Untangling the AI Money Loop

A few months ago, I was sitting with a friend who works in finance.

We were looking at the stock chart of a big AI chip company on his laptop. The line went almost straight up. He zoomed out. Still up. He zoomed out more. The graph looked like a rocket.

He smiled and said, “This is the future. AI is eating the world.”

I looked at the same chart and felt something different.

Curiosity. And a small, quiet worry.

Because behind every rocket chart, there is a story. Some stories are grounded in real value. Some stories are mostly hope, hype, and clever accounting.

And to live wisely in this age of AI, we need more than excitement. We need to understand the story behind the numbers.

This essay is not financial advice. It is an invitation to think more clearly about what I call the AI money loop.

Because when huge companies are investing in each other and also buying from each other, the money starts to move in circles. And circles can be healthy. Or dangerous.

What Is the AI Money Loop?

Let’s start simple.

Imagine I own a bakery. You are my friend and you own a coffee shop.

I give you a big loan so you can remodel your shop, with one condition: you must spend the money on my bread and cakes. Then I tell the world, “Look at my bakery, sales are booming!”

On paper, this is true. I did sell more bread. But the money for those sales came from me in the first place.

That is what people mean when they talk about circular financing.

In the AI world, it looks like this:

• A big tech company invests billions into an AI startup.
• The AI startup then signs a huge contract to buy chips or cloud services from the same big company (or its close partner).
• The big company books that contract as revenue.
• Investors see strong sales and get even more excited about AI stocks.

Money goes out as “investment” and comes back as “revenue”.

On one level, this is allowed and fully legal. On another level, it creates a question in our mind:

How much of this growth comes from real, independent demand, and how much comes from the money loop itself?

Two Stories About the Same Loop

There are two main stories people tell about this situation.

Both stories use some of the same facts. They only differ in how they interpret those facts.

Story 1: The Skeptic

The skeptic says, “This looks like a bubble.”

In their view:

• The demand is partly artificial.
  When a chip company funds an AI lab that then spends that money on the chip company’s products, it feels like the company is helping to create its own sales. The revenue is real, but it depends on that investment stream continuing.

• We have seen this movie before.
  In the late 1990s, some telecom equipment makers gave big loans to network operators. The operators used those loans to buy more equipment. When the internet bubble burst and many operators failed, the equipment makers had to write off both the loans and the “growth” they had reported earlier.

• The profits may be thinner than they look.
  Many AI labs are losing money. Their big spending on compute is funded by investors who hope that one day these models will become very profitable. If that profit never arrives, some of the “demand” we see now could vanish.

In short, the skeptic sees a snake eating its own tail. Capital goes around in a circle, inflates valuations, and makes everyone look rich on paper, until something breaks.

Story 2: The Builder

The builder, or the “bull,” tells a different story.

In their view:

• This is strategic ecosystem building.
  In many industries, big suppliers help their customers get started. Car makers finance fleet purchases. Cloud platforms give startups free credits. The goal is to grow the whole ecosystem. AI is just doing this at a larger scale.

• The money creates real assets.
  The billions spent on AI are building huge data centers, laying new fiber, and filling racks with powerful GPUs. These are physical things. They exist in the world. They can run software, serve customers, and generate revenue for many years.

• Compute demand is growing fast.
  Each new generation of AI models needs far more computing power. Generative AI already uses many times more compute than older “perception” AI. If we move toward more “agentic” systems, which can plan, reason, and act, the need for compute could grow again by a big factor.

So in the builder’s story, this period is like the early days of building railways or the power grid. A lot of money goes into steel and cables. The returns come later, once society starts to run on that infrastructure.

Both stories sound reasonable.

That is why we need critical thinking to move past slogans like “AI is a bubble” or “AI will change everything”, and look closer at what is actually happening.

The Hidden Layer: Incentives

In other essays in this series, we talk about the incentive layer. The idea is simple: if you want to understand any message, follow who benefits from it.

Let’s apply that here.

• Chip makers have a strong incentive to tell a story of endless compute demand.
  Their valuations depend on the belief that AI workloads will keep growing.

• Cloud providers have an incentive to show rising AI revenue.
  This helps justify massive spending on data centers and supports their stock price.

• AI labs have an incentive to tell a story of huge future profits.
  They need this story to raise money and to explain their current losses.

None of this means anyone is lying.

It does mean that every actor is motivated to present the most optimistic version of the future that still feels believable.

Critical thinking does not ask us to be cynical. It asks us to see the incentives clearly, so we are not hypnotized by the story.

The Real Question: Who Pays In The End?

To keep our thinking clear, we can use a simple test.

Whenever you hear about a big AI deal, ask yourself:

“Who is the final, independent customer that pays for this, and what problem are they solving?”

Because in the long run, the money must come from someone who:

• is not funded by the same supplier,
• is solving a real problem,
• is willing to pay from actual profits or budgets.

For example:

• A bank that uses AI to reduce fraud.
• A hospital that uses AI to support doctors and save time.
• A small company that uses AI to serve more customers with fewer people.

These are real use cases. They do not depend on the money loop. They depend on AI actually making work better, faster, or cheaper.

So the key question for this “compute boom” is:

When we strip away the internal deals between giants, is there enough real, external demand to justify all this infrastructure?

How To Think About The AI Boom Without Getting Lost

We cannot control what big companies do. We can control how we think about it.

Here are a few simple practices you can use when you see headlines about AI investments, GPU shortages, or massive data center plans.

1. Name the Story You Are Being Sold

Ask yourself:

• What world does this article or post want me to imagine?
• Is it a world of endless progress and profit?
• Or a world of collapse and crisis?

Simply naming the story helps you step back for a moment. You are no longer inside the story. You are watching it.

This small step already gives you more freedom of mind.

2. Follow the Money Path

Try to trace the path of money with simple language:

• Who gives money to whom?
• In what form: investment, loan, or payment for a product?

If company A invests in company B, and B spends that money on A’s services, note that.

You do not need advanced finance skills for this.

Just write it on paper like a small map:

Big Cloud → invests in → AI Lab → signs contract with → Big Cloud

If the line makes a circle, your “money loop” alarm can blink gently.

It does not mean the loop is always bad. It means you should look more carefully.

3. Separate Assets From Utilization

Remember that physical assets can be both real and wasted.

A stadium is real. If nobody uses it, it becomes a ruin.

Data centers and GPUs are similar. They are real. They cost money. They use power. The key question is: Will someone pay enough, for long enough, to use them?

So when you hear about a new “AI supercomputer,” ask:

• Who will use this capacity?
• What will they pay for?
• Are there clear examples today, or only guesses about tomorrow?

4. Ask What Would Break The Story

Every strong story contains hidden assumptions.

Try asking:

• What must stay true for this AI compute boom to make sense?
• What would have to change for it to become a problem?

For example:

• If model performance stops improving as fast as before, do we still need 10 times more compute?

• If enterprises move slowly to adopt AI, can the builders fill all that capacity?

When you know which assumptions sit under the story, you can watch them over time.

What This Means For You, Personally

You might not be an investor or an AI engineer.

Still, this money loop affects you:

• Through your pension funds and index investments.
• Through the products you use.
• Through the way your job might change as companies chase AI dreams.

So what can you do?

If you are an employee

When your company talks about “AI transformation,” listen for clarity:

• Do leaders explain specific use cases and expected results?
• Or do they mainly repeat buzzwords from outside?

You can ask gentle questions:

“Can you share a concrete example of how this will help our customers or our team?”

These questions are not hostile. They are a service. They bring the conversation from story to reality.

If you are a founder or builder

You may receive offers that mix investment with infrastructure deals.

Before saying yes, ask:

• Does this deal really help me build a sustainable business?
• Or does it mostly help someone else hit their sales targets?

Sometimes vendor financing is helpful. Sometimes it ties you to one supplier and pushes you to overbuild before you have real customers.

Choosing a smaller, cleaner deal can be an act of freedom.

If you are a curious citizen

You do not need to know every technical detail to think clearly.

You only need a few questions:

• Where does the money come from?
• Who gets paid?
• What real work does this enable?
• Who would suffer if the story stops?

These questions help you stay grounded in a time of big promises.

The Deeper Skill Behind All This

In the Ministry of Meaning, we talk about four core human skills: critical thinking, empathy, storytelling, and imagination.

The AI money loop touches all four.

• Critical thinking helps you see the difference between real growth and circular growth.
• Empathy reminds you that behind every deal there are people, jobs, and emotions.
• Storytelling is what companies use to justify their moves, and also what you can use to explain your own view.
• Imagination lets you picture both good and bad futures, so you do not fall in love with only one.

The goal is not to become anti-AI or pro-AI.

The goal is to become awake.

A Closing Thought: Riding The Money-Go-Round With Eyes Open

The AI compute boom may turn out to be one of the greatest infrastructure projects in history. It may also contain pockets of serious overbuild and some painful crashes.

Both can be true at the same time.

The circular deals are real. The revenues are also real. The risk sits in the gap between the story and the independent demand that must support it.

You and I cannot predict the exact future of this market.

What we can do is practice a habit:

• Step back from the hype.
• Follow the incentives.
• Look for the final customer.
• Ask what would make the story fail.

When you do this, something important happens.

You feel less like a passenger on the AI roller coaster. You start to feel more like a clear-minded observer, able to choose which rides you join and which ones you watch from a distance.

The verdict on the AI money loop is still open.

But your inner clarity does not need to wait.

You can start building it today, one question at a time.