One of the recurring errors made by people who build things is the assumption that everyone else is engaged in the same basic enterprise. The engineer sees a defective system and asks how it can be repaired; the activist often sees the same defect and asks what moral significance can be extracted from it, who may be blamed for it, and what social position may be gained by being publicly identified with its correction. These are not merely different temperaments. They are different relationships to reality.

For the engineer, a solved problem disappears, and this is usually counted as success. For the activist, however, a solved problem can be strangely inconvenient, because the problem may already have become the organizing principle of an institution, a constituency, a profession, or an identity. A permanent grievance can sustain organizations, reputations, fundraising, media attention, bureaucratic authority, and a stable distinction between the righteous and the accused. Once this structure exists, the problem is no longer merely something to be solved; it is also something upon which an ecosystem depends.

This helps explain why activist politics so often drifts toward leveling. Excellence naturally produces distinctions, because some institutions work better than others, some people are more capable than others, some firms become larger than others, and some societies organize themselves more successfully than others. A political mentality that interprets distinction primarily through the language of power and injustice will therefore tend to see exceptional performance not simply as an achievement, but as a condition requiring moral explanation. The higher something rises, the more urgently it must justify its height.

An advanced civilization cannot function indefinitely under that assumption. Technological leadership requires large concentrations of capital, abundant energy, tolerance for unequal outcomes, unusual people with unusual talents, ugly and expensive infrastructure, and institutions willing to choose competence over emotional symmetry. Semiconductor fabs, power plants, transmission lines, laboratories, data centers, and frontier research organizations do not emerge from a social order devoted chiefly to ensuring that nobody appears too powerful, too rich, too strange, or too far ahead of anyone else.

The great irony is that the technological sector itself spent years behaving as though this tension did not exist. Major technology companies became increasingly comfortable with institutions that treated political and factual disputes as matters of moderation, safety, trust, misinformation, disinformation, harmful speech, and reputational hygiene. The exact motives varied, and not every moderation decision was ideological, but the general institutional temptation was obvious: once a bureaucracy possesses the power to decide which claims respectable people may encounter, it becomes dangerously easy to substitute administrative authority for argument.

The governing question then changes. Instead of asking whether a claim is true, institutions begin asking whether it is responsible, acceptable, safe, harmful, or endorsed by the proper authorities. Those categories are attractive because they permit an institution to exercise power while retaining the self-image of neutrality. Yet they also create a remarkable vulnerability, because political judgments begin masquerading as technical procedure.

Such arrangements can appear stable for a while, especially when the same institutions control distribution, prestige, and professional legitimacy. But information control has an unpleasant relationship with time. Evidence accumulates, archives survive, political coalitions change, former orthodoxies lose their aura, and claims that once seemed too dangerous for polite discussion eventually become ordinary subjects of retrospective debate. At that point, the censor confronts a problem more embarrassing than simple error: the public can see not only that an institution was wrong, but that it tried to prevent others from being right in public.

This is why censorship ages so poorly. The censor is not merely making a judgment about the present; he is placing a wager on the future and assuming that tomorrow’s observer will inherit today’s taboos. Sometimes that wager succeeds. Quite often, it becomes ridiculous.

The technology industry is now discovering another unpleasant property of ideological coalitions: they are not permanent alliances but temporary convergences of interest. A company may spend years accommodating a political movement, financing its institutions, hiring from its networks, repeating its language, and implementing its preferences, only to discover that none of this purchases permanent immunity. The corporation that was yesterday praised as enlightened can tomorrow be redescribed as a monopoly, an environmental threat, a concentration of private power, or an obstacle to justice.

The point is not that every criticism of technology companies is illegitimate. Some are entirely legitimate. Nor is it sensible to assume, without evidence, that every protest against data centers or energy infrastructure is orchestrated by a foreign adversary. The more serious strategic point is that a society trained to respond to emotionally effective formulas rather than to measurable outcomes creates an obvious vulnerability to manipulation.

A foreign rival does not need to invent every grievance from nothing. It merely needs to understand the host society’s existing fractures. A modest amount of money, media amplification, organizational support, or narrative reinforcement can sometimes increase the salience of disputes that already exist. The participants need not be dishonest, bribed, or consciously disloyal. In fact, sincere participants are often more useful than cynical ones, because they provide the movement with authentic conviction while remaining unaware of whatever larger strategic interests may benefit from the resulting disorder.

The sophisticated version of political manipulation therefore requires no theatrical conspiracy. One finds an existing grievance, gives it a vivid vocabulary, identifies a villain, rewards repetition, punishes skepticism, and allows domestic status competition to do most of the work. Open societies are especially susceptible because they permit precisely the freedom of organization and communication that hostile actors can exploit. The answer to that vulnerability is not repression; it is a culture less easily manipulated by moral theater.

Here the technology industry’s predicament becomes almost comic. Institutions that once believed they could safely cooperate with ideological movements in disciplining disfavored speech now find those same movements increasingly willing to discipline technology itself. Data centers become symbols of greed, energy consumption becomes a moral offense, infrastructure becomes an environmental scandal, and technological scale becomes evidence of illegitimate power. The machinery of accusation remains the same; only the target changes.

There is a certain justice in this, but justice is not the same thing as strategy. Technology executives may deserve to learn that political movements cannot be domesticated merely because their current enemies are convenient. They may deserve to discover that a moral bureaucracy built to punish someone else does not cease being a moral bureaucracy when its attention shifts. But a technologically advanced country cannot afford to destroy the institutions on which its technological position depends merely to teach their executives a lesson in political prudence.

Burning the ship because one dislikes the officers remains a poor naval doctrine.

The larger issue is that advanced nations possess no permanent entitlement to remain advanced. There is no historical mechanism guaranteeing that a wealthy country will continue to lead merely because it led before. Technological supremacy is contingent and expensive. It requires continual investment, toleration of failure, enormous energy inputs, dense concentrations of expertise, capital formation, and a political culture capable of distinguishing genuine abuses from the mere existence of scale.

Artificial intelligence makes the problem particularly stark because the relevant inputs are physical as well as intellectual. Models require chips, chips require fabs, fabs require capital and energy, data centers require electricity and transmission, research requires exceptional talent, and all of these things require institutions willing to permit construction before every possible constituency has been morally satisfied. A society that makes each of these inputs increasingly difficult in the name of preventing every conceivable harm should not be surprised when technological capacity migrates elsewhere.

Reality is not especially interested in the nobility of the explanation. A country may celebrate justice while allowing its grid to decay, regulate concentration while exporting its most capable firms, denounce inequality while making itself uniformly poorer, and congratulate itself on having restrained dangerous ambition while discovering that ambitious people have simply left.

This is where the leveling instinct becomes strategically dangerous. A civilization that cannot tolerate internal excellence may eventually discover that it has achieved equality chiefly by lowering the ceiling. The result is not necessarily justice, and certainly not strength. It may simply be mediocrity administered with impressive moral vocabulary.

The proper response is not blind loyalty to technology companies, any more than it is blind hostility to activists. It is to recover a more demanding standard: institutions should be judged by what they produce, movements by what they solve, and policies by the conditions they leave behind. Environmental politics should eventually produce a livable environment and sufficient energy. Housing politics should eventually produce housing. Education policy should eventually produce educated people. Technology policy should eventually permit technology to exist.

A society that loses the ability to ask these questions becomes vulnerable to endless symbolic politics, because symbolic politics is much easier than competence. One can always declare another emergency, invent another moral vocabulary, identify another class of villains, and demand another sacrifice. The problem is that civilizations are not sustained by the elegance of their accusations.

They are sustained by people who can make complicated systems continue to work.

Alignment, then, should not mean submission to Silicon Valley, to activist organizations, to political parties, or to whichever faction presently possesses cultural prestige. It should mean alignment with measurable reality, institutional competence, technological capacity, and a willingness to correct error without turning every correction into a morality play.

History is quite tolerant of hypocrisy and almost infinitely tolerant of slogans. What it is not tolerant of is incapacity.