Who Really Makes Money From the AI Boom? Follow the Money Behind Project Stargate

When people talk about the artificial intelligence boom, they often focus on the companies building AI models. But some of the biggest opportunities may sit somewhere else.

Behind every AI model is an industrial supply chain involving investors, cloud companies, chip manufacturers, construction firms, energy providers, networking companies and data-centre operators.

Project Stargate provides a fascinating case study because its enormous scale makes that supply chain easier to see.

The headline number is only the beginning

Stargate was announced with an ambition of up to $500 billion in AI infrastructure investment by 2029. OpenAI, Oracle and SoftBank formed the core of the original joint venture, with MGX involved as a financial backer and Microsoft, NVIDIA and Arm playing technology roles.

But the book makes an important distinction.

It is misleading to imagine Stargate as one giant cheque for $500 billion.

The project has evolved into a network of overlapping agreements, site developments and bilateral deals. Partner disagreements have affected the unified joint-venture structure, while individual construction and compute arrangements have continued.

That distinction matters if you want to understand where the money actually goes.

Think of Stargate as an AI supply chain

At the top sits capital.

Investors provide money to acquire land, construct facilities and secure future computing capacity.

Then comes construction: architects, engineers, concrete, steel, cranes and specialised contractors.

Power infrastructure follows — generation, substations and transmission.

Then come cooling systems, networking equipment, servers and AI chips.

Only after all of that does the familiar AI layer appear: model training, inference and applications.

The book's deep-dive supply-chain section illustrates this progression from capital → land → construction → power → cooling → hardware → data centres → AI models → businesses and people.

That is why the AI economy is becoming much larger than software.

Oracle's transformation

One of the most interesting Stargate stories is Oracle.

Historically known primarily as a database and enterprise technology company, Oracle's role in building and operating Stargate infrastructure has positioned it much more prominently in AI cloud computing.

The book highlights Oracle's five-year, $300-billion-plus compute agreement with OpenAI and its reported AI-related contract backlog exceeding $450 billion.

But there is an important caution.

A backlog is not the same thing as money already earned.

The book explicitly distinguishes between future contractual obligations and revenue already received. Those contracts still have to be delivered, paid for and, in some circumstances, renegotiated.

That is an important lesson for anyone analysing the AI economy.

NVIDIA sells the machinery

NVIDIA occupies another critical position.

AI models require enormous amounts of computing power, and GPUs have become one of the most important pieces of the AI infrastructure stack.

The company is therefore positioned as a supplier to the infrastructure race rather than simply another AI application company.

Arm contributes chip architecture, while Microsoft brings cloud-scale software and data-centre experience. The book distinguishes these technology partners from equity partners that actually own stakes in the joint venture.

This is the classic infrastructure lesson:

Sometimes the most reliable position in a gold rush is selling the tools.

The financing model carries risk

There is another side to the story.

Stargate's financing relies partly on forward-purchase commitments — effectively, commitments to buy future computing capacity that help justify building infrastructure today.

That structure makes sense when demand continues growing.

But it also creates risk.

If AI demand or revenues grow more slowly than expected, infrastructure built on expectations of future demand could become harder to justify financially.

That does not mean Stargate will fail.

It means the scale of the bet deserves to be understood rather than reduced to a headline.

What can businesses learn from Stargate?

There is a broader business lesson here.

AI is creating opportunities at multiple layers:

• Infrastructure

• Energy

• Construction

• Hardware

• Cloud computing

• Software

• AI applications

• Professional services

• Training and skills

A business does not necessarily need to build an AI model to participate in the AI economy.

It may supply the infrastructure, support the companies building it, provide specialist services or create applications that become more valuable as computing becomes cheaper.

For African entrepreneurs, this distinction is particularly important.

The book argues that Africa may not need to compete immediately for giant AI campuses. A more accessible opportunity is using increasingly capable AI tools to solve local problems in areas such as agriculture, logistics, education, healthcare and financial inclusion.

Follow the money

Project Stargate is therefore more than a story about OpenAI.

It is a story about who supplies the AI economy, who finances it, who builds it, who powers it and who eventually captures the value it creates.

If you want to understand where the economic power of AI may actually sit, follow the infrastructure — and follow the money.

Explore the full Project Stargate book:
Get Project Stargate for Free