how technology companies, private investors, banks and local governments can become financially connected through a single infrastructure development.

When I first encountered the figure associated with Project Jupiter, I had to examine it carefully. County officials had authorized an industrial revenue bond structure with a maximum amount of $165 billion. That is an astonishing number. But it would be misleading to tell you that taxpayers borrowed $165 billion or that the county handed a technology company that amount of cash. The authorization establishes a ceiling for a particular financing and property arrangement. It does not mean the entire amount was issued, spent or guaranteed by the county.

The structure involves several companies, including entities known as Red Chiles, Yucca Growth Infrastructure and Green Chile Ventures. Oracle is associated with the project's planned computing operations. Our investigation examined county agreements, financing descriptions and reported construction-loan arrangements to understand how these participants fit together.

What makes the project particularly interesting is the separation between ownership, financing and use. A technology company can occupy or obtain computing capacity from a facility without directly owning every building or financing every construction expense. Infrastructure investors can provide equity, banks can arrange construction debt, and separate project entities can hold property interests and contractual obligations.

Financial reporting described approximately $18 billion in privately arranged construction financing associated with the development. That is separate from the much larger industrial revenue bond authorization. Confusing those figures would obscure what the agreements actually accomplish.

The county's involvement introduces another dimension. Industrial revenue bonds can support arrangements that provide significant tax advantages under applicable law. Project Jupiter also involves negotiated community benefits and payments associated with the development. We examined provisions involving promised contributions, guarantees and remedies if certain obligations are not fulfilled.

But these agreements require careful reading. A projected economic impact is not the same as tax revenue already collected. A promised community payment is not necessarily equivalent to the total value of the tax advantages. And a contractual guarantee may cover only specified obligations, rather than every possible loss associated with the project.

That distinction matters because the public is frequently presented with enormous investment figures alongside promises of jobs, economic activity and community improvements. Those benefits may materialize, but their actual value depends on construction progress, employment, operational activity and compliance with the agreements.

Project Jupiter also raises questions about electricity. A computing campus of this scale requires extraordinary power infrastructure. Plans involving dedicated generation and a microgrid introduce environmental, regulatory and financial considerations that extend beyond the buildings themselves.

And then there is the question of timing. Construction financing, tax incentives, community commitments and long-term operating agreements can have different expiration dates and enforcement provisions. A project can remain financially attractive to certain participants while exposing others to different kinds of risk.

This is why Project Jupiter became such an important case study. It allows us to examine actual contractual structures rather than relying exclusively on industry announcements.

The lesson is not that $165 billion disappeared into a secret arrangement. The lesson is that a headline figure can conceal a complicated distribution of ownership, tax advantages, financing obligations and financial risk.

And once we understand that structure, the next question becomes unavoidable: when the construction is delayed, the electricity is unavailable or the anticipated customers fail to generate enough revenue, who is contractually required to keep paying?

Investigation 6 — Who Gets Paid First, and Who Takes the Loss?

One of the most important lessons I learned while investigating these enormous data-center projects is that the people financing a building, the companies occupying it and the investors expecting a return do not necessarily face the same risks. In fact, some of the most revealing details are found in the contracts governing what happens when construction does not proceed according to plan.

Consider Project Jupiter in New Mexico. Financial reporting has described arrangements under which Oracle could remain responsible for certain carrying costs during delays in making the facilities operational. That is significant because investors financing construction want protection against circumstances that prevent a completed facility from generating its expected rental income.

But I want to be precise. The full private lease agreements have not been available for independent examination in our investigation. We cannot claim that every payment is guaranteed under every circumstance or that Oracle is responsible for unlimited losses. What we can establish is that reported contractual protections are an important part of the financing structure.

Now consider how the financial obligations are arranged. A construction lender provides capital and expects interest payments and repayment of principal. An equity investor contributes money in anticipation of profits after the project's senior obligations have been satisfied. The infrastructure owner expects revenue from a tenant or computing customer. The tenant expects to earn enough from its own operations to justify those payments.

If everything works, the arrangement can benefit all of them. But if construction stalls, the risks begin to separate.

A secured lender may have claims against project assets and contractual rights that provide priority over equity investors. An equity investor may absorb substantial losses before senior debt is impaired. A corporate guarantor may become responsible for specified obligations if another party fails to perform. And the technology company using the facility may face costs even when its anticipated computing revenue has not arrived.

This is where the financial strength of the companies involved becomes important. A contractual guarantee does not manufacture money. Its value depends on the guarantor's actual obligations, financial resources and ability to honor the agreement.

We also examined how infrastructure investments are valued. Some private investment vehicles hold assets that do not trade daily on public exchanges. Their reported values depend partly on appraisals, financial models and assumptions about future rental income. Those methods can be legitimate, but they do not eliminate uncertainty about what an asset would sell for during financial distress.

Another risk involves the equipment itself. The buildings may last decades, but advanced computing processors can lose economic competitiveness much faster. If a facility's equipment becomes outdated before the financing is repaid, its owners may need additional capital to remain competitive.

And then there is electricity. Imagine a completed data center that cannot operate at its intended capacity because sufficient power is unavailable. The construction costs still exist. Interest may continue accumulating. Investors still expect returns. But the revenue required to support those obligations may be delayed.

This is why the exact language of leases, guarantees, completion requirements and termination provisions matters so much. Two projects with similar construction costs can expose their investors to very different financial consequences.

There is another distinction I want the audience to understand. A private lender losing money does not automatically mean taxpayers must repay that lender. Public exposure arises only through separate mechanisms, such as government guarantees, tax arrangements or utility cost recovery.

So who gets paid first, and who takes the loss? The answer depends on the contractual hierarchy, available assets and financial condition of the responsible parties.

During this gold rush, some participants earn fees immediately. Others receive payments over many years. And some provide capital whose return depends entirely on the project succeeding.

The real danger appears when projected future revenue is treated as though it were already guaranteed. Because when expectations fail, the contracts determine who discovers that the risk was theirs all along.

Investigation 7 — The Electricity Gold Rush: Who Gets the Power, and Who Pays the Bill?

There is something about this data-center expansion that I believe deserves far more attention than it receives. Everyone talks about artificial intelligence, the computer chips and the billions being invested. But none of these enormous facilities can operate without electricity. And when I began following the electricity requirements, I discovered another financial system developing alongside the data-center gold rush.

A traditional data center already consumes substantial power. But the latest artificial intelligence facilities can require hundreds of megawatts, and some proposed campuses envision gigawatt-scale demand. A single gigawatt represents one billion watts of electrical capacity. When developers begin discussing multiple gigawatts for one campus, they are describing requirements comparable to the generating capacity of several large power stations.

That electricity has to come from somewhere. Utilities may need new generation facilities, substations, transformers and transmission lines. Developers may construct dedicated power systems or negotiate arrangements with existing generators. And these projects can require enormous amounts of water for certain cooling technologies, although actual consumption varies considerably by design and location.

Project Jupiter provides an important example. Its proposed microgrid includes substantial on-site electricity generation. During our investigation, we examined changing emissions projections associated with the project's planned fuel-cell infrastructure. Those changes demonstrated why environmental claims must be evaluated against the actual equipment, operating assumptions and permit applications rather than promotional descriptions.

But here is where the financial question becomes particularly important. When a utility constructs infrastructure to serve a large data center, who ultimately pays for it?

In some arrangements, the developer directly funds dedicated facilities. In others, the utility invests in infrastructure and seeks to recover approved costs through electricity rates. The outcome depends on contracts, tariffs and regulatory decisions. It would be inaccurate to claim that every data center automatically forces ordinary customers to pay its electricity bill.

However, federal regulators have recognized the potential problem. The Federal Energy Regulatory Commission has been examining how large electricity users connect to transmission systems and how the resulting infrastructure costs should be allocated. Federal energy officials have also emphasized the importance of preventing costs attributable to large new customers from being shifted onto households and other businesses.

Think about what happens if a utility builds expensive infrastructure for a data center that promises decades of electricity demand. The utility may finance those improvements based on the expectation that the customer will remain connected and purchase substantial amounts of power. But what happens if the project is canceled, the tenant fails or the computing equipment becomes economically obsolete?

The infrastructure may remain. The debt associated with constructing it may remain. And regulators must determine whether the developer, utility shareholders or other electricity customers are responsible for any unrecovered costs.

That is why minimum-payment agreements, security deposits, termination penalties and financial guarantees matter. They are not minor contractual details. They can determine whether ordinary families and small businesses are protected from the financial consequences of speculative industrial expansion.

There is another issue. Electricity is not an unlimited resource available at every location. Large new customers can compete for generation capacity and transmission access. Depending on regional conditions, that demand can influence infrastructure planning, wholesale electricity prices and the timing of investments needed to maintain reliability.

And the companies constructing data centers are not the only businesses positioned to profit. Power developers, equipment manufacturers, construction contractors and infrastructure investors may also receive substantial revenue from the expansion.

So when someone tells you that a technology company is investing billions in a new data center, I want you to ask a second question.

Who is building the electrical system required to operate it?

Because the price of the computers is only part of the investment. The electricity infrastructure may serve communities for generations, and the financial arrangements made today could influence what those communities pay long after the original developers have collected their returns.

Investigation 8 — Government Incentives: The Public's Hidden Investment

When a major corporation announces that it is bringing billions of dollars in investment to a community, the news is usually presented as an economic victory. Local officials talk about construction jobs, new infrastructure, business opportunities and future tax revenue. And those benefits can be real. But during our investigation, I discovered that the public announcement often tells only part of the financial story.

Behind many large industrial developments are agreements involving tax exemptions, reduced assessments, infrastructure commitments and negotiated payments to local governments. These arrangements are intended to attract investment that might otherwise locate somewhere else. But they also raise an important question: how much public economic value is being exchanged for the promise of future development?

Project Jupiter in Doña Ana County, New Mexico, provides a remarkable example. As we examined earlier, the county authorized an industrial revenue bond structure with a maximum amount of $165 billion. That figure does not mean the county borrowed $165 billion from taxpayers or guaranteed repayment of the entire amount. Industrial revenue bonds can function as part of a property and financing arrangement that provides tax advantages to a qualifying private development.

The distinction is important. A government does not have to write a billion-dollar check to provide an economically valuable incentive. It can also reduce taxes that would otherwise apply, subject to the governing laws and agreements. The financial benefit to the developer and the opportunity cost to the public can be substantial even when no direct cash subsidy changes hands.

Our investigation examined reports estimating billions of dollars in potential tax benefits associated with Project Jupiter. But those estimates must be distinguished from actual taxes that would otherwise have been collected. The appropriate comparison depends on the project's value, applicable exemptions, the development that would have occurred without the incentives and the precise terms of the agreements.

In exchange, governments may negotiate payments in lieu of taxes, community-benefit commitments, infrastructure improvements and employment expectations. Project Jupiter's agreements include provisions addressing community payments and specified guarantees. But not every promised economic benefit is equally enforceable. A contractual payment with a defined deadline is different from a forecast describing thousands of potential jobs or billions in projected regional activity.

And that is where I believe the public deserves clarity. How many permanent employees will the facility actually require after construction? How much local tax revenue will be collected? What infrastructure must the community provide? What happens if the project is delayed, substantially reduced or abandoned?

Construction employment can generate meaningful economic activity, but it should not be confused with permanent employment. A facility requiring thousands of workers to build may need far fewer people to operate. Likewise, projected economic output is not the same as money deposited into a county's treasury.

There is also the question of competing priorities. Land, electricity capacity, water, transportation infrastructure and public administrative resources all have alternative uses. When a government commits those resources to one development, the public should understand both the anticipated benefits and the opportunities being forgone.

I am not arguing that every incentive is a mistake. Some projects may produce economic benefits that substantially exceed their public costs. Others may fail to meet expectations. The answer requires examining the actual agreements and measuring the results rather than accepting either the developer's promises or the critics' assumptions.

The larger discovery is that a project can be privately financed while still receiving considerable public economic support. Private banks may provide construction loans, private investors may own the facilities, and technology companies may sign the leases. Meanwhile, local governments can influence the project's profitability through tax treatment and infrastructure arrangements.

So when we ask who gets rich during the data-center gold rush, we also need to ask what the public receives in return.

Because the public's investment is not always a check written from the treasury. Sometimes it is the revenue a community agrees not to collect, based on the expectation that something more valuable will arrive tomorrow.

Investigation 9 — The BIS, IMF and World Bank: Why Is the Financial System Preparing for AI?

When we began our earlier investigations into data centers, one of the questions that concerned me most was why the world's financial institutions were moving toward artificial intelligence, digital payment infrastructure and increasingly sophisticated information systems at the same time technology companies were building enormous computing facilities. I wanted to know whether these developments were connected, and whether the buildings appearing across America were part of a much larger transformation.

So I began examining the Bank for International Settlements, the International Monetary Fund and the World Bank. These institutions have different responsibilities. The BIS provides a forum for central-bank cooperation and conducts financial research. The IMF focuses on monetary cooperation and financial stability. The World Bank finances and supports economic development. They are not interchangeable organizations, and their research does not automatically establish a coordinated construction program.

But what I found was significant.

The BIS has conducted research into how artificial intelligence and advanced analytics could change financial operations. Project Aurora examined methods for identifying suspicious financial activity across institutional boundaries. Other BIS research has explored automated payment-system simulations and the use of artificial intelligence to assist financial supervisors in analyzing complicated institutional information.

Consider what these systems could accomplish. Instead of examining financial activity only within one institution, advanced analytical systems could identify relationships among transactions, organizations and patterns that would be difficult for individual investigators to recognize. That could improve fraud detection, financial supervision and the identification of criminal activity. It also raises important questions about privacy, access to information and the consequences of incorrect automated assessments.

But I want to distinguish research from deployment. A central bank studying artificial intelligence does not mean it has surrendered control of its payment systems to autonomous machines. A successful experiment does not prove that the same system is operating across every commercial bank.

The IMF has examined another dimension of this transformation. Its research addresses the growing dependence of financial institutions on common technology providers, the possibility of correlated decisions by artificial intelligence systems and the financial risks associated with enormous AI investment commitments. It has also raised concerns about interconnected financing arrangements among technology companies.

Meanwhile, the World Bank has examined how countries can finance the physical infrastructure needed to participate in the artificial intelligence economy. That includes data centers, electricity systems, telecommunications networks and mechanisms intended to reduce investment risk.

Now consider what these findings mean together. Financial authorities are researching increasingly sophisticated computational systems. Development institutions are examining how computing infrastructure can be financed. Technology companies are constructing facilities intended to provide enormous amounts of computing capacity. And financial institutions are supplying capital to support that expansion.

Those are documented developments. They demonstrate that artificial intelligence infrastructure is becoming strategically important to both private industry and public institutions.

But there is something we have not established. We have not found evidence that the BIS, IMF and World Bank jointly directed Project Jupiter or that a particular data-center campus was secretly constructed to implement a universal financial-control system. Making that claim would require contracts, procurement records or other evidence connecting the institutions directly to the project.

What we can say is that the financial system is preparing for substantial technological change, and the companies controlling computing infrastructure may become increasingly important to that transformation.

And this brings me back to our earlier investigations. We questioned whether the future value of data centers might involve much more than storing information or answering questions through artificial intelligence.

The institutional documents now provide a reason to take that question seriously.

Because the gold rush is not only about selling computer chips. It is also about constructing the infrastructure upon which increasingly important financial and economic activities may depend.

And that leads directly to our final investigation: what are these enormous facilities ultimately being built to accomplish?

Investigation 10 — What the Data Centers Are Really For: Following the Evidence to Its Conclusion

After more than 1,300 research checkpoints, examining financial disclosures, government agreements, banking institutions, construction projects and the enormous investments flowing into artificial intelligence, I want to return to the question that started this entire investigation. What are these data centers really for? And were we correct when we began questioning their purpose years ago?

The first answer is straightforward. These facilities provide computing power. They train artificial intelligence models, operate cloud services, process enormous quantities of information and run applications used by businesses, researchers, governments and individuals. Those are legitimate commercial activities, and their importance should not be underestimated. Artificial intelligence is already changing how software is developed, how information is analyzed and how organizations operate.

But our investigation revealed that the buildings themselves are only one part of the story. Behind them are financial arrangements involving technology companies, banks, investment funds, electricity providers and governments. The facilities are being financed today because investors expect computing capacity to become increasingly valuable tomorrow. Some companies earn revenue selling the processors. Others collect construction fees, interest payments, rental income or management fees. The actual profitability of those arrangements depends on their costs, contracts and future demand.

We also discovered that the money does not always move in a straight line. Technology suppliers can invest in their customers. Infrastructure providers can receive financial support from companies purchasing their services. Artificial intelligence developers can make enormous computing commitments while simultaneously receiving investment from businesses that benefit from those commitments. These arrangements can support genuine economic growth, but they also create financial dependencies that deserve scrutiny.

Then we followed the electricity. A data center cannot function without reliable power, and the infrastructure required to provide that electricity can be extraordinarily expensive. Depending on the agreements, those costs may remain with developers or become part of utility investments subject to regulatory recovery. That is why questions about electricity rates, water availability, tax incentives and public infrastructure belong in this investigation.

But perhaps the most revealing discovery came when we returned to the international financial institutions we had examined in our earlier articles.

The Bank for International Settlements is researching artificial intelligence applications in payment systems, transaction analysis and financial supervision. The International Monetary Fund is examining the risks created by technological concentration and interconnected AI investments. The World Bank is studying how countries can finance the infrastructure necessary to participate in an increasingly computational economy.

These developments do not prove that a single organization is directing every data-center project. They do establish that computing infrastructure is becoming important to activities extending far beyond consumer artificial intelligence.

And that brings us to the question of control.

Ownership of a data center does not automatically confer control over every organization using it. But when financial institutions, businesses and governments depend on a relatively small number of computing providers, those providers can acquire considerable economic influence. Their pricing, availability, security, technical standards and contractual terms can affect organizations that rely on their services.

This is where I believe our earlier investigations were asking the right questions. We recognized that information infrastructure could become increasingly important to economic power. What we did not yet understand was the extraordinary financial machinery being assembled to construct it.

So what are these data centers really for? They are being built to supply computing services that investors, businesses and governments expect to become essential to future economic activity. That is the documented commercial purpose. Whether individual projects will fulfill those expectations remains uncertain.

And there is one question the financial documents cannot answer for us.

Will the economic value created by these facilities ultimately justify the extraordinary costs of building and operating them?

Because during a gold rush, the people selling equipment, financing construction and collecting fees can earn money long before anyone knows whether the miners will find enough gold.

And if the promises fail, the most important question may no longer be who expected to become rich.

It may be who was left holding the bill.

Conclusion — Who Gets Rich, and Who Gets the Bill?

Tonight, I want to return to where this investigation began. Years ago, when we first examined the expansion of data centers in *The Rise of the Information Empire* and *The Currency War Beneath the Cloud*, I wondered whether something much larger was developing beneath the surface of the technology industry. I questioned why so much computing infrastructure was being constructed, why financial institutions were modernizing their systems and why governments were becoming increasingly interested in artificial intelligence and digital infrastructure.

At the time, I could see developments that appeared connected, but I could not establish every relationship. And that distinction matters. There is a difference between recognizing where technology may be heading and demonstrating precisely how institutions are financing, developing and using it.

After more than 1,300 research checkpoints, I have a much clearer understanding of the financial machinery behind this expansion. I have examined banks arranging enormous construction loans, investment firms acquiring infrastructure, technology companies making multiyear computing commitments and governments offering substantial economic incentives. I have also examined the extraordinary electricity requirements and the financial risks that can accompany them.

And I discovered something that should concern anyone trying to understand the economics of this gold rush.

Not everyone involved has to wait for artificial intelligence to become enormously profitable before earning money. Semiconductor manufacturers can generate revenue from equipment sales. Banks can collect financing fees and interest. Developers can receive construction payments. Infrastructure owners can collect rent. Investment managers can earn fees for managing capital. But revenue is not the same as profit, and none of these arrangements guarantees that every participant will succeed.

The question is whether enough lasting economic demand will eventually exist to support the obligations being created today.

We also learned that private investment does not always mean the public has no financial exposure. Tax incentives can reduce future government revenue. Electricity infrastructure can create costs that must be allocated among developers, utilities and customers. And communities may commit valuable resources in anticipation of economic benefits that take years to materialize.

But I want to be equally clear about something else. We have not demonstrated that every data center is part of one secret plan, or that every financial institution is pursuing the same objective. The evidence does not justify those conclusions. What it does establish is that computing infrastructure is becoming increasingly important to commerce, banking, government and the operation of modern economies.

That brings us back to our earlier investigations. Were we close to understanding what was happening? In important ways, yes. We were asking whether control over computing infrastructure could become a source of economic influence. Today, the financing agreements, institutional research and infrastructure commitments give us much stronger evidence for examining that question.

But I also believe there is a lesson here that extends beyond technology.

In Luke 14, Jesus spoke about counting the cost before beginning to build. And in Proverbs 22:7, Scripture reminds us that the borrower is servant to the lender. Those passages do not tell us whether a particular data-center investment is wise. But they remind us that financial obligations have consequences, and promises about tomorrow do not eliminate responsibilities incurred today.

So who gets rich, and who gets the bill? The answer depends on who earns actual profits, who holds the contracts, who carries the debt and whether the projected demand becomes reality.

Tonight, we followed the money. But the money is only the beginning.

In Part Two, I want to examine what happens when this infrastructure becomes increasingly embedded in banking, payments, credit decisions, digital identity and government operations. Who controls the information? Who establishes the rules? And what protections exist for ordinary people when automated systems begin influencing decisions that affect their lives?

Because I am beginning to understand that the most important question is not simply who owns the computers.

It is who determines what those computers are allowed to do.

And that is why this program is called *Cause Before Symptom*. We do not stop at what is visible. We keep investigating until we understand the decisions, incentives and consequences behind it.

God bless you, and thank you for joining me tonight.

Bibliography

Endnotes

  1. Bank for International Settlements, Project Aurora: The Power of Data, Technology and Collaboration to Combat Money Laundering across Institutions and Borders (Basel: BIS Innovation Hub, May 2023), https://www.bis.org/publ/othp66.pdf. Project Aurora investigated the use of advanced analytics and collaborative approaches to identifying suspicious financial transactions across institutional boundaries. The findings concern experimental methods, not evidence of a universal operational surveillance system.
  2. Bank for International Settlements, “Project Aurora,” https://www.bis.org/project/aurora. The project demonstrates the BIS Innovation Hub's interest in improving financial-crime detection through data analysis, institutional cooperation, and privacy-enhancing technologies.
  3. International Monetary Fund, “AI: Deployment and Disruption,” IMF Annual Report 2026, https://www.imf.org/annual-report/2026/in-focus/ai-deployment-and-disruption/. The IMF identifies financial vulnerabilities arising from increasingly debt-financed artificial intelligence investments, including circular financing arrangements in which technology companies act as one another's customers, investors, and financiers.
  4. World Bank, Digital Progress and Trends Report 2025: Strengthening AI Foundations (Washington, DC: World Bank, 2025), https://www.worldbank.org/en/publication/dptr2025-ai-foundations/report. The report examines the computing, connectivity, data, and institutional foundations required for countries to participate in the artificial intelligence economy.
  5. Arman Shehabi et al., 2024 United States Data Center Energy Usage Report (Berkeley, CA: Lawrence Berkeley National Laboratory, December 2024), https://doi.org/10.71468/P1WC7Q. The report estimates that U.S. data centers consumed approximately 176 terawatt-hours of electricity in 2023, representing 4.4 percent of national electricity consumption. Its projected 2028 consumption range is approximately 325–580 terawatt-hours, or 6.7–12 percent of U.S. electricity consumption.
  6. Sarah Josephine Smith et al., United States Data Center Energy Usage Report: 2025 Update (Berkeley, CA: Lawrence Berkeley National Laboratory, June 2026), https://datacenters.lbl.gov/publications/united-states-data-center-energy-2025. The updated analysis projects that data centers could account for approximately 9.5–15.3 percent of total U.S. electricity consumption by 2030, with a central estimate of 11.8 percent. These figures are projections rather than measured future consumption.
  7. Oracle Corporation, Annual Report on Form 10-K for the Fiscal Year Ended May 31, 2026 (Washington, DC: U.S. Securities and Exchange Commission, 2026), https://www.sec.gov/Archives/edgar/data/1341439/000119312526277521/orcl-20260531.htm. Oracle's financial disclosures provide evidence concerning infrastructure investment, operating cash flow, financing obligations, and remaining performance obligations. Remaining performance obligations represent contracted future revenue subject to applicable accounting and performance conditions, not cash already collected.
  8. Doña Ana County, New Mexico, Project Jupiter Industrial Revenue Bond Series 2025A: Executed and Recorded Documents (November 2025), https://cms2.revize.com/revize/donaanacounty/Documents/About%20Us/Economic%20Development%20Projects/IRB%20Series%202025A%20-%20Executed%20%26%20Recorded.pdf. These documents include agreements concerning Project Jupiter's industrial revenue bond arrangements and community-benefit commitments.
  9. Doña Ana County, New Mexico, Project Jupiter Community Benefits Agreement (executed November 12, 2025), included in the recorded Series 2025A documentation. The agreement identifies participating project entities, community-investment commitments, county support obligations, and contractual remedies. The existence of a community-benefit agreement does not establish that every projected economic benefit is contractually guaranteed.
  10. Doña Ana County, New Mexico, Chapter 56: Industrial Revenue Bonds, County Code, https://ecode360.com/9675191. Industrial revenue bond arrangements must be distinguished from general-obligation government borrowing. An authorized maximum bond amount is not equivalent to the amount issued, the amount spent on construction, or a direct taxpayer repayment obligation.
  11. Oracle Corporation, “Project Jupiter Expands Community Outreach Effort Highlighting Benefits and Impacts of Its Investments in New Mexico,” June 16, 2026, https://www.oracle.com/news/announcement/project-jupiter-expands-community-outreach-effort-highlighting-benefits-and-impacts-of-its-investments-in-new-mexico-2026-06-16/. Oracle's public descriptions of anticipated economic benefits should be treated as company statements and forecasts, not independent confirmation that all projected benefits have occurred.
  12. Oracle Corporation, “Project Jupiter Is Expected to Generate More than $13 Billion in Economic Impact for New Mexico,” October 13, 2026, https://www.oracle.com/news/announcement/project-jupiter-is-expected-to-generate-13-billion-in-economic-impact-2026-10-13/. The projected economic-impact figure is not equivalent to guaranteed county tax revenue or money already received by local residents.
  13. Federal Energy Regulatory Commission, “Interconnection of Large Loads to the Interstate Transmission System,” Docket RM26-4-000, https://www.ferc.gov/rm26-4. The proceeding addresses proposed reforms affecting large electricity users, including data centers, and specifically examines whether such customers should pay the full cost of transmission upgrades required for their connections.
  14. Federal Energy Regulatory Commission, “FERC to Act on Large Load Interconnection Docket by June 2026,” April 16, 2026, https://www.ferc.gov/news-events/news/ferc-act-large-load-interconnection-docket-june-2026. FERC identified the rapid expansion of large electrical loads as a matter affecting transmission reliability, affordability, and equitable integration.
  15. Lawrence Berkeley National Laboratory, “Data Centers,” Energy Technologies Area Publications, https://eta-publications.lbl.gov/research-areas/data-centers. The laboratory's research addresses data-center energy consumption, cooling, electrical demand, and the implications of large new loads for electricity systems.
  16. International Monetary Fund, “AI: Deployment and Disruption,” IMF Annual Report 2026. The IMF warns that expensive AI investments may fail to produce anticipated economic returns and that interconnected financing arrangements can transmit financial stress among participating companies.
  17. Federal Energy Regulatory Commission, “Interconnection of Large Loads to the Interstate Transmission System,” Docket RM26-4-000. Electricity infrastructure cost allocation depends on applicable tariffs, contractual arrangements, and regulatory decisions. The existence of a large data center does not by itself establish that residential customers are subsidizing its electricity consumption.
  18. Doña Ana County, Project Jupiter Community Benefits Agreement. Public incentives, community-benefit payments, employment projections, and anticipated tax receipts represent different categories of economic value. Their financial significance must be evaluated separately.
  19. World Bank, Digital Progress and Trends Report 2025. The expansion of AI infrastructure is relevant to broader economic development, public services, commercial applications, and access to computing capacity. These documented purposes do not establish that all infrastructure projects operate under a single coordinated institutional plan.
  20. The Holy Bible, King James Version, Luke 14:28–30. The passage concerning counting the cost before building a tower provides the scriptural reference used in the opening and conclusion.
  21. The Holy Bible, King James Version, Proverbs 22:7. The passage concerning the relationship between borrower and lender provides the scriptural reference used in the conclusion.

#DataCenterGoldRush, #DataCenters, #ArtificialIntelligence, #AIInfrastructure, #FollowTheMoney, #BigTech, #Nvidia, #Oracle, #OpenAI, #CoreWeave, #ProjectJupiter, #WallStreet, #GlobalBanking, #BIS, #IMF, #WorldBank, #PrivateEquity, #EnergyCrisis, #ElectricityCosts, #TaxpayerMoney, #DigitalEconomy, #FinancialSystem, #EconomicPower, #CauseBeforeSymptom

DataCenterGoldRush, DataCenters, ArtificialIntelligence, AIInfrastructure, FollowTheMoney, BigTech, Nvidia, Oracle, OpenAI, CoreWeave, ProjectJupiter, WallStreet, GlobalBanking, BIS, IMF, WorldBank, PrivateEquity, EnergyCrisis, ElectricityCosts, TaxpayerMoney, DigitalEconomy, FinancialSystem, EconomicPower, CauseBeforeSymptom

Submit a Comment

Your email address will not be published. Required fields are marked *

Subscribe To Our Newsletter

TikTok is close to banning me. If you want to get daily information from me, please join my newsletter asap! I will send you links to my latest posts.

You have Successfully Subscribed!