For the last few years, it has felt like artificial intelligence — especially ChatGPT from OpenAI — was unstoppable.

Millions of users.
Billions in funding.
Headlines calling it the future of humanity.

From the outside, it looked like the king of tech had already won.

But behind the scenes, the numbers tell a far more fragile story.

Some analysts now warn that the very technology making AI powerful may also make it financially unsustainable.

And that raises a deeper question — not just about business…

…but about what happens when humanity places its hope in machines.

---

The Multi-Billion Dollar Burn

OpenAI reportedly generates massive revenue.

Yet internal projections suggest something shocking:

They could be losing tens of billions per year within just a few years.

Not because the product failed.

Not because nobody uses it.

But because AI itself is brutally expensive to run.

Unlike normal software, AI doesn’t scale cheaply.

Every improvement demands enormous increases in:

• computing power
• electricity
• data center capacity
• specialized chips
• engineering talent

And those costs don’t rise gradually.

They explode.

---

The Scaling Law Trap

Traditional software improves through smarter coding.

AI improves through brute force.

To make a model significantly smarter, companies must pour in exponentially more:

• processing power
• training data
• energy
• hardware

Training a frontier AI model can now cost hundreds of millions to over a billion dollars for a single training cycle.

And that’s just the beginning.

You must retrain again.
Upgrade again.
Buy new chips again.

It’s like building a rocket…

…that burns money every second it flies.

---

The $40,000 Chip Reality

Modern AI depends heavily on high-end chips from companies like Nvidia.

Each advanced AI chip can cost $30,000–$40,000.

But one chip isn’t enough.

Training a single major AI model may require tens of thousands of them running together.

And here’s the cruel part:

These chips become outdated fast.

Unlike a factory machine that lasts decades, AI hardware may need replacing every 18–36 months.

Imagine buying an entire airline fleet…

…and replacing it every two years.

That’s the economics of modern AI.

---

Electricity: The Hidden Giant

Then comes the real monster:

Power consumption.

Large AI supercomputers can require electricity comparable to millions of homes.

This is no longer just a software problem.

It’s an infrastructure problem.

National grids.
Transformer supply.
Energy contracts.

AI is no longer just code.

It’s industrial.

---

The Free User Problem

Here’s something most people don’t realize:

Every time someone uses free AI, it costs real money.

Real electricity.
Real chip wear.
Real server load.

When hundreds of millions of users log in daily, the free tier alone can cost billions annually.

Which creates a brutal equation:

The more successful the platform becomes…

…the more expensive it is to operate.

---

The Competition Squeeze

Meanwhile, competitors like:

• Google
• Meta

are releasing rival AI systems.

Some are cheaper.

Some are open-source.

Some are bundled into existing ecosystems.

Users can switch instantly.

There’s no lock-in.

No moving cost.

No painful migration.

Just one click.

---

The Race Toward AGI

Because of these pressures, AI companies are betting everything on one breakthrough:

Artificial General Intelligence.

A system capable of performing nearly any intellectual task a human can.

If achieved, the payoff could be enormous.

Entire industries could transform overnight.

But if the breakthrough takes too long?

The funding gap could reach tens or even hundreds of billions.

No company — not even the biggest — can burn cash forever.

---

The Most Likely Outcome: Not Collapse, But Absorption

Here’s the surprising part.

Experts don’t predict a dramatic bankruptcy.

Instead…

They predict consolidation.

OpenAI could eventually be absorbed by a larger tech giant like Microsoft, which already provides cloud infrastructure and financial backing.

Not a crash.

A quiet takeover.

The independent pioneer becomes part of a larger empire.

---

The Christian Perspective: When Humanity Trusts Its Own Creation

Now let’s step back.

Because this story is bigger than AI.

It’s about the human heart.

Throughout history, humanity repeatedly builds something impressive…

…and then begins to trust it more than God.

We saw it in the Tower of Babel.

“Come, let us build ourselves a city…” (Genesis 11:4 NKJV)

The tower wasn’t just architecture.

It was human confidence without God.

And today?

AI risks becoming the modern Babel.

Not evil in itself.

Not forbidden.

But dangerously easy to idolize.

---

Technology Cannot Save the Human Soul

AI can:

• answer questions
• write books
• generate images
• assist research

But it cannot:

• forgive sin
• change the heart
• grant salvation
• give eternal life

Only Jesus Christ can do that.

As Acts 4:12 declares:

“Nor is there salvation in any other…”

The danger isn’t AI existing.

The danger is humanity expecting AI to fix what only God can fix.

---

The Lesson: Earthly Empires Always Look Invincible — Until They Aren’t

Rome looked eternal.

Babylon looked unshakable.

Empires rise.

Technologies boom.

Markets explode.

But Scripture reminds us:

“Do not put your trust in princes, nor in a son of man, in whom there is no help.” — Psalm 146:3

Replace “princes” with:

• governments
• corporations
• technologies
• artificial intelligence

The warning still stands.

---

Final Thought

ChatGPT may thrive for decades.

Or it may be absorbed into something bigger.

Or the entire AI landscape may reshape in ways nobody expects.

But one truth never changes:

Human innovation is temporary.

God’s kingdom is not.

And the greatest danger isn’t that AI might fail.

It’s that people might believe it never will.

---