Care (AI Evaluative Criterion)
Care is the second criterion in the evaluative framework this volume proposes, and it is the criterion that connects formal analysis to ethical practice. Care is the sustained attention brought to the encounter with AI output—the discipline of reading each generated paragraph, examining each line of code, questioning each plausible claim, refusing to accept smoothness as evidence of quality. It is not a feeling but a practice: the practice of treating the AI collaboration not as a productivity tool delivering outputs for consumption but as an occasion for the exercise of judgment. Care is detectable not through self-report (every builder claims to care) but through the output's formal properties: the evidence that outputs were rejected, that iterations occurred, that the final configuration was shaped by discriminating evaluation rather than received by default. Segal's confession of his failures of care—the Deleuze passage accepted because the surrounding prose was smooth, the moments when Claude's eloquence outran his thinking—is itself an exercise of care: the willingness to expose the seam where the smooth concealed a fracture rather than letting the surface stand unexamined. Care operates as the counter-force to the seductive efficiency of the smooth, the refusal to be absorbed by the reflection, the insistence on depth rather than its simulation.
In the AI Story
The concept of care as an evaluative criterion is implicit in Krauss's method but not explicitly theorized by her. Krauss's attention to the formal properties of artworks—her insistence on looking closely, on not being satisfied with the first impression, on reading the surface as evidence of structure—is itself a practice of care. But Krauss did not articulate care as a criterion because her domain (the visual arts) did not face the specific challenge AI production introduces: the elimination of the temporal and material friction that traditionally forced care into creative practice. A painter cannot avoid caring about the surface—the medium requires it. The brush resists, the pigment behaves unpredictably, the canvas imposes physical constraints. AI output arrives smooth, immediate, and accommodating, and the absence of material resistance means care must be willed rather than imposed by the medium.
The Deleuze failure Segal recounts is the paradigmatic case of care's absence and recovery. Claude produced a passage connecting flow and Deleuzian smooth space—elegant, seemingly insightful, and wrong. Segal initially accepted it because the prose was too smooth to trigger suspicion. The error was caught only when something nagged, prompting verification the next morning. The nag was care—the background cognitive process monitoring for the discrepancy between surface plausibility and structural accuracy. That monitoring is not automatic; it is a cultivated capacity built through practice and degraded by the habit of accepting smooth output without examination.
Care is threatened by the economic pressures of the AI moment. When the board-room arithmetic converts productivity gains into headcount reductions, the time that care requires—the pause between prompt and acceptance, the iteration cycle refining output, the verification checking plausible claims against sources—becomes economically costly. The builder who cares produces better output, but the builder who accepts smooth output produces more output, and in a culture that rewards velocity, the incentive structure tilts toward carelessness. Organizational dams—structured pauses, protected verification time, cultural norms valuing quality over quantity—are the institutional mechanisms required to preserve space for care against the current of optimization.
Krauss's practice of close reading—the refusal to accept the artwork's self-presentation, the insistence on seeing through the surface to the structure, the discipline of returning to the work repeatedly until its formal logic becomes legible—is the template for care in the AI age. The practice is teachable, though it cannot be codified into a procedure. It requires the cultivation of attention, the development of perceptual sensitivity to the markers that distinguish the careful from the careless, and the construction of institutional environments that reward care rather than penalizing the time it consumes.
Origin
The concept is developed in this volume's second and sixth chapters as the necessary complement to specificity. Specificity identifies what makes a configuration irreplaceable; care identifies the practice producing specificity. The term is chosen for its resonance with the ethics literature (care ethics, the primacy of caring) and its grounding in the phenomenology of attention (Murdoch's attention as moral practice, Weil's attention as generosity).
Key Ideas
Care is practice, not feeling. The sustained discipline of evaluation, iteration, and rejection—distinct from the emotion of caring about outcomes.
Detected in output, not self-report. Evidence that iterations occurred, that smooth outputs were refused, that judgment shaped the final configuration.
Counter-force to the smooth. The aesthetics of the smooth eliminates friction—care is the deliberate reintroduction of evaluative friction into frictionless workflows.
Economically costly. Care requires time that velocity-optimized organizations cannot afford—the dams required are structural, not individual.
Cultivated through practice. Not an innate capacity but a discipline developed through sustained engagement with resistant material and eroded by the habit of accepting smooth output.
Appears in the Orange Pill Cycle
Further reading
- Murdoch, Iris. The Sovereignty of Good. Routledge, 1970.
- Tronto, Joan. Moral Boundaries: A Political Argument for an Ethic of Care. Routledge, 1993.
- Noddings, Nel. Caring: A Feminine Approach to Ethics and Moral Education. University of California Press, 1984.