Part V: Choice — The Road and the Person
The first four parts were about dismantling. Dismantling the manufacture of narrative, the flow of costs, what AI uncovers when it lifts the veil, the structure of consciousness and capability. After the dismantling, everything lies spread out on the table — mechanisms laid bare, illusions debunked, what remains has been located.
This part dismantles no more. It sets out to answer two questions: in a world you have already seen through, what do you choose to do? And — after choosing, how do you walk the road?
Chapter 14: Why "The Benevolent Have No Enemies" Fails — and Why It Still Matters
An Answer That Cannot Stand
If someone asks: in the AI era, what should a person learn, what should a person do?
Four characters rise to the surface: ren zhe wu di — the benevolent have no enemies.
Then you immediately discover that these four characters can barely stand today.
Not because they are wrong. But because the premise they depend on — that good and evil have clear contours — has been dismantled over the past several decades.
The benevolent figures of antiquity appeared in environments with distinct contours: war, famine, tyranny, invasion. The enemy was visible; the direction was broadly clear. But most of today's problems have no absolutely evil person to point to. AI replacing jobs — companies pursuing efficiency is not wrong. Sky-high property prices — investors seeking returns is not wrong. Declining birth rates — young people choosing not to have children is not wrong. Capital concentration — the system's operating logic dictates it. Algorithms controlling attention — platforms optimising user engagement is business common sense.
Each gear, viewed individually, is reasonable. Together, the machine grinds people to pieces.
In this kind of environment, "benevolence" is no longer a question of moral judgement but a question of cognitive ability. You need first to understand how institutions shape individuals, how environments compress the space of choice, how incentive mechanisms distort judgement — and only then can you talk about "treating others well." This is not about exonerating the wicked; it is about finding the machine's blueprints instead of raging at the gears.
But choosing to treat others well after understanding the causes — this step is always the person's own decision.
Two Ways the Benevolent Die
Through long observation of these four characters, two most dangerous possibilities ultimately crystallise.
The first: you think you are benevolent, but you are not. You are merely a self-satisfied hypocrite. This possibility always exists and cannot be verified externally — a true hypocrite and a benevolent person who occasionally errs can behave identically. But there is one criterion: a hypocrite will not seriously question whether they are a hypocrite.
The second is more chilling: you truly achieve "benevolent and without enemies," but you die with nothing left.
In an environment where narrative power is monopolised, right and wrong, black and white, can be easily obscured. In the workplace, the most ruthless people are protected by the system because they produce the prettiest numbers fastest. Managers prefer executors without boundaries because they maximise short-term performance. Those who control the narrative can easily define the benevolent as "inefficient" or "insufficiently decisive," while the most ruthless and thorough are described as "having vision" or "trustworthy."
The benevolent are not invincible; the benevolent are the most easily sacrificed. Failing to acknowledge this reality turns "the benevolent have no enemies" into nothing more than a self-consoling slogan.
Five Blows from Utilitarianism
Having acknowledged this reality, one must still pass through an even more severe tribunal.
Utilitarianism asks only one question: outcomes. It does not ask about motive, character, or inner peace. The moral value of an action is determined entirely by the degree to which it brings the greatest happiness to the greatest number. This logic happens to be AI's native tongue — the essence of an algorithm is utilitarian calculation: maximising user engagement, click-through rates, and profit.
Place "the benevolent have no enemies" on utilitarianism's tribunal. Can it hold up?
Below are five attacks utilitarianism can launch. Every one of them lands.
The first blow: the benevolent usually end badly, and their failure leads to even greater suffering. Confucius travelled the kingdoms for fourteen years without successfully implementing his governance philosophy in any of them. During his lifetime, he did not help a single person in the "present."
Utilitarianism's verdict: had he been more politically savvy, more willing to accept imperfect compromises, he might have actually saved more people. The benevolent pursue the perfect "Tao" at the expense of the imperfect but effective "technique."
The second blow: the benevolent are frequently exploited, and the result of exploitation is that the wicked gain more resources. The benevolent are sacrificed; the wicked are promoted, gain more power, and influence more people.
Utilitarianism's verdict: the existence of the benevolent objectively provides nourishment to the wicked. Your goodness is not neutral — it is a resource being consumed, and those who consume it use it to strengthen themselves.
The third blow: "the benevolent have no enemies" is a form of cognitive anaesthesia. These four characters let people find comfort in failure, reducing the motivation to change reality. "I may be oppressed, but I am benevolent, so I am invincible" — this mentality is essentially the same as "compensation in the next life" in religion.
Utilitarianism's verdict: if "the benevolent have no enemies" leads you to choose endurance over resistance in the face of injustice, it is a tool for maintaining the status quo.
The fourth blow: AI-era data supports utilitarianism. Global poverty rates have dropped dramatically; average life expectancy has risen dramatically; literacy, infant survival rates, educated populations — nearly every indicator is improving. These achievements are not because of "benevolence" but because of market mechanisms, technological progress, and profit-driven incentives.
Utilitarianism's verdict: capitalism has lit more lights than any benevolent figure in history. You may not like its methods, but you cannot deny its results.
The fifth blow — the most lethal: "benevolence" cannot scale. Utilitarianism pursues principles that can be extended to an entire society. But "benevolence" depends on individual character, and individual character cannot be replicated. By comparison, a well-designed system — even one motivated by self-interest — can make a million selfish people produce good outcomes. Adam Smith's "invisible hand" is far more effective at scale than Confucius's "benevolence."
Utilitarianism's verdict: rather than teaching people to be benevolent, design a system that forces selfish people to do good. The former is romance; the latter is engineering. The AI era needs engineers, not saints.
The Knot Utilitarianism Cannot Untie
Five blows land; every one stands. If this were a debate competition, utilitarianism would probably score higher.
But debate competitions judge arguments, not the proposition. The proposition is not "which is more logical" but "how should a person live."
Utilitarianism has a structural defect: it needs someone to define what constitutes "greatest happiness." In the AI era, this power of definition will inevitably fall into the hands of a few. When a few use AI to calculate "greatest happiness," what they are actually calculating is "the happiness-distribution scheme most advantageous to me." Utilitarianism assumes the calculator is neutral. The calculator is never neutral.
Once the power to define "greatest happiness" is monopolised, utilitarianism becomes the very thing it most opposes — an ideology in service of power.
Every historical action taken in the name of "the greatest happiness of the greatest number" eventually had to answer: what about those not in the "greatest number"? Utilitarianism's answer: they are an acceptable cost. "The benevolent have no enemies"'s answer: no one should be a cost.
This is the fundamental divergence between the two. Not efficiency, not logic, but whether you treat people as ends or as means.
Capitalism has indeed lit more lights than any benevolent figure in history. But it has also extinguished countless lights it deemed "not worth lighting." Adam Smith wrote The Theory of Moral Sentiments before writing The Wealth of Nations — he knew the "invisible hand" required a precondition: basic empathy between human beings. Without empathy, market mechanisms produce not prosperity but plunder. No one remembers which state in the Spring and Autumn period had the highest GDP; everyone remembers Confucius — deep in the human cognitive structure, some things matter more than "results."
AI exposes the ultimate limitation of utilitarianism at precisely this juncture. AI can calculate, optimise, and maximise anything defined as a "target." But it cannot decide what the target itself should be. Setting the objective function requires a judgement about "how a person should live." The core of this judgement points in the same direction: people are not means; people are ends.
Utilitarianism won the arguments but lost the proposition. It answers "what is most effective"; "the benevolent have no enemies" answers "why do it at all." The former is methodology; the latter is existential philosophy.
Four Classics as One: Not Just Self-Cultivation, but a System
Up to this point, "the benevolent have no enemies" is merely a personal conviction. But the full context of these four characters requires four classical texts to restore.
The I Ching is the operating system: the cognition of patterns. It does not teach you to be good or evil; it tells you that the operation of all things has its rhythm — yin and yang wax and wane, firmness and gentleness complement each other, prosperity leads to decline, adversity turns to hope. A benevolent person who does not understand patterns is merely a kind blind man whose goodness will be digested by the system without leaving a trace. Utilitarianism says the benevolent do not understand efficiency, yet the I Ching's concepts of timing and position are more precise than any theory of efficiency — it does not tell you to persist blindly but to judge the "timing" and "position," doing the right thing at the right moment.
The Tao Te Ching is the supreme principle: the philosophy of governance. Laozi's teaching is not passive withdrawal; it is counter-intuitive governance wisdom — "the highest good is like water; it does not compete yet excels at overcoming." The best institutional design does not mandate goodness; it creates an environment where goodness will not be punished. But the Tao Te Ching alone is overly idealistic. "Governing through non-action" holds as a supreme principle, but it lacks a mechanism for descending from the ideal to reality — when the environment itself has become so distorted that goodness is punished, how do you "follow the flow"? It is the view from the mountaintop, yet it does not tell you how to climb.
The seven chapters of Mencius are precisely the landing mechanism that completes this link. The theory of innate goodness posits that humans are born with four sprouts of moral feeling; benevolence is not sainthood achieved by defying one's nature, but the expansion of inborn good tendencies. "The people are the most important; the altars of land and grain come next; the ruler is least important" establishes a legitimacy foundation for governance — a ruler who treats his people as means deserves to be overthrown. And Mencius's most crucial demonstration is his response when King Hui of Liang opened with the question of profit: "Why must Your Majesty speak of profit? There are also benevolence and righteousness." This is not pedantry; it is a fundamental rejection of the utilitarian framework. If the king speaks of profit, the ministers speak of profit, the commoners speak of profit, and all compete for profit from top to bottom, the system collapses. When everyone pursues maximising efficiency, maximising KPIs, and society competes top-to-bottom for "numbers," people are devoured by numbers — this logic holds equally in the AI era.
The Analerta provides the daily value anchor. Regardless of how patterns operate or strategies adjust, "the person" is always the end, never the means. The designers of systems themselves require character; otherwise systems become mere instruments of power.
The four together: the I Ching is the terrain map; the Tao Te Ching is the view from the summit; Mencius is the path of ascent; the Analects is the foothold at every step. "The benevolent have no enemies" was never a lone warrior's battle cry; its full meaning is: a person who understands patterns, grasps principles, knows how to implement, and upholds values has the capability to create a world where more people can live as human beings.
"Knowing It Cannot Be Done, Yet Doing It Still"
Even with the complete system, one must still face a cruel reality: knowing how to do something and being able to do it are two different matters.
In the Analects, someone commented on Confucius: "Is he not the one who knows it cannot be done, yet does it still?" The tone carries mockery. He saw very clearly: Confucius's ideal was unrealisable in his time. Confucius was not unaware; he knew and did it anyway. Mencius was the same — knowing full well that King Hui of Liang only wanted to hear about "profit," he spoke of "benevolence and righteousness" nonetheless.
This is precisely the most underestimated layer of meaning in "the benevolent have no enemies." "No enemies" is not a description of outcomes — it does not say you will certainly win. "No enemies" is a description of state — you treat no one as an enemy, so no one can defeat you in the capacity of "enemy." You can fail, you can be destitute, you can be forgotten, but you have not been defeated, because you were never competing on the battlefield of "defeating others."
The evidence history provides is worth noting. Those who "knew it could not be done yet did it still" are often the ones who, after the fact, rewrote the rules of the game. Confucius's failure — fourteen years of sustained failure — defined the spiritual baseline of Chinese intellectuals for more than two millennia. Mencius's "the people are most important" became the intellectual weapon for later generations to question the legitimacy of tyranny. Those who abolished slavery were considered lunatics in their time; human-rights advocates were imprisoned; pioneers of democracy were executed. Every major advance in human civilisation has the same core action: re-recognising "the enemy" as "a person." This action was, in its time, irrational and destined to fail. But accumulated to a critical mass, it rewrote the entire set of game rules.
There is an underestimated case from the business world. Sega was long mocked for "not knowing how to do business" — losing to Sony in the console wars, losing to Nintendo in the portable market. By utilitarian standards, a total failure. But this company's collaboration with NVIDIA drove GPU graphics technology forward; its Yakuza (Ryu ga Gotoku) team persisted for nearly twenty years in a genre — yakuza-themed games — that was entirely non-mainstream, ultimately creating an independent game category. Sega is not a paragon of "the benevolent have no enemies," but a paragon of "knowing many battles will be lost, yet still walking in their own way." Whether luck favours or not is a post-hoc evaluation; the persistence itself is a decision made beforehand.
Good and Evil in One Body: The Jensen Huang Case
Speaking of the historical connection between Sega and NVIDIA, we must turn our gaze to NVIDIA's helmsman.
In June 2026, Jensen Huang appeared on a Korean variety show and, before a global audience, said: intelligence is now almost a commodity; society produces large quantities of smart people every day, and AI amplifies this trend. What truly matters is humanity — human qualities. When seeking collaborators, what he values most is whether a person is generous, kind, and genuinely wishes for others' success (SBS TV Korea, June 2026).
This statement deserves serious attention. The speaker is not a philosopher in an ivory tower but a business leader who commands the world's most powerful computing resources. In Chapter 1, you saw his remark at the Seoul press conference distorted by Chinese-language media; in Chapter 6, you saw him pledge his word about RTX Spark on the Computex stage. Now comes the third angle — the same person.
But stopping here would be a serious mistake. Because the same Jensen Huang has another side.
NVIDIA internally promotes "Speed of Light" execution. Huang requires that employee emails not exceed six lines and directly manages fifty senior executives. Former employees describe interacting with him as "like sticking your fingers in an electrical socket." In his Stanford commencement address, he wished graduates "ample doses of pain and suffering," because he believes NVIDIA's success came from thirty years of repeatedly struggling on the brink of bankruptcy. He has spoken of how, during NVIDIA's early days, a former executive called his business plan "the worst startup pitch ever," then turned around and called Sequoia Capital to recommend the investment. Huang's conclusion: that executive was his benefactor.
There is a dangerous narrative logic here — "the person who scolds you is actually helping you"; "pain is a necessary cost of success." But "there will be light in the darkness" and "light must be accompanied by darkness" are two different propositions. The former is observation; the latter is rationalisation. Huang survived the inferno and then misattributed his personal experience as the foundational logic for everyone's path to success. Far more people are destroyed rather than forged in the same high-pressure environment — he cannot see them, because those people are no longer in the room.
The deeper problem is this: it is not hypocrisy. Huang almost certainly genuinely believes kindness is important, while simultaneously genuinely believing extreme pressure is necessary. In his worldview, the two do not contradict — kindness is the direction; cruelty is the method.
This is "good and evil in one body" — a single person in whom goodness and wickedness genuinely and sincerely coexist, while their own narrative system unifies both into a coherent story. If he were hypocritical, you could unmask him. But he is sincere, so you cannot discern from his words where the problem lies.
This is far more terrifying than "saying one thing, doing another." When suffering is redefined as grace, when psychological oppression is packaged as "for your own good," when unreasonable pressure is called "training," even the language of victimhood is taken from the victims. You will not feel that you are being harmed; you will feel that you are not strong enough.
And this logic is being replicated across the entire corporate world. A genuinely kind company and a company using kindness as packaging can write corporate values that are word-for-word identical. You cannot distinguish truth from falsehood from the manifesto, just as you cannot discern a person's character from their CV. The only way to tell real from fake is not to watch what the other party says, but to watch what they do under pressure.
When the market rewards quarterly performance rather than long-term character, when investors care about share price rather than employee well-being, the entire system automatically selects for those most skilled at "appearing kind." Individual convictions are crushed by institutional logic, and the institution feels no guilt about it.
Slow Poison
More dangerous still, this process does not happen suddenly. It is gradual.
Most people do not wake up one morning and decide to abandon goodness. Under institutional pressure, goodness is eroded bit by bit. Sacrificing a colleague's interests for the KPI — just work requirements. Staying silent about injustice to protect your position — just being pragmatic. Making a decision you do not quite agree with to hit the numbers — just a compromise. Each time is small; each time has a reasonable explanation. Ten years later, looking back, you have gone from a good person to a "sensible" person.
If during this process you accepted the narrative that "pain is grace," you lose even the capacity to realise you are changing. Each time you are hurt, you tell yourself: the cost of growth. Each time you hurt someone else, you tell yourself: for their own good. The rationalisation of suffering passes down, from top to bottom, generation after generation.
This is the true challenge "the benevolent have no enemies" faces in the contemporary world — not someone holding a knife demanding you abandon goodness, but an entire environment acting like slow poison, making you abandon it without noticing. And the most sophisticated poison is the kind that makes you believe the poison is medicine.
Between the ideal and reality, the gap is not a wall but a gradual slope — you think you are still standing in place, but you have already slid far. Being able to recognise the existence of this slope is the most essential capability the benevolent need in the contemporary world.
Then Why Still Do It
Having passed through utilitarianism's five blows, the system of the Four Classics, the lucidity of "knowing it cannot be done, yet doing it still," the case of good and evil in one body, and the mechanism of slow poison — the "benevolent have no enemies" that survives is no longer the original four characters.
It is no longer a declaration about character but a proposition about choice.
Having no enemies does not mean "defeating everyone." Having no enemies means "not needing to defeat anyone to prove yourself."
Using the word-image-meaning framework to locate it: "the benevolent have no enemies" resides at the "meaning" layer, where as a direction no one would object. The real difficulty lies in "image" — in what situations to compromise, in what situations to hold firm? When to borrow utilitarianism's blade, and when to set it down? Then descending to "word" — how do you express it so it does not become another slogan to be scrolled past?
AI will equalise knowledge, skills, and even creative output. But trust, responsibility, conscience, and empathy will not be equalised. You can use AI to write a novel or build a company, but you cannot outsource "facing another person."
Utilitarianism says this is irrational. AI's neural networks say this is not the optimal solution. The entire incentive structure of modern society tells you: goodness is an inefficient strategy. They are all correct. But none of them can answer one question: if everyone chooses the optimal solution, what kind of world is left?
This is not "the benevolent have no enemies." This is harder than having no enemies. It is not that you have no enemies — you see every structural adversary, hear every reasonable argument utilitarianism makes, understand why they exist, and then choose not to treat a single person as an enemy.
This choice is not the ending. It is the beginning.
The world is seen clearly; the choice is made. But choice is not free — the next question is: how heavy is this bill?