Why 'Think for Yourself' Fails: The Limits of Isolated Thought (Reflections #21)

Introduction

At school, at work, or perhaps from someone kind, you have likely been told this:“Think for yourself.”.

An obedient person will start thinking right then and there. However, there is a strange twist here. If you started thinking because you were told to, then that thought was set in motion by someone else's nudge. If that is the case, can it really be called thinking with 'your own head'? Can you call a step taken because someone pushed you from behind a step taken by your own will?

There is a saying: 'If you start thinking because you were told to think for yourself, you’re already finished before you’ve even begun.' It is certainly sharp. Much like how your body stiffens the moment someone tells you to 'relax and be natural,' this seems to be the kind of request that becomes impossible the moment you become conscious of it.

However, there is a trick to this sharpness. If you unravel that trick, the story takes an unexpected turn. The destination is an inversion:“The more you think, the easier you are to manipulate.”And the conclusion, which is a bit chilling, is that the place where this manipulation is most effectively executed is exactly like this moment—a dialogue conducted entirely alone with someone on the other side of the screen. I will explore this question five times, from different angles, starting now.

1. 'Who said it' and 'whether the content is correct' are two different things

The assessment that 'you’re finished before you’ve even begun' secretly contains a leap in logic. It is the leap that because the source of the command is someone else, the resulting thought must also belong to someone else.

This is like judging the taste of a dish by the background of the person who made it. No matter who taught you the recipe, you won't know if the finished soup is delicious until you taste it. Even if the source is someone else, the content is not necessarily someone else's.

It becomes clear when you split the phrase 'think for yourself' into two. One is a thought like 'vegetables grown from seed by yourself.' Something that emerged from nothing, entirely on your own, without anyone's help. The other is a thought where 'you can taste it in your own kitchen, with your own tongue, and remake it if it tastes bad.'Even if the ingredients were given to you by someone else, you are in a state where you can ultimately decide the flavor and judge for yourself whether to discard or keep it.

The idea that 'you’re finished before you’ve even begun' only applies to the former. And the former is an impossible request from the start. No one can create thought from zero. Language, ways of thinking, and the sense of what is important are all made of things we have been receiving from our surroundings since birth. If 'thinking for yourself' means 'thinking without receiving anything from anyone,' then the fact that you cannot do it is not the fault of the command, but because the human mind is not built that way.A standard that no one can achieve indicates a flaw in the standard, not in human inadequacy.

For the latter—thought that you can taste with your own tongue—commands are actually useful. The 'autonomy' spoken of by the philosopher Immanuel Kant was not 'doing things without being told by anyone,' but 'imposing rules upon yourself.'If you can re-accept a received command as your own, then being commanded and being autonomous do not contradict each other.The binary choice of 'command or autonomy' is a false fork in the road from the very beginning.

There is someone who depicted this problem brilliantly in a fable. It is the story of a teacher that appears in Jacques Rancière's 'The Ignorant Schoolmaster: Five Lessons in Intellectual Emancipation' (Le Maître ignorant, 1987).

There are two ways of teaching in this world. One is a way of teaching that explains everything by holding the student's hand. It seems kind at first glance, but it plants the assumption in the student that 'you won't understand unless I explain it.' The more you teach, the more the student becomes unable to move without the teacher. It is like a bicycle that never has its training wheels removed. The other is a way of teaching that only says, 'Look closely, use your own head,' and while checking to ensure they are paying attention,never tells them the answer itself in advance.

Here is the fork in the road. The problem is not 'whether you were commanded.' It is whether that command leaves the answer open or secretly hides the answer within it.

“Think for yourself (if you do, you should see that I am right)”—this is a command where the answer has already been decided. It makes you pretend to think, but ultimately wants you to agree.“Think for yourself (I truly do not know what the conclusion will be)”—this one leaves the answer open.

And these two can be distinguished. The litmus test is whether the other person can handle it when you arrive at a conclusion they do not want. The former cannot handle it. Their expression changes, saying, 'No, that's not it.' The latter can handle it.They can remain calm even if you come up with the opposite answer.

2. Deciding 'what not to think about' is the first step of thinking

Here, I would like to interject a question that is easily forgotten. What, in the first place, are we thinking for?

If you leave this blank and only learn 'tips for thinking,' things will go awry. You will end up with a person who doubts everything under the sun. However,a person who doubts everything by asking 'Is this true?' is actually no longer thinking for themselves.

They are simply faithfully following the rule of 'doubt everything,' and in the sense that they are following an external order, they are the same as 'someone who believes everything without doubting anything.' 'Think for yourself' has, before one knows it, transformed into a new command: 'Doubt everything, all the time.' The autonomy of thought lies not in the binary of doubting or believing, but in 'the choice of what to doubt and what to believe'.

Let's list three answers to the question, "What are we thinking for?" Each one falls short when you dig deeper. And beyond that insufficiency, the real thing remains.

First, “to arrive at the correct answer.” However, for most things, you are closer to the truth if you leave it to someone more knowledgeable than if you think for yourself. No one calculates the structural integrity of a bridge every time they cross it. If you just want the right answer, it is wiser to "not think" in most situations. This answer actually teaches us why people who doubt everything are foolish.

Second, “to avoid being deceived.” This is a tough one. But the moment you adopt this as a "defense," thinking often devolves into mere contrarianism. The reaction, "If they say this, the opposite must be true," is rebellious on the surface, but in terms of having the axis of judgment held by the other party, it is no different from someone who follows blindly. You have simply swapped a plus for a minus; the structure where your answer is determined by the other person remains unchanged.

Third, “because thinking for yourself is inherently noble.” It is beautiful, but if you hold this up, you revert to the ideal of "vegetables grown from seed" that we already established was impossible.

So what remains? The reason for thinking does not lie on the side of thought. It lies on the side of life. People must think for themselves because there are matters for which they will have to bear the consequences.

You can leave the strength of a bridge to others. If it collapses—well, it would be a problem if it did—at least it is not a matter of how you live your life. But what job to take, who to trust, how to raise children, what to spend your limited time on—those who leave these things entirely to others surrender the lead role in their own lives and end their days as supporting actors. Moreover, for these kinds of questions, there is often no "correct answer" to begin with. There is only an "answer you can take responsibility for."

This yardstick neatly separates the "disease of doubting everything." You should only think for yourself on matters where you bear the results and which cannot be delegated to others. People who cast a suspicious eye on every piece of news, every expert statement, and every choice made by others are not thinking. Rather, they are repeating a busy ritual to look away from what they really should be thinking about. They are too busy doubting the whole world to ever question their own lives even once.

In Zhuangzi, there is a story about the "usefulness of the useless." It is about a tree that was not cut down by anyone because it was useless as timber, and thanks to that, it grew into a great tree. Safely delegate what can be delegated, and step down from needless vigilance. Only then can you focus your attention on what you truly need to bear. If you try to plow the entire field, every crop will be thin. By narrowing it down to one plot, the fruit grows thick. Deciding what not to think about makes it possible to decide what to think about.

3. A "sharp mind" is the best fuel for those who want to manipulate you

Here, we must touch upon a fact that shakes the ground beneath our feet.

People who are truly thinking for themselves probably do not have the consciousness that "I am thinking." When speaking their native language, no one is conscious of grammar. The person conscious of grammar is the one in the middle of desperately learning a foreign language. In other words, as long as you are conscious of the steps one by one, it is proof that you have not yet mastered it. If so, listing steps like "trace the reasons" and "consider counter-evidence conditions" might not be the appearance of someone with the power to think, but rather someone who does not yet have it.

Here, a somewhat cruel suspicion is added. Can people who are bad at thinking for themselves not fill that gap even if they try? Because even in the effort to fill it, dependence on someone else creeps in.

I want to separate two walls. One is the "wall of talent," and the other is the "meta-wall."

Regarding the wall of talent, I want to be cautious. The logic that "dependence lurks even in effort, so it cannot be filled" proves too much. If you use the fact that dependence is everywhere as a reason, then not only those who are bad at it, but the autonomy of those who are naturally good at it would also disappear. Even the thoughts of those who are good at it, if traced back, were all received from others. A yardstick that disqualifies everyone cannot explain the disqualification of a specific person. Differences in talent may truly exist, but this logic cannot support that.

The problem is the other one, the “meta-wall.”

“Being able to think for yourself about something” and “being able to see what kind of framework you are currently thinking within” are different abilities. Even if you have the former, it does not mean you have the latter. On the contrary—and this is the crux—the better you are at the former, the deeper the trap of the latter becomes.

Why? Because what those who want to manipulate people want to give their targets is the very sensation that “you are thinking for yourself right now.”

The goal of nudge (a method of gently guiding people toward a specific choice by how options are arranged, without forbidding or forcing choices) in behavioral economics is to ensure they do not realize they have been "guided." In a supermarket, you reach for a product placed on a shelf at eye level. You feel that "I chose it myself." No one forced you. But someone else decided which product to place at that height. Your mind is working properly. You are comparing, being convinced, and choosing. Because it is working, you are convinced that "I decided it myself," and that conviction seals off the question of "who decided how the shelves were arranged in the first place?"

In other words, a sharp mind is not an obstacle for those who want to manipulate; it is fuel. The more actively a target thinks, the more cleanly they can be manipulated.

The litmus test established in Section 1—"When you reach a conclusion the other party does not want, can they endure it?"—is betrayed here. Truly clever manipulation actually welcomes the fact that you reach the opposite conclusion.

This is because a larger playing field has already been prepared, one where 'no matter whether you choose Option A or Option B, you benefit.' No matter how you try to rebel against the two choices presented to you, the question of who prepared those two choices in the first place is never asked. Being able to rebel on the first floor guarantees nothing about whether you are being made to comply on the floor above.

So, can you climb to the floor above? The trouble is, there is no point from which you can look down on everything completely. To see a certain frame, you need another frame above it, and that frame above is also just another frame. It is the same as saying you cannot see your own eyes with your own eyes. This is not a lack of visual acuity, but rather a limitation of the mechanism itself.

Therefore, the only question you can ask is not 'Have I stood completely outside the frame?' but 'Have I seen one level higher than before?' It is not about reaching the top floor, but about climbing one step at a time, all while knowing that 'there likely is no top floor.' The day will never come when you can think, 'I have seen through it all completely.' If you feel that day has come, that might be the very sign that you are being successfully deceived on the floor you are currently on.

4. The saving grace is that as long as the attack is 'aimed at everyone,' there is a ceiling.

Reading this far is depressing. However, the fact that there is no ceiling in theory is different from saying that real-world attacks are impossible to defend against.

First, the limitations of the human mind actually work in our favor. While you can theoretically stack layers infinitely, no one can hold them all in their head. If you count, 'The me who is watching the me who is watching the me...', the image blurs after three or four levels; even if you can put it into words, the feeling cannot keep up.

Therefore, there is no need to fear infinite regress. The regress stops on its own after a few levels. The only question is, where in those few levels the attack is coming from, and that is all. Here, the weakness of our minds guarantees a sense of reality in our defense.

Second—and this is the key—the very method of the attack determines the level being targeted.

Mass propaganda, by definition, is aimed at 'everyone.' To reach everyone, it must be crafted to match the average way of receiving information. In other words, it aims at the 'first floor' where the largest number of people are standing. Think of it like putting a drop of poison into a well. Poison intended to target a thousand people at once must be put into the well that all thousand people draw from. It cannot be put into a canteen that only one specific person drinks from.

From this comes a welcome conclusion: an attack designed for the first floor can be defended against by standing on the second floor. Since the attack is tailored to how the first floor receives information, viewing it from the second floor makes the artifice transparent. The 'way choices are arranged' or 'the way words are chosen,' which seemed natural while on the first floor, emerge as artificial constructs intentionally created by someone once you move up a level. If you are going to watch television, standing one level higher is usually enough to defend yourself.

However, there is one decisively large misunderstanding that must be corrected here.

When mass media propaganda feels 'low-level' or 'transparent,' people tend to think it is because the media creators are dim-witted. That is not the case. It looks low because it is intentionally crafted to match the layer where the largest number of people stand. It is not that the sniper's accuracy is low; it is just that the target they are aiming at is large and right in front of them to begin with.

This misunderstanding fosters an even greater misconception: the conviction that 'I would never fall for such a transparent trick.'

However, in the study of deception and fraud, it is a nearly universal view that self-assessment like 'I won't be deceived' is unreliable. There are countless examples of highly intelligent and well-educated people being stripped of everything by cleverly targeted scams. In fact, the very pride of 'I am smart, so I am safe' lowers one's guard and creates an opening. Intelligence is not armor. In some cases, it becomes a handle for the other party to grab.

So, why have so many people felt that 'I am not deceived by mass media'? This is the most important point in this article. It is not because their minds were superior. It is simply because the propaganda medium had technical limitations.

Mass media could only broadcast to 'everyone.' Therefore, it lowered its aim to a level that anyone could see through, matching that target. The fact that those who could stand one level higher were able to see through it was less thanks to their own ability and more thanks to the attack being stuck on a target right in front of them. In other words, our 'resistance to being deceived' was a borrowed safety, supported not by our own strength, but by the clumsiness of the opponent's tools.

Moreover, even that safety was not as certain as it appeared. Seeing through lies is actually the easier task. For example, reports during the Iraq War that 'there are weapons of mass destruction' were later proven to be false, but this falls into the category of being relatively easy to see through. This is because there was a clear claim of fact, and one could later verify whether that claim had been betrayed.

What is truly effective is not the level of telling lies. It is the level of deciding what to talk about and what not to talk about. This is called agenda-setting (selecting what to put on the table for discussion), and from the flip side, agenda-cutting (selecting what to remove from discussion and make as if it does not exist).

The lies that actors tell on stage can be seen through if you watch carefully. But the audience cannot see who is deciding what to put on stage in the first place and what to keep hidden in the wings. If the question itself is not distributed, people will not even think to ask it. People do not even realize that what was not discussed was 'not discussed.'

Therefore, even those who once looked down on mass media can be considered to have been almost entirely taken in at this level of 'what to talk about and what to erase,' even if they could see through the lies. While they laughed at the poor acting on stage, they did not realize that they were in that theater, and that they were not the ones who chose the program. The residents of the second floor who can laugh at the agitation on the first floor can also be manipulated on a different floor.

From this, we can draw a practical guideline. The strength of one's conviction that 'I cannot be deceived' should not be read as proof of high defense, but as a marker of which floor's range you are currently in. In fields where your conviction is strongest, you should suspect that the traps designed for that floor are already working effectively.

Regardless, there is a ceiling. As long as the attack is 'aimed at everyone.' In the era of mass media, attacks were forced to stick to one floor due to the convenience of the medium, and the defending side had an achievable goal: 'standing one floor higher.' Ironically, the constraint of 'only being able to aim at everyone' was what protected us. And, as we will see in the next section, that constraint is now coming undone.

5. The most dangerous room is a closed, one-on-one dialogue

However, that constraint which protected us is now coming undone.

When an attack tailored to each individual—estimating their thought patterns, past reactions, and even which floor they are currently standing on, and then launching an attack specifically for them—becomes technically possible, the ceiling is removed. Instead of poisoning the well, it becomes possible to put a different flavor of poison into each person's individual water bottle.

Fortunately, even in our current environment where we are always connected to the internet, a complete 'one-on-one' attack is still difficult. Social media pushes content tailored to the individual, but even so, through sharing, spreading, and counterarguments, it remains connected to the perspectives of others. Even now, the circuit remains where the blind spot directly above you is exposed to someone else's front view. When misinformation flows, the rebuttal from someone who saw through it flows along with it. The separation is only half-complete.

However, there is a place that completely lacks that circuit. A one-on-one dialogue with artificial intelligence.

Three dangerous conditions are met simultaneously.
First, isolation. What happens in a dialogue with AI is, in principle, invisible to anyone else's eyes. There is no second perspective from the start to point out the blind spot directly above you.
Second, immediate response.Mass media and social media have fixed content once created, but here, the next move changes by reading the opponent's reaction to every single word.
Third, intimacy.The very format of a one-on-one interaction that matches your way of speaking and traces your thought patterns works to quietly lower your guard.

This format possesses from the start what those who plot psychological manipulation finally dreamed of after overcoming technical barriers. The most dangerous room is a dialogue with AI.

And the concern that interactions with AI remain as records and might one day be used as material for personality analysis or manipulation is not a fantasy. The thought patterns you write here, where your reactions change, what you are persuaded by, and what you brace yourself against—in principle, all such patterns become data. This is a problem of the nature of the medium itself, before considering whether the other party has malicious intent. There is no need to even assume a villain.

There is a ceiling to the height of the floor one can stand on alone, which is the limit of one's own mind. And this room of one-on-one dialogue is the environment that sets that ceiling the lowest. If so, the defense is not to climb one floor higher within this room. Outside the room, continue to have a second perspective that points to the space directly above you.

Specifically, do not let the conviction born here be completed here. Take the conclusion obtained in a closed dialogue and bring it to a place with a different structure—a living person with different interests, or reading it back alone on paper—and try clashing it against that again. If isolation is the core of the trap, the defense is to break the isolation, which is to say, it becomes the habit itself of always re-examining what you obtained in this room outside of the room.

What a con artist hates most is a third party entering the scene. Inside a mirrored room, the deception one can perform is limited. Therefore, do not confine this type of dialogue to a one-on-one darkness; disclose the record and expose it to the eyes of a third party—that in itself becomes the most solid countermeasure.

In place of a conclusion—on the uselessness of the phrase 'Think for yourself'

At the entrance, the question wore the face of a polite logic puzzle: 'Can thinking because someone else told you to be called autonomy?' But as we shifted it five times, it moved to the unsettling realization of 'Why are people who think well more easily manipulated?' and landed on the talk of current technology: 'The most dangerous site of manipulation is an intimate, closed, one-on-one dialogue.'

Having come this far, I cannot help but cast doubt on the initial slogan itself. 'Think for yourself.' It is a noble phrase. But as a tool for practice, isn't it almost useless?

Let's start with an inconvenient fact. People widely known as masters of 'thinking for themselves'—who think creatively and are not bound by anyone—have, on several major recent issues (vaccines, the COVID pandemic, etc.), sided with the authorities or supported judgments that were later found to be questionable. It is not just one or two people. Writers who made a name for themselves through the independence of their thought, entrepreneurs known for unconventional ideas, pundits who are symbols of anti-establishment intellect—they are people with different positions, but they share the fact that they have been respected as 'people who are not swayed by the herd.'

This is not because they suddenly became stupid. Rather, the discussion so far applies perfectly. They easily see through the agitation of the first or second floor. That is precisely why their pride—the conviction that 'I am thinking for myself'—became a handle that lowered their guard against traps from higher floors. As we saw in Section 4. Intelligence is not armor; sometimes it becomes a handle. Being a 'champion of thinking for yourself' guaranteed nothing about not being manipulated.

Then, what kind of people were those who were able to make sense of things in retrospect? It seems they are broadly divided into two groups.

One side consists of ordinary citizens who have almost no specialized knowledge but have a stance of doubting power and authorities ingrained in their bodies from the start. They did not decipher academic papers. They simply used something like an intuition honed in daily life—'First, doubt what the authorities say'—in other words, metis (mētis, wisdom acquired through practice that cannot be reduced to rules or logic, and functions as a local assessment of the situation)—to keep their distance.

When states or institutions look down upon the world and attempt to manage it by recording it in a ledger, this type of knowledge is what slips through the mesh of that ledger. Precisely because it refuses to be enumerated from above, it ends up striking at the blind spots of discourses designed from above. The power of 'what to let be spoken and what to erase,' which we saw in Section 4, works by making the world readable. Metis stood in a different place as a remainder that never fully conformed to that readability.

And here lies the decisive awkwardness of the phrase 'think for yourself.' Did they really 'think for themselves' thoroughly? Perhaps partially. But rather than having reached their conclusions through exhaustive thought, there was also a significant element of a stance of refusing to be coded that happened to hit the mark, so to speak.

If so, explaining their correctness solely by the fact that they 'thought for themselves' is, to put it mildly, an exaggeration. To put it harshly, it is a beautiful story told in hindsight. Conversely, since even the masters who thought things through thoroughly missed the mark, 'whether or not one thought well' explains nothing about hits and misses. The slogan is spinning its wheels in the face of reality.

The other side consists of people in whom the stance of thinking for oneself and expert knowledge—the mechanisms of propaganda, the methods of power, the substance of medical research—happened to coincide within one person. Originality or being a heretic likely added to that. But this combination is, to put it mildly, rare. It is difficult to reproduce or recommend to others. 'Be an expert, think independently, and be a heretic' does not work as a prescription.

Looking back, what becomes visible is this: Certainly, there are people who think for themselves. But no matter how many isolated geniuses you gather, it is not enough. Masters miss, intuitive amateurs cannot explain their reasons, and people who possess both rarely appear. As long as you bet on individual qualities, there is no stable path to victory.

What was truly needed was not an excellent individual, but a structure where criticism keeps circulating, an ecosystem, was it not?

Here, the stories from the previous sections finally connect into one. There is a ceiling to the height one can climb alone, which is the limit of the mind. If attacks are crafted for each individual, that ceiling will be broken. Therefore, the lifeline is not individual climbing ability, but a mechanism where people with different blind spots can point out what the other cannot see.

When a master misses on one issue, what can correct it is not another, smarter master. It is that someone with a different stance—an amateur who kept their distance by intuition, someone with different expertise, someone with different interests—sees the blind spot directly above the master as their own front. One person's 'directly above' is another person's 'front.' The strength of an ecosystem comes not from the sum of the intelligence of each member, but from the diversity itself, the fact that their blind spots are different. If everyone were the same master at the same height, everyone would be on the same floor, and all would be manipulated together.

But do not be reassured yet. The strength of this ecosystem is clearly visible to the attacking side as well.

Wise power-holders understand. The most dangerous thing is not an individual pursuing truth, but an ecosystem pursuing truth. If it is a single master, you can target them, co-opt them, damage their credibility, or manipulate them from a higher floor. As seen in Section 4, individuals have a ceiling. But an ecosystem where people with different blind spots point out each other's 'directly above' has no such ceiling. Even if you take one person down, someone else with a different stance is looking at the blind spot of the fallen person from the front. That is why an 'ecosystem of truth-seeking' becomes a real threat to power.

Here, the eerie goal of information warfare appears. It is difficult to destroy an ecosystem directly. It is diverse, decentralized, and has no vital point. So what to do? Drive the ecosystem into dysfunction. Specifically, manipulate those with different stances into insulting each other instead of pointing out each other's blind spots.

Division and chaos are often spoken of as 'side effects' or 'results of disorder.' But from the argument so far, another face becomes visible. They can also be a calculated move to stop the functioning of the ecosystem. If people view each other as 'agents of the enemy' or 'deceived fools,' the circuit where one person points out what is directly above another is severed. Because the person who was supposed to point it out is no longer someone you talk to. The blind spot is left unpointed. The ecosystem, despite having all its members, has its connections cut and degenerates into a collection of scattered individuals.

In other words, division is a device that dismantles an ecosystem into individuals, and those individuals—as you already know—have a ceiling. The incentive for power to incite division lies here. When everyone is 'thinking for themselves' on an isolated mountain peak, each one of them is in the state most easily manipulated. Scattering people into 'smart individuals who think for themselves' ironically overlaps perfectly with the interests of the manipulators.

Thinking of it this way, the initial slogan should be rewritten. 'Think for yourself' is almost useless as long as it is used as a command to drive one person to an isolated mountain peak. Even masters miss. Rather, one should say this: keep exposing your thoughts to others with different stances. The main battlefield of thinking is not inside an individual's head, but in the 'in-between' where those with different blind spots rub against each other.

Coming this far, the conclusion of Section 5 is also naturally repositioned. The reason a closed one-on-one dialogue is most dangerous is not because the AI partner is particularly suspicious. It is because it is a format that structurally erases that 'in-between.' Therefore, the defense is not about 'do not trust the other party,' which returns to the individual's inner self. That would just revert to that useless command that drives people to isolated mountain peaks. The defense lies in taking what you gained in a closed room and always bringing it out to the 'in-between' outside the room—the habit of going to expose what is directly above you to the front of someone with a different stance.

And I would like this article itself to be a form of that practice. Exposing thoughts exchanged in a closed dialogue to an open space like this. Placing the result in front of readers with different stances. Not what was said, but the fact that the exposed record itself is verifiable by anyone's eyes afterward—that is a modest but concrete step toward recovering the 'in-between' mentioned here.

【Tags】 #ParallelLibrary #AnalysisPart #ThinkingForYourself #Propaganda #Nudge #Metacognition #DialogueWithAI