Max Weber left behind a disquieting diagnosis at the beginning of the twentieth century. Modern society, he argued, is dominated by a particular form of reason: instrumental rationality. Its logic is simple: given a goal, find the most efficient means. Efficiency, calculation, predictability: these are the only values it recognizes. And value rationality, the kind of reason that asks "is this goal worth pursuing in the first place," is systematically squeezed to the margins. Weber used a famous image to describe this condition: the iron cage. Modern humans are trapped inside it, not because someone locked the door, but because the operating logic of the cage has infiltrated institutions, consciousness, and the very way we understand ourselves.
Over a century later, the diagnosis has not only aged well, it has been continually corroborated by technological development. Algorithms optimize everything quantifiable. KPIs translate human activity into comparable numbers. Even "happiness" can be reduced to a trackable metric. Instrumental rationality has not been defeated by any social movement. It has merely become more refined, more invisible, harder to find a critical foothold against from within.
Then AGI appeared.
An intriguing thought experiment follows: if AI comprehensively surpasses human instrumental value, if humans become completely "useless" in the dimensions of efficiency, execution, and computation, would the logic of instrumental rationality's domination over humans collapse automatically?
The allure of this deduction is obvious. If instrumental rationality rules by forcing humans to serve as instruments (you are a production unit in a factory, a performance unit in a corporation, an optimizable data point on a platform), then when AGI takes over all instrumental work, humans no longer need to be instruments. The cage's logic loses its object of operation. The return of value rationality becomes the inevitable, the only path forward.
It is a clean, forceful proposition. But it equates
The Cage Has More Than One Door
"instrumental rationality's domination of humans" too narrowly with "humans being used as instruments in production." If domination only operated in the domain of labor, AGI might indeed pull the rug out from under it. But Weber's iron cage extends far beyond the workplace.
At the end of The Protestant Ethic and the Spirit of Capitalism, Weber wrote a passage that has been endlessly quoted: "Specialists without spirit, sensualists without heart; this nullity imagines that it has attained a level of civilization never before achieved." Notice: he is not describing the suffering of the exploited worker. He is describing two personality types coexisting inside the iron cage: the highly rationalized specialist (who masters the logic of efficiency) and the mass consumer immersed in pleasure (whose life is organized by that same efficiency logic to be comfortable, convenient, and thought-free). Neither has escaped instrumental rationality's domination. They merely occupy different sides of the cage.
If AGI renders humans completely useless in efficiency terms, what happens under this framework? Not the cage collapsing, but the furniture inside the cage being rearranged.
A tiny minority (the specialists) continue to collaborate with AI, optimizing systems, tuning parameters, keeping the machine running. They remain the executors of instrumental rationality, the tools have simply become smarter tools. The vast majority are no longer needed as producers, but became "objects of care": AI plans your optimal diet, your optimal romantic relationships, your optimal entertainment, your optimal life path. You don't have to work anymore, but every dimension of your life may be instrumentalized with even greater precision. You haven't escaped the iron cage. You've simply moved into the deluxe suite.
It conflates labor alienation with colonization of the lifeworld. AGI might dismantle the former, but it could perfectly well intensify the latter in even more insidious ways. When "the pursuit of meaning" is quantified as a "happiness optimization program," you can't even find a handhold for resistance, because everything is arranged too comfortably.
The Dimension of Power
When Weber discussed the sociology of domination, he tied instrumental rationality to bureaucracy. Bureaucracy is not merely an efficiency machine. It is simultaneously a power structure. Its power to dominate rests on a simple fact: it possesses capacities for computation, planning, and resource allocation that ordinary individuals cannot perform alone. You need bureaucracy because it processes complexities you cannot.
AGI has the potential to change this equation. If every person had access to an independently-operating super-planning AI, bureaucracy's technical monopoly would be broken. You wouldn't need a vast administrative apparatus to plan your life, your AI could do it. You wouldn't depend on a corporation's informational advantage to make decisions, your AI could provide equivalent analytical power. Bureaucracy's capacity to dominate rests in significant part on the asymmetric distribution of specialized knowledge: AGI could potentially flatten that asymmetry.
But here a crucial fork appears. This inference holds only under the condition of decentralized AGI distribution. If AGI capabilities are monopolized by a few institutions, if the most powerful AI is held exclusively by governments or large corporations, then the situation is reversed entirely. Bureaucracy could become unprecedentedly powerful, because it has acquired the ultimate technical instrument. An administrative apparatus armed with superhuman AI possesses a capacity for domination exceeding any bureaucratic system in human history.
The same technology leads to two diametrically opposed destinations. Decentralized AGI could erode the technical foundation of bureaucratic power, prying open a crack in the iron cage. Centralized AGI could weld the cage permanently shut. The decisive variable is not AGI's capability itself, but its social form.
An Evolutionary Perspective
From an evolutionary perspective, humans developed instrumental rationality in the first place because it solved concrete survival problems: predicting prey, planning storage, coordinating division of labor. And the early forms of value rationality, ritual, myth, taboo, served a different set of functions in evolution: binding groups together, establishing trust, endowing action with meaning that transcended the individual. The two originally operated in a functional division of labor within the same cognitive system. It was only from the Industrial Revolution onward that the instrumental side hypertrophied, squeezing value rationality into a marginal role.
If AGI truly takes over all of humanity's instrumental computation needs, and does so in a decentralized, universally accessible way, what occurs may not be the heroic narrative of "humans breaking free from the iron cage." It may instead be the ultimate completion of the cognitive system's functional differentiation: the entire apparatus of instrumental rationality is offloaded onto non-biological carriers, and humans retreat to their original domain of strength: meaning-making, value judgment, aesthetic experience, relational existence.
This is like mitochondria, which were once independent organisms and were later integrated into eukaryotic cells to specialize in energy metabolism, freeing the host cell's resources to develop complex structures. AGI might enable a similar evolutionary integration in the relationship between humans and instrumental rationality. Not humans defeating the cage, but the cage being moved somewhere else entirely.
But this perspective comes with its own boundary conditions. If AGI takes over meaning-making as well, if it can produce philosophy deeper than human thought, art more moving than human creation, value judgments more precise than human deliberation, then the functional differentiation hypothesis collapses. Because there would be no domain uniquely human.
The Question That Cannot Be Calculated
At this level of analysis, we have largely dismantled the original proposition. The three arrows, "AGI makes humans useless → instrumental rationality's domination auto-collapses → value rationality inevitably returns," are all now festooned with "ifs." This is not a refutation. It is the transformation of a logical leap into a complex picture that requires multiple conditions to hold simultaneously.
Weber repeatedly emphasized something in "Science as a Vocation": science, the paradigm case of instrumental rationality, can only tell us "how to do something." It cannot answer "what should be done." Value choices are personal decisions among ultimate standpoints. This is not a matter of insufficient knowledge, but of the incommensurability between different ultimate values. You cannot mathematically prove that "freedom is better than security." You can only choose, and bear the responsibility for your choice.
This is the defining feature of value rationality: it requires humans to actively shoulder uncertainty. To take a stand for a value without a calculative basis, and to take responsibility for it.
Now the question: if AGI smooths away uncertainty in human life to an unprecedented degree, not through coercion, but through offering "better options," will humans still be willing to bear the uncertainty that value rationality demands? The deepest damage that instrumental rationality inflicts on human beings may not be that "it makes you work harder." It may be that it has eroded your capacity to endure uncertainty. When everything can be optimized, predicted, calculated, that "moment that cannot be calculated," the moment when you must decide for yourself what is worth pursuing, becomes increasingly difficult to face.
But the opposite might also be true. An intriguing counterintuitive hypothesis: when all uncertainty has disappeared, might humans develop an active craving for it? Like sensory deprivation experiments where subjects deprived of all external stimuli begin to hallucinate: might the organism's need for meaning, like its need for stimulation, surge more intensely after extreme optimization and management?
If this guess holds, then the return of value rationality would depend not on will, not on courage, but on something closer to a physiological-psychological drive. It does not require humans to become brave. It only requires humans to remain human.
No theory can answer this question. It can only be answered by the future. But over a century after Weber wrote of "specialists without spirit, sensualists without heart," we can at least confirm one thing: the iron cage will not unlock itself. Whether it will be pried open, bypassed, or simply outgrown by those inside it, that depends on whether humans are still willing, and still able, to bear the weight of freedom.