On Faking It
If taste can be modeled, predicted, and monetized, what’s left for the people who built their lives around it? A dispatch on the necessary slippage between preference and performance.
Thirst Behavior is a project about wine, taste, and the social performances that form around them, the rituals of enjoyment shaped by status, media, money, and desire. Wine is the entry point; the real subject is how taste gets made, circulated, and contested, especially in New York and the Hamptons, where leisure and tastemaking overlap most visibly.
You can find my most recent feature here.
Friday features are always free, but if you want the weekly intel and ongoing education—plus to support independent wine writing that doesn’t do scores or sponsors—upgrade whenever it feels right.
If you’d like to work together or have consulting inquiries, pitches, or scoops, please email me at bodhilanda@gmail.com.
My friend Meg Maker recently sent me a set of questions about the ‘nature of taste’—what it is, how it develops, whether it can be taught—which, on the surface, felt straightforward enough. I used to think I was going to be a contemporary art philosopher, so I’ve attempted at various times in my life to seriously grapple with these ideas. But in thinking through them in relation to wine, things get more interesting. The word itself is already unstable. In English, “taste” refers both to a literal, bodily sensation and to something more abstract: a capacity for judgment, the assignment of value to aesthetic experience, a set of preferences that are fundamentally personal and also socially legible. The more I tried to pin it down, the less clear it became where the boundary lay between sensation and interpretation, instinct and evaluation.
Also clattering in my head is Kyle Chayka’s piece for the The New Yorker about the sudden fixation on “taste” in tech circles, where it’s being recast—somewhat optimistically—as the last defensible human advantage in an AI-saturated landscape. If machines can now produce text, images, music, code—more or less anything—on demand, then the scarce resource is no longer creation but selection. The person who can sift through the expanding field of generated content and identify what matters begins to look, from a certain angle, like the most valuable actor in the system. Taste, in this formulation, becomes a kind of filter: a way of imposing order on an ever-expanding abundance of slop.
As usual, our philosopher of broken-brain millenial psychology has arrived at a conclusion painfully obvious for those of us already living inside this moment. Recommendation engines, algorithmic playlists, endlessly refreshing feeds—these are all attempts, in one form or another, to formalize preference, to translate something that feels intuitive and subjective into a set of patterns that can be modeled, predicted, and, crucially, monetized. If you liked this, you will like that. If people who resemble you tend to choose X, then X can be surfaced again, and again, and again. Taste, flattened into preference, becomes legible as data. YouTube and Spotify have proven that recommendation algorithms are the key to monetizing taste in the age of information overabundance. And all other industries, wine included, are clamoring to reproduce this kind of success.
But something gets lost in that translation, and it’s not a minor detail. It is, I would say, the entire point.
To say that taste is what you like is true, but insufficient. Taste is also how you make that liking intelligible to someone else. It is not only a private orientation toward the world, but a public act of communication—an attempt to render a subjective experience legible, persuasive, even contagious. When we talk about someone having “good taste,” we are rarely referring to a static list of approved objects. We are responding to a way of moving between objects, a way of noticing, a way of drawing connections and assigning significance. Taste, in other words, is not just selection. It is articulation.
This becomes especially clear in fields where taste is supposed to function as a form of expertise. As a sommelier, I am, in theory, tasked with knowing which wines are “good,” but that knowledge, on its own, is inert. A bottle does not declare its own value in any meaningful sense. It has to be positioned, described, related to something else. It has to be made to make sense within a particular moment, for a particular person, at a particular table. The work is less about identifying the right object than about constructing the conditions under which that object will be experienced as meaningful. The same is true, in a different register, of deejaying, where the track itself matters far less than the sequence, the timing, the way it lands in a room that is already in motion. In both cases, what we call taste is inseparable from performance.
This is precisely the dimension that resists abstraction, and therefore the one most likely to be ignored. What gets captured instead is the residue—the list, the recommendation, the output of a process that is far more contingent and relational than it appears from the outside. Chayka describes a version of this dynamic in his essay, pointing to what he calls “taste-washing”: the attempt to confer a veneer of human sensitivity onto systems that are, at base, incapable of having an experience. The machine does not taste anything; it recognizes patterns in the traces left behind by those who do.
When taste is reframed as a set of preferences that can be modeled, it becomes tempting to treat it as a problem of optimization. Better inputs, better outputs. More data, more accurate predictions. But this assumes that the goal of taste is to reliably produce satisfaction, to deliver more of what one already likes with increasing efficiency. And while that may be a reasonable description of consumer behavior, it is a poor account of taste as such. Taste, if it is to remain interesting, depends on its own instability. It evolves, contradicts itself, encounters things it cannot immediately process. It is shaped as much by misrecognition as by recognition, by moments of friction as much as by moments of ease.
There is also, perhaps uncomfortably, the question of value. If taste is now being positioned as a monetizable skill—the ability to extract signal from noise, to curate, to recommend—then it is worth asking what, exactly, is being sold. In my own case, Thirst Behavior could easily be read as attempts to package and distribute taste: a set of preferences, a sensibility, a guide to what is worth paying attention to. And that’s exactly what this poject is! But that framing is, at best, incomplete, and at worst, a kind of misdirection.
Because the thing itself—the wine, the restaurant, the song—is rarely the point. What is being produced, more often, is a set of communicational and affective strategies that allow those things to take on meaning in the first place. The recommendation is a device; the object is a vehicle. What matters is the way value is constructed, how it circulates, how it becomes persuasive within a given social field. To write about taste is not simply to rank or endorse, but to participate in that process, to make visible the mechanisms by which something comes to feel significant at all.
From this perspective, the current attempt to systematize taste begins to look slightly misplaced. It is not that machines will never be able to approximate preference—they already do, and with increasing precision—but that preference is only one component of a much larger, messier phenomenon. The part that remains difficult to automate is not the selection of objects, but the translation of experience: the leap from sensation to meaning, from private response to shared understanding.
If taste is going to become valuable again, it won’t be because it can be cleanly extracted and monetized, reduced to a set of preferences and optimized outputs. It will be because it continues to resist that kind of reduction—because it operates less as a system of discernment than as a performative strategy. Something enacted in real time, in relation to other people, shaped as much by misfires and adjustments as by certainty. Moments of friction, moments of play—these are not bugs in the system, but the very conditions that make taste function as a social mechanism at all.
This is the part of taste that resists automation—not the identification of the thing, but the negotiation around it. The small, human adjustments that turn a preference into a moment, and a moment into something that feels, however briefly, like meaning.
This raises a slightly uncomfortable possibility: that what’s being monetized here is not the object, exactly, but the ability to make the object matter in the first place. To frame it, to position it, to move it through a network of references and associations until it takes on weight. This is true of wine, obviously, but it’s just as true of music, of art, of anything that circulates as culture.
In that sense, the current fixation on taste is not entirely misguided. It’s just misnamed, because the thing being valued is not taste itself, but the capacity to fake it. If people would rather feed photos of my wine list into ChatGPT than ask me what to drink, what exactly is my job supposed to be?
Maybe the answer is that it changes. Taste-makers have to lean harder into the parts of perception that don’t quite resolve—to hold competing impulses at once, to organize and destabilize, to introduce a few bugs into their own systems of meaning making.
Effective communication in the recommendation economy will look less like expertise and more like performance. Humor, play, and, occasionally, the willingness to create confrontation; just enough to remind people that having the right answer was never the point in the first place.