Split View: 교육은 어디로 가는가 — 배움의 과거와 미래

Where Education Is Going — The Past and Future of Learning

• Opening — The World That Begins When the Bell Rings
• 1. How Today's School Was Born
  • Education Is Old, but School Is New
  • The Prussian Model and the Birth of Mass Schooling
  • Why It Took That Particular Shape
• 2. What That Model Did Well, and Where It Strains Today
  • What It Genuinely Achieved
  • Where It Strains Today
• 3. The History of Knowledge Opening Its Doors
  • The Printing Press — The First Democratization of Knowledge
  • Public Libraries — Knowledge Opened for Free
  • The Internet and MOOCs — The End of Distance and Cost
• 4. Bloom's 2 Sigma Problem — Tutoring as the Holy Grail
  • The Question One Study Posed
  • What 2 Sigma Means
  • The Grail, and an Old Dilemma
• 5. Adaptive and Self-Directed Learning — From Lecturer to Mentor
  • Attempts Toward Personalization
  • The Axis of Self-Directed Learning
  • From Lecturer to Mentor
• 6. Credentials or Real Skill — The Signaling Problem
  • What Do We Learn For
  • The Uncomfortable Question of Signaling Theory
  • A Balanced View
• 7. Equity and Access — Technology Can Narrow or Widen the Gap
  • The Same Tool, Diverging Outcomes
  • The Path by Which the Gap Widens
  • The Conditions That Narrow the Gap
• 8. AI in Education — Genuine Promise and Genuine Risk
  • The Genuine Promise — A Patient Teacher for Everyone
  • The Genuine Risk — Shallow Answers, Dependence, and Cheating
  • What Cannot Be Replaced — The Relationship Between People
  • The Place That Is Neither Hype nor Doom
• Closing — What Changes and What Does Not
  • Questions to Ponder
• References

Opening — The World That Begins When the Bell Rings

In the morning, a bell rings.

Students take their seats in classrooms sorted by age.

One teacher stands at the front, and thirty or so children open to the same page.

After a set number of minutes, the bell rings again, and everyone moves to the next subject.

Most of us find this scene deeply familiar.

So familiar, in fact, that it is easy to think this is simply what education is.

But step back a little, and the picture is neither natural nor eternal.

The school we know is closer to an invention, made roughly two hundred years ago.

And inventions can always be redesigned.

This essay looks at three things together.

First, how today's school came to have the shape it has.

Second, how knowledge has opened to more and more people, from the printing press to the internet.

Third, where personalized learning and AI might take the future of learning, without either doom or hype.

Education is how one generation hands itself down to the next.

So asking where it is going is, in the end, asking what kind of future we intend to pass on.

1. How Today's School Was Born

Education Is Old, but School Is New

Teaching and learning are as old as our species.

Parents taught children to handle fire, and craftsmen passed the skill of their hands to apprentices.

But "school" in the form we know it — gathering children of the same age and marching them through a fixed curriculum — is surprisingly recent.

For most of history, systematic education was a privilege of the few.

Only the children of priests, nobles, and wealthy merchants learned to read.

The great majority lived their whole lives unable to read or write.

The Prussian Model and the Birth of Mass Schooling

The template that most of the world's schools now resemble is often traced to nineteenth-century Prussia.

Prussia established, relatively early, a system of primary education that was state-funded, compulsory, and meant for every child.

It included age-graded classes, a standardized curriculum, fixed timetables, examinations, and a bureaucracy to manage it all.

This model soon spread across the world.

The Industrial Revolution added powerful momentum to that spread.

As factories and cities grew, society suddenly demanded that large populations be able to do basic reading, writing, and arithmetic.

It also needed habits of punctuality, following instructions, and moving through fixed procedures.

Why It Took That Particular Shape

The important point here is that this model did not arise out of foolishness.

Under the constraint of teaching as many children as possible with a limited supply of teachers, it was a highly reasonable solution.

The "one-to-many" structure, where one teacher instructs many at once, was a great achievement that turned education from a privilege of the few into a right of the many.

Sorting students into grades by age made administration and promotion simple.

A standardized curriculum was an attempt to guarantee a consistent level no matter which town or school a child attended.

Bells and timetables were the order that kept millions moving in step.

In short, today's school was an engineering compromise that made the ideal of "education for all" real within the conditions of the industrial age.

2. What That Model Did Well, and Where It Strains Today

What It Genuinely Achieved

We should not casually look down on this old model.

It actually raised human literacy dramatically.

In just a few generations, reading went from a privilege of the few to a basic skill of the many.

And school did not transmit knowledge alone.

It was also a place to mix with peers, learn rules, and practice cooperating with strangers.

By offering a common ladder on which a child could learn something regardless of a parent's background, school became a channel of social mobility.

Where It Strains Today

But a tool designed under the constraints of two hundred years ago cannot be perfect under today's conditions.

The most frequently noted problem is that it teaches everyone at one speed.

Within a single classroom, some students sit bored, hearing again what they already know, while others are pushed on to the next unit without understanding.

There is one teacher and many students, so the pace must be aimed at the average.

Standardization gave fairness, but it brought uniformity along with it.

It cultivates the ability to arrive quickly at a set answer, but things that are hard to measure, like curiosity and creativity, are often pushed to the margins.

There is a clear gap between the "ability to conform and repeat" that the industrial age demanded and the "ability to question and create anew" that this age asks for.

That gap is the starting point from which we now ask again about the future of education.

3. The History of Knowledge Opening Its Doors

Before we talk about the future of education, one large current deserves attention.

It is the fact that the number of people who can reach knowledge has steadily grown through history.

The Printing Press — The First Democratization of Knowledge

In the middle of the fifteenth century, Gutenberg's movable-type printing changed the world.

Before it, a book was an extraordinarily expensive object, copied out by hand one letter at a time.

A single book could cost as much as a house.

Printing made it possible to produce the same book in hundreds and thousands of copies.

Knowledge was no longer trapped in the scriptorium.

Ideas spread quickly across borders, and more people began to read and judge for themselves.

Public Libraries — Knowledge Opened for Free

The second great step was the public library.

No matter how many books are printed, they are out of reach if you cannot afford to buy them.

Public libraries, built across the nineteenth and twentieth centuries, created places where anyone could reach knowledge for free.

Even a child from a poor home could, at least in the library, read the same books as a wealthy child.

The library was a space that built the idea "knowledge does not belong only to those who can buy it" out of brick and shelving.

The Internet and MOOCs — The End of Distance and Cost

And then came the internet.

If the printing press made books cheap, the internet drove the cost of copying and delivering information to something near zero.

Now one person's lecture can reach millions on the other side of the planet at the same moment.

In the early 2010s, the so-called MOOC, the massive open online course, arrived amid great expectations.

Courses from the world's finest universities were opened for free to anyone with an internet connection.

For a while, people expected this to replace the university.

In reality, it did not go that far.

The share of people who finished a course was low, and it favored above all the minority who could drive themselves.

Even so, MOOCs proved one thing clearly.

What had blocked world-class learning was no longer "access" but "motivation" and "guidance."

4. Bloom's 2 Sigma Problem — Tutoring as the Holy Grail

The Question One Study Posed

In 1984, the educational scholar Benjamin Bloom published a short but deep paper.

Its title was "The 2 Sigma Problem."

The question Bloom and his colleagues asked was simple.

When learning the same material, how much does achievement change with the way a student is taught?

They compared several situations.

The ordinary classroom, where one teacher instructs many students.

And tutoring, where one teacher is devoted to a single student or a very small group.

What 2 Sigma Means

The result was striking.

Students who received tutoring pulled noticeably ahead of those taught in the ordinary classroom.

Bloom expressed the difference as roughly "two standard deviations," or 2 sigma.

Put simply, it means a student who was middling in a normal classroom could rise to near the top of the class with tutoring.

That is an enormous difference.

Why does this happen?

In tutoring, the teacher can align entirely with that one student.

When the student understands, they move ahead quickly; when the student is stuck, they stop right there and explain again.

There is immediate feedback, questions flow freely, and the pace is fitted to that person.

The Grail, and an Old Dilemma

Bloom called this result a "problem."

We already know tutoring is that effective, but assigning a dedicated teacher to every student was practically impossible.

There are nowhere near enough teachers, nor enough cost that society can bear.

So he asked a question.

Could we find a method as effective as tutoring yet as affordable as the classroom?

That question became the holy grail of education research for decades afterward.

Personalization, fitting learning to each individual, turned out to be the goal education had long dreamed of.

5. Adaptive and Self-Directed Learning — From Lecturer to Mentor

Attempts Toward Personalization

People have tried to answer Bloom's grail question in several ways.

One of them is so-called "adaptive learning."

Adaptive learning adjusts what to show next according to the learner's responses.

Get a question right and it offers something a little harder; get it wrong and it returns to an earlier concept to shore it up.

As a simple example, many of today's language apps follow this principle.

Long-studied learning techniques such as "spaced repetition" and "active recall" are also a form of personalization, in that they tune the timing of review to each person's rate of forgetting.

The Axis of Self-Directed Learning

Another axis is self-directed learning.

In an age when knowledge is already open in libraries and on the internet, "whether you can learn on your own" matters as much as "what you know."

Self-directed learning is the ability of a learner to set their own goals, choose materials, manage their pace, and check their own understanding.

This is not innate but a skill that is cultivated.

As the low completion rate of MOOCs showed, even when access is open, learning does not carry far without this capacity.

From Lecturer to Mentor

This current makes us rethink the role of the teacher.

In an era when knowledge was scarce, the teacher was almost the only channel through which knowledge flowed.

The teacher was the source of information, and the lecture was the way that information was handed out.

But in an age when knowledge is everywhere, the value of simply relaying information declines.

Instead, other roles grow more important.

The role of a mentor who sets direction for what to learn, works through the sticking points together, kindles motivation, and helps make sense of why one learns.

It is the often-cited shift from the "sage on the stage" to the "guide on the side."

In this change the teacher does not disappear but moves, if anything, into a more human role.

6. Credentials or Real Skill — The Signaling Problem

What Do We Learn For

There is a question hard to avoid when talking about education.

Do we really go to school in order to learn, or in order to prove?

A diploma, a degree, and various certificates do two things at once.

One is that they may mean you genuinely acquired something.

The other is that, even if you forget all of the content, they send a signal that "this person is diligent and capable enough to pass through such a gate."

The Uncomfortable Question of Signaling Theory

Economics has what is called "signaling theory."

This view holds that much of a degree's value comes not from what was actually learned but from the signal it sends about the person.

An employer cannot directly see a candidate's diligence or persistence.

So they take a record of finishing a hard program to the end as a proxy for guessing at the kind of person they are.

Pushed to its limit, this view arrives at a somewhat cynical conclusion.

If most of a degree's value is signaling, then we are spending all that time and money less on the learning itself than on the paper that proves it.

A Balanced View

Of course, this is one extreme.

Much knowledge and skill can only be used if it is actually learned, and that learning is genuinely real.

What a doctor has learned, what a carpenter has learned, what a programmer has learned is real ability before it is any signal.

Still, the signaling problem offers an important implication for the future of education.

As knowledge opens and paths to learning on one's own widen, "what you can actually do" may matter more and more than the sign that says "where you learned."

Here lies room for portfolios and records that show skill directly to gradually stand in for the old signboard.

7. Equity and Access — Technology Can Narrow or Widen the Gap

The Same Tool, Diverging Outcomes

Each time a new educational technology appears, a similar hope repeats.

"Now anyone can reach the best learning."

And that hope tends to be only half right.

Technology can narrow the gap, or it can widen it.

Which of the two it becomes depends less on the tool itself than on the conditions around it.

The Path by Which the Gap Widens

Consider one end.

Picture a child with a good device and fast internet, a quiet space to study, and an adult nearby to guide them.

For this child, online lectures and learning tools become wings.

Now consider the other side.

There is a child with an unreliable device, slow internet, no space to concentrate, and little guidance to steer by.

Even the same online lecture rarely does much for this child.

So the well-equipped side pulls further ahead, and the less-equipped side falls behind.

This is the so-called "digital divide," and beyond it, a divide in the capacity for self-directed learning as well.

The Conditions That Narrow the Gap

Yet there are clearly cases where the same technology narrowed the gap.

A rural student far away hears a great lecture from the city, and a student of limited means climbs the wall with free materials instead of costly private tutoring.

What tends to make the difference is things like these.

Whether devices and connectivity were actually provided.

Whether human guidance was present so no one wandered alone.

And whether the capacity to learn on one's own was cultivated from the start.

Technology only opens a possibility; who that possibility reaches depends, in the end, on our design and our will.

Forget this, and you get the paradox of a tool meant to narrow the gap widening it instead.

8. AI in Education — Genuine Promise and Genuine Risk

Now we arrive at today's hottest topic.

Artificial intelligence, especially AI that converses naturally with people, is entering education.

There is much hype here, and much fear.

All the more reason, then, to set promise and risk side by side and look at them calmly.

The Genuine Promise — A Patient Teacher for Everyone

Recall Bloom's 2 sigma problem from earlier.

Tutoring is effective, but there were far too few people to assign a dedicated teacher to everyone.

AI has the potential to shake exactly this old wall.

A partner that does not tire, does not rush, and does not grow irritated no matter how many times you ask the same question again.

A partner that changes its explanation to match the learner's pace and reaches for a different analogy at the sticking point.

If such a tool were given cheaply to everyone, it would spread the benefits of tutoring far more widely.

Educators such as Salman Khan pin great hopes on exactly this possibility — "a patient personal tutor for everyone."

Even a student too shy to raise a hand can ask an AI again and again.

A student who has fallen behind can quietly shore up an earlier concept unseen.

The Genuine Risk — Shallow Answers, Dependence, and Cheating

But the same tool can be used in the opposite way.

The first is the problem of shallow answers.

AI produces a plausible answer instantly, but that answer is not always right.

At times it states something confidently and wrongly.

If the learner lacks the power to judge for themselves, there is a risk of accepting a wrong answer at face value.

The second is the problem of dependence.

Grow used to getting answers too easily, and the muscle of wrestling with a problem and thinking for oneself may fail to develop.

There are moments in learning when a certain amount of difficulty and wandering is, in fact, the medicine.

If a tool removes all of that friction, learning can grow as shallow as it grows comfortable.

The third is the problem of cheating.

Before a tool that will write the assignment for you, it grows ever harder to tell "what one did oneself."

This demands that we redesign the way we assess.

What Cannot Be Replaced — The Relationship Between People

And there is one most important thing.

The fact that education is not the transmission of information alone.

A good teacher not only gives knowledge but believes in a person, holds expectations, and lifts them back up when they fall.

Saying "you can do this" at the moment someone wants to give up.

Helping to find, together, the meaning of what one learns for.

Such things are hard for even the finest tool to fully take over.

The motivation and courage to learn grow, for the most part, out of the relationship between one person and another.

The Place That Is Neither Hype nor Doom

So the balanced position can be summed up like this.

AI shines most when seen not as something that replaces the teacher but as a tool that lightens the teacher's load and makes personalization possible.

A picture in which repetitive explanation and basic practice are entrusted to the tool, while people concentrate on more human work.

The work of kindling motivation, setting direction, and drawing out growth through relationship.

Both shunning the tool out of fear and entrusting everything to it while erasing the human are choices that have lost their balance.

The diagram below shows the change in this old structure very simply.

[The Industrial-Age Classroom — One to Many]

             teacher
                |
   +-----+------+------+-----+
   |     |      |      |     |
 pupil pupil  pupil  pupil pupil
 (same pace, same content for all)


[Tutoring / AI-Assisted — Closer to One to One]

  teacher (mentor) ......... for each student:
        \                    custom pace and feedback
         \
   [ AI tool ] --- student A (moves fast)
   [ AI tool ] --- student B (repeats slowly)
   [ AI tool ] --- student C (different analogy)

  People handle motivation, relationship, meaning;
  tools help with repetition, practice, personalization.

Closing — What Changes and What Does Not

The outward form of education will change greatly in the years ahead.

Bells, age-sorted classrooms, and uniform timetables may loosen bit by bit.

Learning will be fitted more to the individual and freed further from the constraints of place and time.

Tools will grow ever smarter, and the doors to knowledge will open even wider than now.

Yet beneath all that change, some things do not change.

The fact that learning is, in the end, one person coming to understand the world and expanding themselves.

And the fact that this journey still needs guidance, encouragement, and human warmth.

The school of the past realized the ideal of "learning for all" in the manner of the industrial age.

The education of the future may push that ideal one step further.

Toward the ideal of "learning for all, fitted to each."

Whether that future turns out well or not depends not on the tool but on how we use it.

For technology only offers the possibility, while people decide the direction.

Questions to Ponder

1. Among the things you learned in school, what has helped you most to this day — was it the "knowledge," or the "power to learn on your own," or the "relationships with people"?

2. If we could give every student a patient AI teacher cheaply, what work do you think a human teacher must still take on nonetheless?

3. Of a degree's or credential's value, how much do you see as "real skill" and where does "signaling" begin, and how might that balance shift going forward?

4. To make a new educational technology narrow the gap rather than widen it, what do you think must be provided alongside the tool itself?

References

• Benjamin S. Bloom, "The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring" (Educational Researcher, 1984) — the classic study showing tutored students outperform ordinary classrooms by roughly two standard deviations.
• Salman Khan, "The One World Schoolhouse: Education Reimagined" (2012) — the Khan Academy founder's vision centered on personalization and self-paced learning.
• Salman Khan, "Brave New Words: How AI Will Revolutionize Education (and Why That's a Good Thing)" (2024) — a follow-up work on the promise and cautions of an AI personal tutor.
• Sir Ken Robinson, "Do Schools Kill Creativity?" (TED, 2006), https://www.ted.com/talks/sir_ken_robinson_do_schools_kill_creativity — the widely known talk questioning the relationship between standardized schooling and creativity.
• Encyclopaedia Britannica, "Education — History of education", https://www.britannica.com/topic/education — a historical overview of mass schooling and the modern school system.
• Michael Spence and the discussion of signaling theory and the signaling function of education (a classic topic in labor economics) — background on the value of a degree as a signal.