Opening: The Bill for the Free Lunch
There is an old maxim in economics: there is no such thing as a free lunch. And yet we eat free lunches every day.
Map apps guide our way, search engines unfold the world's knowledge, and social media carries word of our friends. For all of this we pay not a cent.
So where did the bill go? It did not disappear. It was simply written in a different unit than currency. Where we go, what we buy, whom we talk to, what we search for at three in the morning — every one of those traces is a line item. Instead of money, we pay with information about ourselves.
Let us begin with a thought experiment. Imagine an invisible assistant who records your every day without omission. This assistant notes when you woke, which cafe you stopped at, what you deliberated over before finally ordering at lunch, which video you fell asleep to at night.
And this record flows out to hundreds of companies you have never met. Unsettling? But this is not a thought experiment. It is close to the reality of nearly everyone who carries a smartphone.
Let us sketch it more concretely. The moment you silence your alarm, your wake time is logged, and the map app on your commute leaves a trail of your route. A single card payment at lunch, one tap of a like in the afternoon, a sentence typed and deleted in a search box before sleep — all of it is marked somewhere as a tiny point on a server.
The trouble is that these points do not stay scattered. They gather, connect, and are interpreted. The picture thus assembled sometimes says more than you know about yourself.
This essay does not set out to judge who is right and who is wrong. The debate over privacy is not a simple drama of good and evil but a hard problem in which value collides with value.
Convenience and freedom, safety and autonomy, efficiency and rights — all of these sit at one table. We will walk slowly around that table, listening fairly to the voices seated in each chair.
The Data Economy: The Trap in the "New Oil" Metaphor
Over the past decade, data has often been likened to the oil of the twenty-first century. The metaphor is powerful but also misleading.
Oil is a finite resource. Once burned, it is gone. Data is not like that. The same data can be copied infinitely, used by many companies at once, and grows explosively in value when combined with other data.
A single location record of yours is nothing much, but add purchase history and search records to it, and suddenly the outline of you as a person sharpens. Where you live, what you like, what worries you carry — all surface in the overlap of data.
The value of data comes precisely from this combination. Scattered points become a line, lines become a portrait. What the advertiser wants is not a single point but that portrait.
Another trap in the oil metaphor is this: it is clear whose oil it is, but not whose data it is. Whether the traces I leave truly belong to me, to the company that collected and refined them, or to both — this question of ownership remains a riddle society has yet to answer cleanly.
Setting the two resources side by side makes the difference sharper.
| Aspect | Oil | Data |
| --- | --- | --- |
| Consumability | Used up when burned | Persists and can be copied |
| Source of value | Its sheer quantity | Combination with other data |
| Ownership | Relatively clear | Murky and contested |
| Tendency to monopoly | Bound to a deposit | The larger the scale, the stronger |
How "Free" Services Actually Work
Simplified, the economic structure of so-called free services looks like this.
[User]
│ provides time, attention, and data
▼
[Platform] ── analyzes behavior → builds a precise profile
│
│ "reach people with specific interests"
▼
[Advertiser] ── pays for advertising
The key here is that the user is not the customer. The real customer is the advertiser, and the user is the source of attention sold to that advertiser. As a familiar saying goes, if you are not paying for the product, you may be the product.
Of course we cannot read this structure as simply evil. Thanks to the ad-supported model, billions of people use powerful tools without bearing the cost. Search, translation, maps, email — if all of these had been paid services, the digital divide would have been far wider than it is. The free model genuinely widened access.
The trouble is that the price is invisible. We agree to the deal scarcely knowing what we hand over, where it goes, or how it is used. A single thoughtless tap on a consent button buried in a long, complex policy is the signature on that contract.
Attention as a New Currency
Seen from another angle, what we actually pay is not only data. Another precious resource is attention. The time spent staring at a screen, the pull that makes us scroll once more — the real battlefield over which many services compete is precisely this attention.
Data and attention reinforce each other. The more they know about us, the longer they can hold us; the longer they hold us, the more they learn about us. This loop is the engine that sustains free services.
Here a balanced observation is in order. This engine is not always harmful. The same recommendation technology can find music we will truly love, and the same analysis can ease the traffic on the road. The problem is not the technology itself but whose interest the power tilts toward, and how.
The Diagnosis Called Surveillance Capitalism
The most influential name given to this phenomenon comes from the scholar Shoshana Zuboff. In a 2019 book she put forward the concept of surveillance capitalism, and it resonated widely.
Zuboff's central claim can be summarized this way. Where earlier capitalism took natural resources as raw material, a new form of capitalism takes human experience itself as raw material.
Our behavior, emotions, and habits are extracted as data and processed into products that predict future behavior. She called these behavioral prediction products.
Going further, Zuboff points to a tendency not merely to predict but to subtly nudge and steer behavior. The timing of a notification, the order of a feed, the flow of recommendations — all of this design, she argues, tilts our choices gently in a particular direction.
To feel the weight of this diagnosis, it helps to dwell on the word raw material. Just as ore dug from a mine is refined into a product, our daily lives are mined as raw material and processed. Only this mine lies not in the mountains but inside the screen in our hand.
Holding the Critique and the Rebuttal Together
Zuboff's diagnosis is powerful, but not everyone in scholarship agrees. For balance, let us hear the rebuttals.
First, some scholars note that business models differ too much from company to company to bundle them under a single banner of capitalism. A firm run on subscription fees and a firm run on advertising have different incentives toward data.
Second, there is criticism that the effect of behavioral steering is exaggerated. If advertising truly steered us with such precision, how do we explain the many absurd recommendations we receive? Prediction remains imperfect, and humans, the rebuttal goes, are harder to handle than supposed.
Third, there is the view that the use of data is not inherently harmful. The same location data improves traffic flow, and the same medical data aids the early detection of disease. A tool becomes a knife or a scalpel depending on its use.
These rebuttals do not nullify Zuboff's insight. They simply remind us that reality is too complex to be fully explained by a single book. Rather than accept any one diagnosis as absolute truth, we are safer alternating lenses.
The Two Views in One Table
Placing the diagnosis and the rebuttal side by side, one sees that neither fully overwhelms the other.
| Issue | Critical diagnosis | Cautious rebuttal |
| --- | --- | --- |
| Nature of data collection | A mining of experience | Also a byproduct of providing service |
| Power to steer behavior | Tilts choices in secret | Prediction remains imperfect |
| Outcome of data use | Drifts toward control | Also serves the public good and convenience |
| Proper response | Change the structure itself | Widen transparency and choice |
The point of this table is not to crown a winner. It is rather to read the two columns in turn and to feel that the truth may lie not in either column but in the tension between them.
Government Surveillance: The Scale Named Safety
There is a domain different in texture from corporate data collection, yet no less important: surveillance by the state.
Here we meet one of humanity's oldest dilemmas. Safety and freedom are often placed on opposite sides of the same scale. It is a tension in which raising one side seems to lower the other.
Two Voices Side by Side
The logic of those who emphasize safety runs thus. To protect society from threats such as terrorism, organized crime, and the exploitation of children, a certain capacity for surveillance is necessary.
Without the authority to trace communications and look into the flow of money, society lies exposed and defenseless. The old phrase that those with nothing to lose have nothing to fear speaks for this position.
The logic of those who emphasize freedom and privacy runs thus. Surveillance powers, once created, rarely shrink, and the temptation to abuse them always follows.
A tool built today to catch a terrorist can tomorrow be used to silence a legitimate dissenter. Moreover, in a society where people feel watched, they begin to censor themselves, withering free debate and creativity. This is called the chilling effect.
The chilling effect is all the more troublesome because it is hard to see. Hesitating to search a certain topic, or writing a sensitive opinion and then deleting it, leaves no mark in any statistic. Yet as those unwritten sentences accumulate, the range of what a society can ask and answer quietly narrows.
Surveillance for safety Protecting privacy
───────────── ─────────────
Early detection of threats Preventing abuse and concentrated power
Efficient investigation Blocking the chilling effect
Maintaining public order Respecting individual autonomy
│ │
└────── Where is balance? ──┘
The Old Metaphor of the Panopticon
The eighteenth-century philosopher Jeremy Bentham conceived a prison design called the panopticon. From a central watchtower one could see into every cell, while the prisoner could not tell whether he was being watched at that moment. The point was not the fact of surveillance but the mere possibility of being watched at any time, which led a person to control himself.
The twentieth-century philosopher Michel Foucault drew this panopticon in as a metaphor for how modern power operates. Surveillance, the insight goes, tames behavior not with the whip but with the gaze alone.
Today we are led to recall this metaphor again. Only now the modern watchtowers are not one but countless, and the watchers are not only the state but corporations and algorithms, dispersed. It is an age in which even who is watching has grown blurry.
There is an interesting difference, too. In Bentham's panopticon the prisoner at least knew he was in a prison. Today's surveillance is often invisible, and we leave traces while feeling we walk the streets freely. It is not that the prison walls have vanished, but that they have turned transparent.
Here too it matters not to raise one side's hand. Neither absolute freedom nor absolute safety exists in reality. Every society endlessly negotiates where to set the weight on this scale. The result of that negotiation tells us the character of that society.
A Small Thought Experiment: A House Built of Glass
Imagine this. Every house in a village is built of glass. No curtains, no walls. At first it seems everyone will turn honest. With nothing to hide, the logic goes, there is nothing to fear.
But after only a few days, people realize that the mere fact of always being watched changes their behavior. A dance they would have danced, a song they would have sung, a book they would have read — all give way to hesitation. Though they have done nothing wrong, the gaze sculpts their conduct.
What this experiment teaches is clear. Privacy is not merely a place to hide what is bad but a space in which to be freely oneself. The curtain is a tool of autonomy before it is a tool of secrecy.
Technologies That Protect Privacy
There is not only debate. Technology makes problems, but it also offers solutions. Let us examine, with metaphors, several technical approaches to protecting privacy.
Encryption: Sending It in an Envelope
A postcard lets the carrier read everything. A letter sealed in an envelope can be read only by sender and recipient. End-to-end encryption is the digital version of that envelope. A message is locked on the sending device and unlocked only on the receiving device, so even the service relaying it cannot see the contents.
The same tension of safety and freedom repeats around this technology. Strong encryption protects the individual but also makes it harder for investigators to reach evidence of crime.
On this, the argument that a backdoor for investigation should be built has long faced the rebuttal that a backdoor ultimately endangers everyone. A key, once made, carries no guarantee that only good people will use it. A door left open for the good is a door open to the bad as well — that is the heart of the latter view.
Anonymization and Its Limits
If we strip names and identification numbers from data, is it safe? Sadly it is not so simple. Many studies have repeatedly shown that even data claimed to be anonymized can re-identify individuals when combined with other data.
A classic line of research found that knowing only birth date, postal code, and gender suffices to single out a surprising number of people. A few points are enough to restore the portrait. This is why removing a name brings little comfort. The true identifier is not the name but the combination of traces that makes a person unlike anyone else.
Differential Privacy: Spraying a Suitable Fog
A sophisticated approach to this limit is differential privacy. The core idea is to deliberately mix a little noise into data so that individuals cannot be identified, while overall statistics remain useful.
The shape of the whole forest is visible, yet a single tree is hidden in the fog. This technique is increasingly used in large-scale statistical collection. Yet spray the fog too thick and the statistics blur; too thin and the individual emerges. Finding the balance between the two is the central task of this technique.
The Principle of Data Minimization
The simplest yet most powerful principle is perhaps not a technology but an attitude: collect only what is needed — this is data minimization. Data never collected cannot leak or be abused. The safest vault, as the saying goes, is the empty one.
Comparing the Technologies at a Glance
The approaches we have seen each answer a different question. None is a cure-all, and they are usually sturdiest when layered together.
| Technology | Core metaphor | Limit |
| --- | --- | --- |
| End-to-end encryption | Sending it in an envelope | Tension with investigative access |
| Anonymization | Tearing off the name tag | Risk of re-identification |
| Differential privacy | Spraying a suitable fog | A trade-off with accuracy |
| Data minimization | The empty vault | Giving up some conveniences |
The small lesson of this table is that protecting privacy is less a single magic solution than a craft of layering several tools to fit the situation. Seal it in an envelope, tear off the name, spray the fog, and collect less to begin with — these layers of defense work together.
The Shadow of the Data Breach
Before we speak of technology and regulation, we should note that collected data always carries a shadow: the leak.
Data carries the risk of seeping out from the moment it gathers. A vast database is an alluring vault to a thief, and a single mistake or a single intrusion can spill the information of millions at once. When we hand over data, we trust that it will be kept safe. That trust is often betrayed.
The important point here is that the harm of a leak is not immediate. Stolen information lies quietly somewhere, then surfaces months or years later in an entirely different shape.
Someone opens an account in my name, or pieces my information together to impersonate me. This is called identity theft. Once leaked, information may drift forever somewhere on the internet, so a danger thought to have passed can revive with time.
This shadow lends fresh weight to the principle of data minimization mentioned earlier, for the less data gathered, the less there is to lose.
At the same time, it commends to us one small habit: not handing over information that is not truly needed, and not using the same key, a password, on many doors. So that if one door is breached, the others may still hold.
[Data collection]
│ we gain convenience
▼
[Accumulated database] ── an alluring target for thieves
│
│ one intrusion, a mass leak
▼
[Delayed harm] ── surfaces months or years later as identity theft and the like
A Pause: Privacy Through History
It is easy to regard privacy as an invention of the digital age, but in fact the concept is far older. Only its shape differed from era to era.
In the days when everyone in a small village knew one another, privacy meant something other than it does today. A neighbor knowing my affairs was natural, and even the concept of anonymity was faint.
As cities grew and we came to live among strangers, a new freedom arose at last: anonymity in the crowd, the freedom to walk a street where no one knows me. The intriguing paradox is that digital technology is taking that anonymity back. We live in the crowd of the city, yet are identified more sharply than ever.
Intriguingly, a famous piece arguing for privacy as a right appeared in the late nineteenth century. It rose from the worry that newly arrived technologies — the portable camera that takes pictures with ease and the newspapers that spread them — were threatening people's private lives. What they urged at heart was the right to be let alone, the right to guard one's own domain undisturbed.
Picture the unease of people in those days. As you walk down the street, someone abruptly points a camera, and the image so captured runs the next day in a corner of the newspaper. To have one's image drift through the hands of strangers without one's consent or control was, for that time, a shock. It is not so far from how we are casually caught today in the background of someone else's photo.
What this history teaches is plain. Each time a new technology arrived, the boundary of privacy was redrawn, and society had to seek a new settlement each time. The camera did so, the telephone did so, and now the smartphone and artificial intelligence do. We stand, in effect, at the most recent scene of a negotiation humanity has passed through many times.
The Arrival of Regulation: GDPR and Its Ripples
When technology and the market alone failed to balance things, society reached for the tool of law. Among such efforts, the most influential case is the European Union's General Data Protection Regulation, commonly known as GDPR. Enforced in 2018, it became in effect the reference point for data regulation worldwide.
Unpacked, the core rights embodied in GDPR look like this.
| Right | Meaning |
| --- | --- |
| Clarity of consent | Clear, freely given consent must be obtained for data collection |
| Right of access | One can confirm the data held about oneself |
| Right to rectification | One can correct inaccurate data |
| Right to erasure | Under certain conditions one can demand deletion of one's data |
| Right to portability | One can carry one's data to another service |
In particular, the right to erasure became widely known under the name the right to be forgotten, confronting head-on the problem of a past that follows us forever in the digital age.
Assessments of the Regulation Also Diverge
The verdict on GDPR likewise does not tilt to one side. For balance, let us hear both.
Those who view it positively credit the regulation with returning to individuals control over their own data and strengthening corporate responsibility. As many companies worldwide adapted to the standard, the global baseline was effectively raised.
Those who view it critically point out that the cost and complexity of compliance fall harder on small firms with fewer resources. Paradoxically, this may end up favoring large firms. There is also the criticism that, like the endless cookie consent pop-ups, only formal consent multiplies while real protection is scarcely felt.
Thus regulation is no cure-all but a delicate tool whose intentions and outcomes can diverge. Designing good regulation is as tricky as writing good code.
The Weight of the Word Consent
One of the core concepts GDPR stressed is consent. But what is consent? Is a button tapped without reading a long, difficult policy true consent, or merely a ritual with the form but not the substance?
A thorny question hides here. When there is in effect only one option — when refusing means you cannot use the service — how free is that consent? The freedom to refuse a tool everyone uses often means, in reality, not freedom but isolation.
So some point to the limits of the consent model itself. Rather than offloading the burden onto a single click by the individual, they argue, it is better to forbid risky processing outright or to mandate safe defaults from the start. Consent is a starting point, not a destination.
Why Privacy Is Valuable — and Whether It Truly Always Is
Here let us step back and pose a fundamental question. Why must we protect privacy? And is it always right to?
Voices in Defense of Privacy
The position that prizes privacy cites several values as its grounds.
Autonomy. Deciding for oneself whom to show what is part of being the master of one's own life. A life with everything exposed leaves less room to govern oneself.
Intimacy. We show different faces to different people. What we say to a friend, to family, to a stranger differs. Privacy makes possible the textures of these differing relationships. If everyone knew everything, the very concept of intimacy would vanish.
Room for growth and mistakes. People change. If a shameful word from the past were forever embalmed and trailing us, no one could learn and change with an easy mind. Forgetting is sometimes a mercy.
Balance of power. If one side can see fully into another, that alone creates an enormous imbalance of power. Privacy is also the minimal shield the weaker side can hold before the stronger.
Among these, autonomy is the one most often misunderstood. A common retort against defending privacy is that if you have nothing to hide, you have nothing to fear. But this narrows privacy to mere secrecy. We put a door on the bathroom not because we do evil there. Some acts are simply entitled to be alone.
Hearing the Skeptical Voices Too
Yet there are voices that question seeing privacy as an absolute good. Let us hear these fairly as well.
Some say that transparency, rather, builds trust. It is the view that a society in which people can know one another may be more honest. Others point out that when data sharing yields results good for all, as in medical research or public safety, excessive privacy protection can block that benefit. There is also the observation that the sense of privacy differs by generation. For those who grew up naturally sharing their lives, the boundary of privacy is sometimes drawn differently than before.
These voices do not mean privacy is unimportant. They simply show that the value of privacy must endlessly negotiate with other values — transparency, the public good, connection. No single value is absolute on its own.
The Age of AI: Boundaries Redrawn Again
The boundary of privacy, we said earlier, is redrawn each time a new technology arrives. The camera did so; the telephone did so. Now artificial intelligence stands in that place.
The question AI poses to privacy has a slightly different texture than before. Where the old worry was who sees my information, the larger problem now is what is inferred from scattered information. Facts I never disclosed — my health, my tastes, even my mood — can be guessed from other traces.
We might call this inferred data: information made about me though I never handed it over. The thorny part is that the old tools, such as consent and deletion, fit this inference poorly. How do you reclaim information you never gave?
Yet the same technology can stand on the side of solutions. Methods that learn on each person's own device without gathering data in one place, and techniques like the differential privacy we saw earlier, have grown more important, not less, in the age of AI. The tool that enlarged the problem also holds a thread to the solution.
Here too the conclusion does not converge on one point. AI is neither the enemy nor the ally of privacy. It is closer to a vast, neutral tool that changes shape according to the rules and defaults on which we choose to wield its power.
What Can We Do
Before a vast structure, it is easy to feel powerless. But there is neither complete powerlessness nor a complete solution. In between, there are things we can do.
First, awareness. Merely knowing the structure of the bargain hidden behind the word free makes us more wakeful users. Even if we cannot read every policy, the habit of recalling, just once, what we are handing over is a starting point.
No grand resolve is required either. Pausing when an app asks for location permission to ask whether it is truly needed, not using the same password on many doors — such small habits make a surprisingly large difference.
Next, choice. For the same function, if an alternative takes less of our data, we can choose it. Rather than thoughtlessly granting every permission request, the small restraint of giving only what is truly needed accumulates.
And voice. Privacy is not preserved by individual effort alone. Caring about and weighing in on which rules a society sets, demanding better defaults — these ultimately change the environment for all.
For privacy is less a wall built alone than a commons built together. However carefully one person guards it, a single photo posted by a friend may catch me in the frame. So privacy, in the end, cannot remain a matter for one alone.
Closing: Living on the Scale
The story of digital privacy is, in the end, a story about balance. We can neither give up convenience entirely nor lay down freedom entirely.
We live on this scale, making small decisions every day. Those decisions, gathered together, shape little by little the digital world we inhabit.
This essay did not tell you which way to move the weight. It only tried to show, clearly, that the scale is there and what sits upon it. The diagnosis of surveillance capitalism, the logic of surveillance for safety, the voices defending privacy, the voices doubting it — all are legitimate guests at this table.
A good society does not give the floor to one guest alone. The process by which many voices find balance amid tension is itself a mark of health.
Next time you thoughtlessly tap a consent button, why not pause and recall this scale. Not to settle the answer, but to hold the question.
A Small Self-Check
These are questions to lightly retrace what you have read. The aim is less to get the answer right than to notice how you yourself think. Try recalling the question first, then weigh it against the note below.
Question one. What was the largest difference between data and oil?
Note. Oil disappears once used, but data can be copied and grows explosively in value when combined with other data. The murkiness of its ownership was an important difference too.
Question two. What is the chilling effect, and why is it hard to capture in statistics?
Note. It is the phenomenon in which people who feel watched begin to censor themselves. A sentence not written after hesitation, a search not run, is recorded nowhere, and so escapes notice.
Question three. What was raised as the limit of the consent model?
Note. That when there is in effect only one option, and when the consent is tapped without reading the policy, it is doubtful how free or substantive that consent is. Consent is a starting point, not a destination.
Questions to Sit With
- For which convenience are you willingly handing over which data? Does that bargain feel fair?
- If you had to yield a little of safety or of privacy, which way does your weight tip? Why?
- How can the right to be forgotten and the value of being recorded be reconciled?
- Is privacy an individual matter, or a commons the community must protect together?
- If you lived in a village where every house was built of glass, how would your day be different?
- If you placed on a scale the good free services have brought and their price, which way does your scale tip?
References
- Stanford Encyclopedia of Philosophy, "Privacy": https://plato.stanford.edu/entries/privacy/
- Stanford Encyclopedia of Philosophy, "Privacy and Information Technology": https://plato.stanford.edu/entries/it-privacy/
- Encyclopaedia Britannica, "Panopticon": https://www.britannica.com/topic/Panopticon
- Encyclopaedia Britannica, "Jeremy Bentham": https://www.britannica.com/biography/Jeremy-Bentham
- European Union, General Data Protection Regulation (GDPR) official text: https://gdpr-info.eu/
- Nature, "Privacy and artificial intelligence" (research coverage): https://www.nature.com/
현재 단락 (1/175)
There is an old maxim in economics: there is no such thing as a free lunch. And yet we eat free lunc...