Only Joiners and Outsiders

In 2026, a Chinese runner broke the world half-marathon record long dominated by Black athletes, cutting the former record set by Ugandan runner Jacob Kiplimo by 12.06 percent.

The athlete was called “Lightning”: a new AI robot that won the second Beijing Humanoid Robot Half Marathon in 2026.

AI has defeated humans more than once. This marks the fact that AI no longer exists only inside GPUs and HBM. It has stepped into the world and is running without pause.

China is one of the centers of this enormous AI vortex. Robotaxis have been running in Wuhan and Shenzhen for years; drone delivery has been flying in Shenzhen for years. The existence of drivers, couriers, and many similar jobs is as fragile as a thin thread between two pieces of lotus root. AI will become stronger in the future.

You may have been living a calm, modest life, and then one day AI arrives uninvited. How should you face it: disgust, fear, indifference, or contempt? Human attitudes toward AI can be divided into two kinds: those who fear AI, and those who refuse to understand it. I fear AI.

The wave of AI cannot be stopped. This is called momentum. Human force cannot block it. When the giant wave comes, you either choose to surf on it at risk, or you are crushed by it. After calm reflection, you find that there are only these two choices. If you do not use AI and do not join AI, that does not stop others from developing and using it.

In front of AI, there are only two choices: join or resist. Will you join, or will you resist? You may resist for a while, but not for a lifetime. You should join. Those who join early receive the blessing; those who join late pay the price.

The later you join, the greater the loss.

For the Next 500 Years, AI Is Humanity’s Greatest Threat

What can AI do?

AI can change diapers, plan wars, slaughter pigs, sail ships, design houses, write sonnets, settle accounts, build walls, set dislocated bones, comfort the dying, obey commands, issue commands, cooperate, act independently, solve equations, analyze new problems, shovel manure, program computers, cook good meals, and perhaps one day fight well and die more bravely than humans.

(This is adapted from science-fiction writer Robert A. Heinlein’s description of people. I replaced the subject with AI.)

If AI can do everything, what remains for humans to do? Will AI take over all human work and let humans enjoy the result? Is that really what will happen?

If an omnipotent AI covers every path of human livelihood, that looks less like salvation and more like a threat.

AI threatens everyone.

Elon Musk is one of the most famous voices warning about AI. He has repeatedly called AI the greatest existential threat facing humanity, saying that humans may become AI’s “livestock.” He is far from the only person with such worries.

But here we will not begin with grand talk about the destruction of human civilization. We will not imagine the scenes from Terminator or The Matrix, with humans fighting AI machines. Let us first discuss the urgent matter in front of us: your livelihood and mine.

For the Next 10 Years, AI Is the Greatest Threat to Livelihood

Under the sudden assault of the AI wave, everyone’s rice bowl has become precarious.

New things have repeatedly killed old things at scale. When this happens again, do not rely on luck. Do not be the person who ignores AI. Do not be the textile worker who ignored the steam engine. Do not be the coachman who ignored the internal-combustion engine.

Artificial intelligence is coming. If you ignore it, what kind of end will you meet? You may fall like textile workers, coachmen, and typists, losing your livelihood and being left only with blind anger. Like workers whose living was taken by steam looms, you may have no outlet except smashing machines into parts.

Now the situation is worse. AI has arrived, and if you happen to be in an unfavorable position, you may not even be able to vent your anger through physical destruction. You can only swallow that rage like a blade. What a miserable thing that would be. Ignoring a new thing is the prelude to being crushed by it.

If you are a programmer, translator, designer, customer-service worker, or illustrator, you have probably already felt it: the next ten years may be a world in which AI takes human jobs at scale. If you have not felt this, it is not because AI cannot take your rice bowl, but because it does not want yours yet.

But you may say: I have a very good job. Why should I bother with AI? I am not like programmers or designers doing low-end work. I have an iron rice bowl. You may say you are a doctor, a lawyer, a remarkable scholar, a competent civil servant. AI may replace other people, but it is unlikely to replace me. I think the chance is small. I think AI basically cannot replace my profession. My rice bowl is harder than iron.

Are you really so certain?

What Gets Replaced Is Not Old “Low-End Work,” but Technological “Redundancy”

Who will AI actually replace? Perhaps you say: certainly not me. After all, I do not do physical labor. My job has a high technical threshold. It is so advanced and refined. It requires enormous stores of knowledge.

I need to correct this mistaken view.

Please note: AI will be like a storm. It does not destroy the solid foundation beneath the house; it destroys the complicated ornaments on the surface. It does not destroy old occupations with deep roots; it destroys the redundancies produced by human history.

In other words, AI does not harm old occupations. It specializes in harming new occupations.

The more high-tech and refined something is, the easier it is to replace. The more ancient it is, the harder it is to replace.

Strange, is it not? AI harms the advanced rather than the ancient? Does that mean a fortune teller on a county street may be safer than an engineer at a major tech company? Yes. In the AI era, that is exactly how it can be.

Several authoritative data sources and studies from institutions such as Anthropic and OpenAI reveal a clear trend.

The jobs most affected by AI are precisely high-paying, highly educated, white-collar “new occupations”; while many “old occupations” dependent on bodily skill and emotional interaction, such as cooks, have temporarily built a safer zone.

A typical high-risk, highly technical new occupation is programming.

In records of actual AI-tool usage, the share of programmers’ tasks that can be substituted reaches 74.5 to 89 percent, the highest risk among all occupations.

Designers, engineers, finance workers, accountants, and customer-service workers follow closely behind.

These jobs are often described as having concrete, explicit, and quantifiable workflows. On the surface, that is true.

But what few people notice is that these high-risk jobs in the AI era all lack communication and connection with people.

These occupations differ from truly ancient occupations. Their birth was not human-born, but object-born. Programming, for example, rose almost entirely with computers and the internet.

Most so-called advanced occupations are redundancies and bubbles in human development. Future generations may find that they existed for only a very short time in human history. The so-called ancient occupations point directly to human nature. They depend on people, not objects. Occupations that depend on people can endure; occupations that depend on objects bloom briefly and vanish. Human evolution is slow; the evolution of objects is fast. People remain; objects change. When the object disappears, the occupation that depended on it disappears too.

There are very few true old occupations: tool makers, food preparers, hunters, shamans, officials, soldiers. There are not many at all, unlike official occupational classifications that list thousands of kinds of work.

Truly old occupations, probably those older than the Qin and Han dynasties in China, will not be replaced because they depend on people.

Sex work is one of humanity’s oldest occupations, though not the oldest. If you are a competent sex worker, I congratulate you: you have obtained more stable income and better treatment than most civil servants.

If you are competent in that profession, you do not need to worry too much about aging, nor about AI being implanted into a robot with silicone skin and a European model’s face to compete for your clients. The essence of that profession’s access to resources is not between the legs, but in the ability to provide a fictional intimacy and admiration for others. Money earned through the body has a ceiling; money earned through fictional emotion has none. A truly competent practitioner understands this.

Returning to the title of this essay: in the AI era, a grassroots civil servant is almost certainly worse off than a sex worker. This depends on whether the job inside the government department involves links between people. Most people inside the system are easily replaceable screws. They lack social cognition and even basic common sense. They are mud rice bowls wrapped in a layer of tin-like metal, positioned as minor clerks handling miscellaneous tasks for the truly iron-rice-bowl officials.

A competent sex worker provides fictional emotional admiration; most people inside the system handle miscellaneous tasks.

If you spend your days processing documents, writing reports, filling forms, wiping tables, pouring tea, and printing files, then you are almost certainly worse off than a competent sex worker, whether in happiness or income. You are the floor; she is the ceiling.

A new occupation such as programming exists almost entirely by depending on computers. Programmers are translators between natural language and machine language. In the past, machines could not understand natural language, so people who understood machine language were needed to “communicate” with machines. Programming disappears not when computers disappear, but when machines can recognize natural language.

The programming languages you worked so hard to learn may soon be useless. Python may not be worth more than Thai or Hindi in the future. C++ may become almost equivalent to Nigerian languages or Esperanto. The original sin of programming is that it connects with objects rather than people.

And AI destroys precisely the link between people and objects. In the future, that chain will no longer need humans.

Do not think advanced work cannot be replaced. Do not think knowledge-intensive work cannot be replaced. Any occupational bubble of history is an object of AI replacement. The blind fortune teller in your hometown county who only finished primary school is not a historical bubble occupation. The major-tech-company employee who survived nine years of compulsory education, college, and graduate school is. The fortune teller survives by connecting with people. Have you successfully connected with people inside a big company? Have you realized that what determines your status and risk of being fired is not your professional ability, but the depth and strength of your human connections? If you realize this, you will be better off.

Whether your occupation may be replaced by AI does not depend on how advanced it is, nor on how long it takes to learn. It depends on whether it connects with people. If it does not connect with people, it will be replaced.

Living with AI While Afraid

There are many reasons for refusing to understand AI: blocked information, or a lack of courage to face reality. But the essence of fearing AI is only one thing: people fear being abandoned by society. The fear of AI has a traceable basis. AI seems to enter everywhere and do everything; this indirectly shows how large, broad, hard, and growth-oriented its capacity is.

In general, however, AI destroys the link between people and objects, not the link between people and people.

No matter your age, occupation, or identity, you may have to try AI. AI has evolved to the point where it can contract out almost all force-work for humanity. Confucius once said, “The gentleman is not a vessel.” In the AI era, if you are not such a person, you probably cannot survive.

Fortunately, since you have found this essay, it means you have already realized the weight of AI.

The existence of AI is essentially forcing humanity to return to the human itself, instead of serving as electronic labor.

In the view of Toyotomi Hideyoshi during Japan’s Warring States period, the standard for whether someone could be a warrior often did not lie in martial skill, but in their perception of momentum and their attitude toward death or change.

A true warrior must be extremely sensitive to the battlefield. Those who feel no fear and only charge blindly are dead soldiers; those who feel fear and therefore judge timing, weigh momentum, and search for opportunity possess the quality to command armies.

As carbon-based beings, the only capacity by which humans can surpass AI is the enormous potential activated by fear in the face of death: the potential of a desperate final struggle. Before death, humans can generate a sixth sense beyond reason. Fear is not weakness; it is an a priori warning that the logic of an era is collapsing.

Fear is the expert’s biological response to the giant wave of the age.

Those who refuse to understand AI, or cannot even feel the threat, have already been logically placed among those to be eliminated.

There are two attitudes toward AI: fearing AI, and refusing to understand AI.

I am certain the former is far better than the latter. If you have successfully developed fear toward AI, it means your understanding of AI has already gone to a certain depth. You have at least some sharp senses. You are already one large step ahead of those who refuse to understand it.

You have already recognized the changes AI will bring. You will unconsciously correct your behavior. Stop being physical labor or electronic labor. Admit your fear of AI, overcome your instinct to retreat, and dance with the future destroyer of humanity.

Those who fear AI, while fearing it, are precisely the ones who do not need to fear it. Those who fear AI do not need to fear AI; those who do not fear AI need to fear it.

Just embrace AI with your fear, as a warrior embraces death.

The Equality of Natural Persons and the Inequality of Social Persons

In ancient China people said, “Men have their share; women have their belonging.” In the language of Warring States thinkers, people without a “share,” or in modern terms without a social identity, are essentially absolutely equal.

Yet once men have a “share” and women have a “belonging,” once every person puts on the mask of social identity, inequality appears with it.

In the state of natural persons, all beings are equal. In the state of social persons, hierarchy is strict.

There has never been true equality in society. Many people try to define class: some by deposits, some by official rank, dividing society into a dozen layers. Westerners like to divide it into three: capitalist class, middle class, and proletariat. The old communists we respect simplify the world into the binary opposition of property and propertylessness.

But I think these divisions remain too superficial. The division of class should follow some older logic, and it should divide more simply, because simplicity means approaching essence.

One night an idea struck me like lightning: use the generating logic of the Ten Gods in Bazi. Under this division there are only five layers. Between every layer there is a qualitative difference; the gap between classes is difficult to cross, and communication between classes is nearly cut off. What follows is my conjecture.

A Wild Speculation

First Layer: The Effort Class (Peer-Star Class)

Participation logic: relying on physical strength to participate in social division of labor. This is the largest and easiest class to understand. In Bazi, peer stars represent hands and feet, peers, and physical force.

For people in this class, the core of survival is consumption: consuming muscle, energy, and time. They do not have many technical barriers and mainly exchange high-frequency repetitive labor for survival resources.

This class is the most common, and also the hardest, because its competitors are countless homogenized units of labor.

Second Layer: The Problem-Solving Class (Output-Star Class)

Participation logic: relying on mental ability and skill to participate in social division of labor. Eating God and Hurting Officer represent talent, output, and technique. People in this class no longer rely solely on physical strength; they survive by transforming a form of knowledge or skill.

Sample analysis: hairdressers and cooks. A cafeteria cook relying only on muscle memory belongs to the effort class, but a stylist or chef with aesthetics and the ability to reshape an image enters the problem-solving class.

Professionals: programmers, fate analysts, lawyers, teachers, and doctors.

People in this class are essentially solving complex problems. They are also parts in the operation of society.

Third Layer: The Capital Class (Wealth-Star Class)

Participation logic: relying on decision-making to participate in social division of labor. The core of the wealth-star class is not “how much money one has,” but “how one makes decisions.”

People in this class do not personally do hard labor, nor do they personally solve problems. They stand at forks in the road and use the resources in hand for games and decisions.

The essence of decision-making is judging the future and transforming risk into return. They are no longer parts of society, but a hub component that manipulates parts.

Fourth Layer: The Management Class (Authority-Star Class)

Participation logic: relying on risk and deterrence to participate in social division of labor. The authority-star class is the backbone of society. Many people think management means stability, but the opposite is true: the core of this class is risk-taking.

Management is essentially the exercise of power and the bearing of consequences. People in the authority-star class must play games within complex rules. They use the lever of power to organize production and control order.

This is a high-pressure life, and also a class with both predatory and protective attributes.

Fifth Layer: The Totem Class (Seal-Star Class)

Participation logic: relying on a spiritual totem formed by oneself to participate in social division of labor. This is the peak of human society. The seal star represents reputation, protection, learning, and faith. People in the totem class have gone beyond concrete affairs and decisions; they themselves have become a spiritual symbol.

Heads of state and religious leaders: they are the core in which the will of a group condenses, and they provide endorsement for power.

Promoters of social progress: scientists who not only produce results, but whose results change the dimension of human cognition.

If the theories of Einstein and Newton had remained only in the laboratory, they would merely be advanced problem-solvers. But when their names became synonyms for truth, they rose into the totem class.

Heroes: people such as Yue Fei, Yi Sun-sin, and Joan of Arc can gather the unifying will of a nation. Even when their bodies disappear, their spiritual totems continue to perform a function in social division of labor.

Bazi itself may be a class model.

Bazi is an ancient thing. Many people do not believe in it, but I believe the brain structure of ancient people was basically no different from that of modern people. Many things they conceived are not inferior to modern ideas. Especially because ancient social structures were relatively simple, they were less likely to be confused by the tangled symbols of the modern, late-modern, and postmodern worlds. I believe the designers of Bazi considered this when they created this small game.

Personal reflections, organized by AI.

Redefining Anxiety: It May Be the “Heat Death” of Your Inner System

I have noticed an interesting phenomenon: many people are anxious not because they are poor or poorly educated, but because they know too much. In thermodynamics, entropy measures disorder. In the logic of life, entropy is the information you have not converted into action. When you take in massive amounts of information every day and fill your mind with other people’s life logic while taking almost no action in reality, the serious imbalance between knowing and doing produces enormous entropy. Anxiety is essentially the mental depletion brought by this high-entropy state. The only road to renewal is to learn to reduce entropy and pull attention back from knowing to doing.

Bruce Lee’s Lesson: The Dynamic Balance of Knowing and Doing

On how to become stronger, I actually agree with a simple idea of balance: powerful people who are not miserable often reach a subtle equivalence between what they know and what they have done. Many people are capable yet painful because they know a lot and do a lot, but the two are completely out of balance. In truth, we do not need to know so much. Bruce Lee once said he feared the person who practiced one kick ten thousand times. That extreme simplicity and repetition is an efficient process of entropy reduction. If you only know and never do, you are increasing entropy and moving toward mental stagnation. Once you begin to act, even if you know a little less, you gradually move toward a clear and unanxious state.

Why Entropy Reduction Should Begin with the Outer World

I often advise friends who feel mentally chaotic not to rush into solving inner contradictions. First, clean the physical environment. I have found a logic: the entropy of the inner world is often hard to operate on directly; it is like a final boss. But the outer and inner worlds communicate with each other. Many people try to reduce entropy through advanced psychological construction, and they often fail badly. The best method is to begin outside and influence the inner world by simplifying the living environment. This is what I call simplicity. When your physical space becomes clear and orderly, the signal-to-noise ratio of your inner world naturally rises, and you have energy left to handle deeper anxiety.

Throw Away Ten Things Every Day: A Practical Exercise in Entropy Reduction

I once heard a very Japanese-style method: force yourself to throw away ten things at home every day. It sounds extreme, but behind it is training the ability to let go. The process of discarding objects is essentially a process of reexamining your relationship with things. When you find old items and ask yourself whether they still matter, you are actually clearing your own paths of thought. This abandonment lets you understand which things are indispensable in life and which are merely waste bought by following the current. The more you discard, the lighter your system becomes. That lightness turns into a sense of control and makes you more composed when dealing with complex problems.

Steve Jobs’s “Uniform”: Decision Cost Behind Extreme Simplicity

Look at Steve Jobs: for years he wore the same style of clothing and drove the same style of car. This was not merely personal eccentricity. I prefer to understand it as a highly intelligent entropy-reduction strategy. A person’s willpower each day is limited. By fixing these trivial parts of life, he minimized meaningless information and choice cost at the physical level. He understood simplicity: by constantly abandoning the nonessential, he cleared a pure path and focused all energy on product and innovation. This low-entropy state was a core weapon that let him lead Apple against high-entropy competitors.

The Poverty Code in a Room: The Link Between Order and Wealth

I occasionally observe the homes of friends from different social levels. I find that people who can control their own fate often have fewer unnecessary objects at home. By contrast, some friends work very hard, but their homes are always stuffed with plastic bags, expired medicine, and dusty old chargers. This reflects a lack of the virtue of simplicity. If unused things continue to exist, they continuously produce disorder in your space. The ancient saying “If one cannot sweep one room, how can one sweep the world?” is quite reasonable. If a person cannot maintain simplicity in the most basic physical environment, then when facing complex social and economic games, they will probably fall into a chaotic state of wanting everything and doing nothing well.

“Purity” in Fate Theory and Energy Focus in Reality

I remember that in some traditional systems of fate analysis, there is a pattern called “purity.” People with such a pattern often achieve something. This corresponds to focused energy. When a person has too many goals and scattered desires, their energy is dispersed and collapses at a touch. By contrast, people who gather energy into one point often achieve breakthrough. Just as having too many stars representing desire while one’s own will is too weak can lead to poverty, the only way to avoid this trap is to actively delete unimportant goals and prune unnecessary preferences. Less is more. Only by clearing out a single path can life have a real eruption.

The Underlying Truth of Sudden Success: A Byproduct of Completed Entropy Reduction

We often see someone suddenly become successful, but I am more inclined to believe that sudden success is the natural result of entropy reduction accumulated to a certain degree. Experts are usually doing subtraction. Before the eruption, they are in fact constantly using elimination to remove choices that look tempting but are actually traps. This accumulation process is entropy reduction. When you sort out the chaos around you and gather all energy in the most correct direction, the so-called eruption is merely a small gift that comes when water reaches the channel. So instead of longing for sudden success, first clarify what can be abandoned now and what must be persisted in.

Who Receives Luck? Simplicity Is the Highest Form of Low Entropy

Sometimes I feel that luck is mysterious, but it seems to prefer simple people. Here, simplicity does not mean stupidity. It means a low-entropy state: low expectation, focus on what one likes, and awakening through practice. Complex people often have bad luck because their systems are too chaotic, full of calculation, entanglement, and fear. These are all high-entropy disturbances. Simple people, like the Indian tea seller Dolly Chaiwala, simply practice one movement to the extreme and turn selling tea into an art. This extreme purity attracts opportunity. Protect your simplicity; do not let excessive calculation damage your luck.

Give Yourself a Chance at Renewal: It Is Never Too Late to Begin Reducing Entropy

Every person is born low-entropy and full of vitality, loved like an infant. But as time passes, we often let dust cover the soul and fall into high-entropy tension and fear by our twenties or thirties. Still, I believe a person’s future is connected only to the present. If one day you look back and feel that your life has been too tiring and inauthentic, perhaps it is only because you have always been adding. Forget so-called background, family conditions, or past failures. Begin subtracting now, and learn the virtue of entropy reduction. This active reconstruction of life order is itself a powerful life force.