We’re Debating the Wrong AI Questions

by Donald J. Claxton, April 1, 2026.

The conversation around artificial intelligence has settled into two familiar lanes.

Safety.
Regulation.

One side warns about what could go wrong.
The other argues about how tightly we should control it.

Both matter.

Neither will determine whether we win.

The outcome of the AI race will not be decided by policy frameworks alone.

It will be decided by whether we can build—and sustain—the physical systems that AI depends on.

Power.
Infrastructure.
Skilled labor.

Those are the constraints.

And right now, they are not keeping up with the promises being made.

The market has started to notice.

Wages in skilled trades are rising.
Demand is outpacing supply.
Work that was ignored for decades suddenly carries a premium.

From one perspective, that looks like opportunity.

It is.

But it is also a warning.

Alex Hormozi often frames these moments correctly at the individual level: when something becomes scarce, the market rewards those willing to do it.

That logic explains why trades are regaining value.

It does not explain whether we have enough people to meet demand.

The market can reward scarcity.

It does not guarantee capacity.

That’s the gap.

And it’s where Mike Rowe has been far more direct.

For years, he has pointed out what most institutions ignored: we pushed an entire generation away from skilled work.

Not accidentally.
Deliberately.

We told people that success meant avoiding manual labor.
We elevated credentials over competence.
We built an education pipeline that produced degrees faster than it produced capability.

Now we face the result.

Artificial intelligence does not run on theory.

It runs in buildings.

Data centers require land, materials, and power.
Power requires generation, transmission, and maintenance.
Every layer depends on people who know how to build, install, and keep systems running under real conditions.

Electricians.
Linemen.
Welders.
Technicians.
Operators.

Without them, nothing scales.

Yet the national conversation continues to orbit abstractions.

We debate model safety.
We debate regulatory authority.
We debate who controls the technology.

Meanwhile, the question that matters most goes largely unasked:

Do we have the capacity to build what we’re promising?

That question is not theoretical.

Projects slow down not because the technology fails, but because the physical work cannot keep pace.

This is not a future risk.

It is a present constraint.

The difference between seeing this as opportunity and seeing it as a problem is the difference between individual gain and national capability.

From the individual perspective, scarcity creates leverage.
From the national perspective, scarcity creates limits.

If too few people can do the work, the work does not get done—no matter how valuable it becomes.

This is what the current AI debate misses.

We are arguing over how to govern a system we may not be able to fully build.

We are focusing on control while ignoring capacity.

We are treating AI as a software revolution when it is equally a physical one.

The countries that lead in AI will not simply write the best code.

They will build the most resilient systems.

They will generate the power required to sustain them.
They will train the workforce required to maintain them.
They will align culture with the reality that not all critical work happens behind a screen.

What must change

First, we have to rebuild the skilled-trades pipeline at scale. A country cannot promise industrial expansion while starving the workforce that makes it possible.

Second, we have to treat power and grid modernization as AI issues, not separate ones. If the electricity does not scale, neither does the technology.

Third, we have to stop talking about trades as fallback work. They are foundational work, and the countries that remember that will outperform the ones that don’t.

Technology does not replace the physical world.

It depends on it.

A nation that cannot build cannot lead.

That is not a slogan.

It is a constraint.

The conversation we need to have is not whether AI will reshape the future.

It will.

The question is whether we are prepared to support the systems that make it possible.

That answer will not come from a policy paper.

It will come from what we train, what we build, and what we choose to value.

We are not short on intelligence.

We are short on alignment.

And until that changes, the outcome is not as certain as the headlines suggest.

ABOUT DONALD J. CLAXTON: 

I’ve worked at the front edge of multiple technology shifts—from early web infrastructure to today’s applied AI systems. Across each one, the same truth holds: what determines outcomes isn’t what gets discussed, but what gets built.

Today, I apply AI in real production environments using CNC and physical manufacturing systems. That perspective removes any illusion about where the limits are. Technology scales only when the infrastructure and people behind it do.