Essay 3 | Taste and the End of Scarcity
Ellison Carter Essay Series: “The Next IT: Intuition and Taste in the Age of AI”
Essay 3 | Taste and the End of Scarcity
For most of human history, the question that governed technical work was: can we do this? Can we bridge the river, cure the infection, model the climate, connect the network? The difficulty of forming solutions—the sheer intellectual cost of training people to develop them, the material cost of testing them, the institutional cost of coordinating them—meant that arriving at any working answer was achievement enough. When solutions are scarce, we use what works and linger less over whether the solution goes beyond what is sufficient to what is good.
That scarcity is ending. Not everywhere, and not all at once, but the direction is unmistakable. Computational systems now generate solutions faster than we can evaluate them. An engineer working with an AI model can produce in an afternoon what once required a team and a quarter. A researcher can test multiple hypotheses before lunch. The constraint is no longer the generation of options but the selection among them. And selection, when it concerns not just efficiency but the shape and consequence of what we build, is the domain of taste.
I am less certain about taste than I am about almost anything else I’ve written in this series. Intuition I can ground in cognitive science, in pattern recognition, in decades of research on tacit knowledge and expert decision-making — and more recently in work on embodied cognition, interoceptive prediction, and the idea that reasoning evolved to be shared rather than solitary. Judgment I can locate in professional practice, in the accumulated fine-tuning that define expertise. But taste? Taste is the capacity I keep circling without being able to pin down, which may be exactly the point. I suspect that taste resists definition not because it is vague but because it operates at the boundary between the personal and the communal, between what I sense is right and what we come to recognize as good. And that boundary is where some of the most consequential decisions of the coming era may be made.
The Long Confinement of Taste
The Western intellectual tradition has thought carefully about taste for centuries, but almost exclusively within the domain of aesthetics. When Immanuel Kant wrote the Critique of Judgment in 1790, he defined the judgment of taste as disinterested: a response to beauty that is free from any concern with the object’s utility or moral purpose. For Kant, taste operates through a “free play” between imagination and understanding, a pleasure that claims universality precisely because it is detached from practical stakes. This was a philosophical triumph of a particular kind. It gave taste intellectual legitimacy by separating it from interest, consequence, and power.
David Hume, writing three decades earlier in his essay “Of the Standard of Taste,” approached the problem from another direction. Where Kant sought a transcendental foundation, Hume grounded taste in the character of the perceiver. A good judge, for Hume, possesses delicacy of perception, practice, the capacity for comparison, freedom from prejudice, and good sense. These are, notably, cultivated capacities—not innate gifts. Hume’s standard of taste is empirical, social, and trainable, which makes it more useful for our purposes, even if his examples never left the world of poetry and painting.
Nearly two centuries later, Pierre Bourdieu turned the entire tradition on its head. In Distinction (1979), he argued that taste is never disinterested. It is a social weapon, a mechanism by which dominant classes naturalize their preferences and enforce hierarchy. To have “good taste” is to have internalized the aesthetic codes of the powerful. Bourdieu’s critique was devastating and necessary: it showed that what we call taste often conceals what we mean by status. But its very success may have reinforced the confinement it meant to expose. If taste is always about social distinction, then it seems to belong permanently to the realm of cultural construction and consumption—art, food, fashion, décor—rather than to the realm of technical production.
I think this confinement is partly an artifact of material conditions. For most of history, the technical disciplines had so little surplus that the question of taste rarely arose. When it takes a decade to train an engineer and a year to test a bridge design, you don’t deliberate over the elegance of your load calculations. You verify them. The work is defined by constraint, and constraint leaves little room for the kind of choosing that taste requires. Taste, after all, implies alternatives: you can only exercise taste when you have more than one viable option and must decide which is better—not merely which functions, but which is fitting, proportionate, responsible, or even beautiful. Engineering under scarcity produces competence. Engineering under abundance demands taste.
What Changes When Solutions Are Abundant
The shift I am describing is not hypothetical. It is already underway in fields I know well. In environmental health, computational models can now generate dozens of exposure scenarios in the time it once took to parameterize one. AI-assisted analysis surfaces patterns that would have taken a research team months to identify. In community health practice, the question is less often whether we can model a hazard but which model, built on which assumptions, deployed in which context, serves which community’s needs. The answers to these questions are not technical in any narrow sense. They require judgment about what matters, sensitivity to who is affected, and a sense of proportion. They require taste.
This is, I think, the surprising claim at the center of this essay: taste has been confined to aesthetics not because it belongs there, but because scarcity made it irrelevant elsewhere. When the fundamental constraint was can we?, taste was a luxury. Now that the constraint is shifting to should we? and which version? and for whom?, taste becomes a core technical competence. Not taste in the Kantian sense of disinterested pleasure, and not taste in the Bourdieuian sense of social distinction, but taste as the practiced capacity to choose well among abundant options in consequential settings.
This is engineering taste: interested, embodied, collaborative, and consequential. It is interested because the choices matter—they affect people, communities, ecosystems. It is embodied because, like the intuition I described in a previous essay, it draws on perceptual capacities developed through repeated exposure and reflection. It is collaborative because, as Hannah Arendt saw in her reading of Kant, judgment that claims validity beyond the individual must be tested in public, refined through dialogue, and accountable to a community of practice. And it is consequential because, unlike the aesthetic judgment, engineering taste operates in a world where the stakes can be high and the outcomes are lived.
Taste as Calibration, Not Refinement
If taste in engineering is not about aesthetic refinement, what is it? I want to propose that it is fundamentally about calibration: the ability to frame the right question, for the right context, along dimensions that interact in complex ways.
Consider what happens when an environmental health researcher decides how to communicate a risk finding to a community. The finding itself may be precise, but its communication requires navigating competing imperatives: accuracy and accessibility, urgency and proportion, scientific rigor and emotional truth. A finding that is communicated with perfect technical precision but in language that excludes the affected community is not well calibrated. A finding that is simplified to the point of distortion in order to generate action is not well calibrated either. The researcher with taste finds a response that serves multiple values simultaneously without sacrificing any of them entirely. This is not optimization—there is no single objective function. It is the exercise of a practiced sensibility that holds competing goods in tension and acts from within that tension rather than resolving it prematurely.
This kind of calibration has always mattered, but it has largely remained concealed, held within what we called experience, seniority, or professional judgment. It never needed its own name because it was exercised so rarely and by so few. When generating a single solution consumed most of the available effort, the calibration, if at all, happened incidentally, embedded in the constraints of the problem itself. Now that solutions are abundant and the generation cost approaches zero, the calibration is all that’s left. And we discover that we have no curriculum for it, no shared vocabulary, and very little institutional support for developing it.
The Education Problem
Our educational systems were designed for a world of solution scarcity. We have taught students to find answers — to solve problems that arrive fully specified, with clear evaluation criteria and a single correct response. This pedagogy made sense when arriving at any working solution required years of training and enormous analytical effort. But it has produced generations of practitioners who are superb at execution and poorly equipped for the work that precedes it: forming questions worth answering in the first place.
The distinction between execution and question formation is, I believe, the essential distinction for the age we are entering. Execution asks: given this problem, can you produce a solution? Question formation asks: what is the right problem? When computational systems can generate hundreds of viable solutions in an afternoon, the bottleneck is no longer the answer. It is the question. And choosing among questions — identifying which ones are good, which ones matter, which ones will open productive lines of inquiry rather than close them prematurely — is an act of taste. “Good” here is not a technical term. It is a judgment that encompasses fit, proportion, consequence, context, equity, and care.
Many programs aspire to teach question formation, and some do it well — particularly those built around studio models, case-based learning, or community-engaged practice. But aspiration and institutional structure are different things. The dominant architecture of professional education still rewards execution through timed exams, problem sets with known solutions, capstone projects evaluated against predefined criteria. Even programs that value open-ended inquiry tend to confine it to a single course or a final-year project rather than treating it as a throughline. The result is that most students graduate having practiced question formation only in the margins of a curriculum designed around answer production. They do not routinely sit with an ill-defined situation and generate competing framings of what the problem might be, then articulate why one framing is better than another when “better” cannot be reduced to a single metric. They rarely encounter the conditions under which taste develops, including exposure to variety, practice in articulating qualitative judgments, feedback from communities affected by those judgments, and the slow accumulation of pattern libraries that inform future discernment. In the second essay, I argued that intuition is a trainable capacity that develops through feedback loops. Taste is the application of that capacity to the formation and selection of questions worth pursuing. If we do not learn to teach it at the center of our curricula rather than at the edges, we will find ourselves in a world with extraordinary computational power and impoverished judgment about what to ask of it.
Taste and the Problem of “For Whom?”
If taste is the capacity to choose well among abundant options, then everything depends on what we mean by “well.” Well for whom? By whose standards? Bourdieu’s critique haunts any discussion of taste because he showed, convincingly, that taste tends to reproduce the preferences and power structures of those who already have the most cultural authority. If engineering taste simply means that senior professionals impose their aesthetic preferences on technical decisions, we will have traded one form of scarcity for another—abundant solutions filtered through the narrow sensibility of a privileged few.
I think the answer lies in Arendt rather than in Bourdieu, though both are necessary. Arendt, in her late lectures on Kant’s political philosophy, reinterpreted taste not as a private aesthetic faculty but as a fundamentally political one. For Arendt, taste is the exercise of judgment in the presence of others. It requires what Kant called an “enlarged mentality”—the capacity to think from the standpoint of others, to imagine how a decision looks from perspectives not your own. This is not empathy in the sentimental sense. It is a disciplined cognitive practice in which you train yourself to hold multiple viewpoints simultaneously, not to achieve consensus but to make your judgment accountable to a wider community of perception.
In Arendt’s reading, Kant’s sensus communis—the common sense that grounds judgments of taste—is not a shared standard of beauty but a shared capacity for communication. It is what makes it possible for individuals to make judgments that are neither merely subjective nor mechanically universal, but responsive — shaped and reshaped through encounter with others. This is exactly the kind of taste we need in engineering. A practice of judgment refined through exposure to diverse perspectives and accountable to those affected by its outcomes.
Building Taste as a Shared Practice
In the first essay of this series, I described intuition as a pattern library built through attentive exposure, and taste as the capacity to respond to what we notice. In the second, I explored how intuition develops through feedback loops with computational systems. Here I want to close the arc by suggesting that taste is what makes those feedback loops meaningful. Without taste, we are simply iterating. With taste, we are converging on something good.
But taste cannot develop in isolation. This is one of its essential features and one of the reasons it has been so neglected in technical education. Intuition can, though maybe less productively so, be cultivated privately—through practice, reflection, journaling, attentive observation. Taste requires a public. It develops through friction, disagreement, and the slow work of articulating qualitative judgments to people who see things differently.
I think we can build such cultures more broadly, and I want to acknowledge that some teams, programs, and organizations have been doing this work for years — often without calling it taste, and often without institutional recognition for doing so. What I am describing is not the invention of something new but the naming and scaling of something that already exists in pockets. We are not merely adding a soft skill to the technical curriculum. We are reclaiming a capacity that the Western intellectual tradition confined to aesthetics, that sociology then exposed as a mechanism of class reproduction, and that technical education never developed at scale because scarcity made it seem unnecessary. Taste, understood as the practiced capacity to choose well among consequential alternatives in collaboration with others, may be the most important competence we are not yet teaching systematically.
Closing
I titled this series “The Next IT” because I think intuition and taste are the next frontier of professional development—not as replacements for analytical skill but as its necessary companions in a world where computation handles more of the analytical burden. The progression across these three essays has been deliberate: from intuition as pattern recognition, to intuition refined through feedback with AI systems, to taste as the capacity to choose well among the abundant options those systems generate.
If that arc holds, then taste is not an afterthought. It is the faculty that influences whether our extraordinary new tools build things that are merely functional or things that are genuinely good — good for communities, good for the people affected, good in ways that survive communal scrutiny. We have never before needed to develop this capacity at scale, because we have never before had this much power to build. That is a question of taste, and we are only beginning to learn how to ask it.
As with the previous essays, this one is an invitation rather than a conclusion. I am genuinely uncertain about many of the claims I’ve made here, and I suspect that uncertainty is productive. If you have found language for what I’m calling taste — in your own field, in your own practice — I would like to learn from it. And if, like me, you have been developing and exercising something like taste for years but largely uncalibrated, untested in public, and only now beginning to find words for it, I would like to hear what that has looked like. We are, I think, at the beginning of a conversation, and writing these essays has been my way of learning how to enter it.
