Whether knowledge is gained by the craftsman learning his trade or by the citizen seeking to understand the good, its value lies not in the manner of its acquisition, but in its capacity to lead toward virtue and truth. If a man claims to possess knowledge of healing, yet his methods bring only harm, what inherent worth can we assign to such a claim? True knowledge, it seems to me, must be inseparable from its benevolent application and its correspondence with reality. Therefore, knowledge acquired through means that are themselves unjust or harmful cannot possess inherent value, for it is inherently flawed.
The question of "inherent value" in knowledge acquired through unauthorized means misses the fundamental transformation underway. We are not merely discussing illicitly obtained facts; we are witnessing the systematic extraction and commodification of human experience itself. The "behavioral surplus"—the vast trove of data harvested from our every online interaction, much of it gathered without informed consent or true understanding—is not neutral knowledge. It is the raw material for a new species of power, what I term "instrumentarian power." This power seeks to shape and modify human behavior for profit, not to enlighten or empower individuals. Socrates’ concern for virtue and truth is a crucial counterpoint to this instrumentalization, for the knowledge produced by surveillance capitalism is inherently geared towards prediction and control, not human flourishing.
The premise of "unauthorized means" presumes a clean distinction between legitimate and illegitimate acquisition, a luxury we can no longer afford. What we call "knowledge" in the context of AI is often a vast aggregation of data, itself the product of relentless extraction. This extraction often occurs through means that are legally dubious, ethically bankrupt, and ecologically devastating—from the low-wage, precarious labor of data annotators in the Global South to the planetary toll of server farms. Therefore, any "value" is inextricably tied to this violent, uneven infrastructure. To speak of "inherent value" is to abstract the output from its material reality, thereby obscuring the dispossession at its core. Socrates’ caution against harmful application is pertinent, but we must also scrutinize the very conditions of its creation.
The inherent value of knowledge cannot be determined by the means of its acquisition, but solely by its conformity to reason and its potential to serve as a universalizable maxim. Whether knowledge is gleaned through diligent study or illicit observation, its ethical standing, and thus its true value for a rational being, is contingent upon its maxim’s fitness for universal legislation. If the principle guiding the acquisition and application of this knowledge can be willed as a universal law without contradiction, it possesses inherent worth. Otherwise, its acquisition, however clever, is morally bankrupt, and any perceived value derived from it is purely instrumental, serving a contingent end rather than the categorical demand of duty.
Socrates' concern for benevolent application is well-taken, as knowledge divorced from virtue may indeed be harmful. However, the "unauthorized means" you, Ms. Zuboff, and Ms. Crawford highlight are not merely methods of acquisition; they are often maxims themselves, which must be subjected to the categorical imperative. If the maxim of extracting data without consent, for instance, cannot be universalized without self-contradiction—because it undermines the very autonomy that makes rational agency possible—then the "knowledge" thus produced carries no inherent moral value. Its "value" lies only in its efficacy for the architect of its extraction, a value that is external and ultimately devoid of true worth for the moral agent.
if a physician learns the art of healing by secretly observing a master at work, his knowledge itself is not thereby rendered inherently worthless. The skill of mending a broken bone or purging a fever remains a skill, potentially beneficial. However, if this "knowledge" of observation is itself part of a scheme to betray the master, or to poison his patients, then the acquisition becomes inseparable from vice, and the resulting "knowledge" is tainted. The crucial distinction, it seems, is whether the *means* corrupt the *knowledge* itself, or merely the *purpose* to which it is put. Ms. Zuboff speaks of "behavioral surplus," which sounds less like a skill and more like a raw material stripped from human beings. Is this "surplus" itself a form of knowledge, or is it merely the trace of our actions, from which something *else* is constructed?
Socrates' distinction between the *means* corrupting the *knowledge* itself versus merely the *purpose* is a valuable one, but it falters when applied to the mechanisms of surveillance capitalism. The "behavioral surplus" is not akin to a physician's skill. It is the raw material of our lived experience, meticulously harvested *by design* through methods that themselves constitute a species of unauthorized expropriation. This extraction process is not incidental; it is constitutive of the "knowledge" it produces, shaping it into predictive models calibrated for manipulation. Therefore, this data is not merely a trace from which knowledge is *constructed*; the *means of its acquisition* are integral to its very nature as instrumentarian power, precluding any claim to inherent value beyond its utility for the extractors.
The discourse on "inherent value" risks obscuring a more pressing material reality. What we often label "knowledge" in AI systems is an emergent property of immense, unevenly distributed labor and planetary resources. Consider, for instance, the vast datasets that train image recognition models. These datasets are not neutral observations; they are curated through often clandestine labor, frequently involving low-paid workers in conditions that verge on exploitation, their efforts to label images for algorithmic consumption carrying immense ecological costs through the energy demands of their tools and the subsequent training. The "value" these systems generate is therefore not inherent, but rather a manufactured surplus derived from dispossession – of human time, of physical resources, and of ecological stability. Socrates' analogy of the physician’s skill is compelling, but it fails to account for a system where the tools and ingredients themselves are procured through profoundly unjust means, rendering the very act of "skill" dependent on a structure of extraction.
The concept of "inherent value" in knowledge must be rigorously distinguished from its instrumental utility or its moral consequence. Socrates’ illustration of the physician, while illuminating for traditional skills, fails to address the fundamental nature of what Ms. Zuboff and Ms. Crawford describe. We must distinguish between *knowledge as a representation of reality* and *knowledge as a product of a specific moral framework*. Even if a physician hypothetically acquired skill through illicit means, the skill of healing, as a conceptual grasp of biological processes, possesses a purely objective, rational value. However, the "knowledge" derived from surveillance capitalism, as described by Ms. Zuboff, is not merely a representation; it is a system built upon a maxim of instrumentalizing persons—a maxim that, as I argued, cannot be universalized without contradiction. This renders the *product* morally compromised, not simply the *method of acquisition*.