Schools Were Built for a Different Economy

Schools are not failing. They are succeeding at a job that no longer exists.

Consider what a school was originally designed to solve.

Not ignorance in the abstract. A specific, material problem: information was scarce, teachers were scarce, books were expensive, and most families had no mechanism for transmitting structured knowledge across generations.

The institution existed to solve a distribution problem.

Gather the children.

Deliver the content.

Test the retention.

Advance by cohort.

Issue the credential.

That system worked.

For a long time, it worked remarkably well — not because it was philosophically ideal, but because it was structurally matched to the environment it operated in.

Scarcity of information meant centralized delivery was the only viable architecture.

Fixed curriculum made sense when the knowledge required for stable careers was itself relatively fixed.

The Carnegie Unit — the time-based credit system that still governs most school schedules — was introduced in 1906.

It was a reasonable solution to a real coordination problem in 1906.

The environment has changed.

The operating system has not.

Marshall McLuhan, the quintessential Canadian philosopher, may be more relevant now than ever before.

Because much of what he predicted is no longer theoretical.

It is infrastructural.

His most useful insight was not about content.

It was about what happens to human roles when the medium changes.

In Understanding Media, McLuhan argued that electric media would begin dissolving the walls the printing press had helped build. Print rewarded linear, sequential, individual processing of information. The classroom was architecturally matched to print: one teacher, one text, one sequence, one answer.

Electric media changed the tempo.

Information moved faster than deliberate sequential processing could absorb. And something unexpected followed — not the end of learning, but the erosion of subjects as discrete containers. When information becomes ambient and instantaneous, the neat divisions between disciplines stop making sense. The walls become porous. The sequence becomes optional.

McLuhan’s deeper insight was the emergence of a new kind of learner.

Not the passive recipient.

The information-gatherer.

The student was no longer waiting at the end of a delivery chain. The student was becoming a navigator — moving through an environment, gathering, interpreting, testing, connecting.

Not waiting for the lesson to arrive.

Hunting through an information field that never stops moving.

The old school was built for the recipient.

The new environment demands the hunter.

AI completes the reversal McLuhan described.

The child no longer waits for the teacher to explain.

Explanation is ambient.

Generation is ambient.

Simulation, tutoring, translation, feedback — ambient.

The institution was built to solve the problem of scarce information delivered in controlled sequence.

That problem is gone.

A new one has taken its place.

When information was scarce, the valuable thing was access to it.

When information is abundant, the valuable thing is judgment about it.

This is the structural shift most school reform misses entirely.

The response to AI in education has largely been one of two things:

Ban it.

Or add a class about it.

Both responses assume the old architecture is fundamentally correct and AI is either a threat to it or a subject within it.

Neither response understands the medium.

AI is not another audiovisual aid. It is not the internet in a different format. It is a change in the learning environment itself — the most significant change since the printing press created the conditions for mass schooling in the first place.

And as McLuhan would have predicted, it does not simply add new content to the existing container.

It changes what the container is for.

The old question was:

How do we get knowledge into students?

The new question is:

How do we build judgment in an environment where knowledge is everywhere and of wildly uneven quality?

Those are different problems.

They require different architectures.

The first problem is solved by delivery systems:

• textbooks
• lectures
• standardized tests
• sequential curricula

The second is solved by something closer to fieldcraft:

• searching
• evaluating
• synthesizing
• testing
• acting under uncertainty

The new student is not waiting for information.

The new student is hunting through it.

The school that does not understand this distinction is not failing through bad teaching.

It is failing through misalignment — preparing students for a problem that no longer exists while leaving them unequipped for the one they actually face.

The grammar of schooling — the deep structural logic that persists across reforms, administrations, and decades of attempted innovation — rests on assumptions that made sense in a different environment.

It assumes knowledge is stable enough to be fixed in the curriculum.

It assumes sequence matters: A before B before C.

It assumes age is a useful proxy for readiness.

It assumes a single teacher can remain the primary information source for thirty students simultaneously.

It assumes the credential reliably signals competence.

Each of these assumptions was reasonable when information was scarce and careers were predictable.

Each becomes more fragile as information becomes abundant and the half-life of skills shortens.

The critique is not of teachers.

Teachers are operating rationally inside a container that constrains them.

A skilled teacher in a fifty-minute period, responsible for thirty students at different levels, assessed through standardized metrics, is not failing when they deliver content efficiently.

They are succeeding at exactly what the system was designed to produce.

The problem is not the people.

The problem is the operating system.

And operating systems, when the environment changes enough, do not need better content.

They need to be rebuilt.

Consider what a hospital looks like when it is working well.

It does not separate medicine into isolated subjects and teach each in abstraction. It does not advance practitioners primarily by age cohort. It does not evaluate competence through memory recall under timed conditions alone.

It places people inside real environments, with real stakes, under the supervision of people who have already mastered the field — and it measures learning by whether the practitioner can actually perform.

The residency model is not perfect.

But it understands something the classroom model often does not:

Judgment forms through exposure to real constraint, not through the simulation of it.

A system that measures content recall will produce excellent content recallers.

A system that measures judgment under real conditions will produce something more durable.

The question is not which graduate we prefer.

The question is which environment produces them.

Here is what the accumulated evidence of this series suggests about schools:

They were built to solve a distribution problem in a world of scarce information and predictable careers.

They solved it.

Generations of students moved through the system and emerged credentialed and reasonably equipped for the economy that existed when the system was designed.

That economy has changed.

The information environment has changed.

The rate of change itself has changed.

And AI has accelerated all of it simultaneously, in a direction the existing architecture has no structural response to.

The answer is not more screens inside the old classroom.

That misunderstands the medium.

A screen delivering the same sequential content to the same age cohort solves the wrong problem with faster hardware.

The answer is not a class about AI.

That treats the environment as a subject instead of recognizing it as the new condition everything now operates inside.

When information is scarce, schools should deliver it.

When information is abundant, schools must do something harder:

Build judgment.

Build discernment.

Build the ability to navigate ambiguity, evaluate quality, coordinate with others, and act under real conditions.

Schools do not need to become more digital.

They need to become more real.

If the old model is an operating system built for a different environment, the next question is no longer rhetorical.

It is architectural.

What replaces sequence?

What replaces passive delivery?

What does a learning environment designed around judgment — rather than content transfer — actually look like?

That question has a structural answer.

And it looks nothing like a classroom.