Everyone is talking about “taste” right now.
I’ve heard the word a hundred times this month alone. From clients, from friends, from the algorithm itself. It’s become the favorite hedge against AI. The thing humans still have and the thing creatives are supposed to have, instinctively.
Which is why I keep getting stuck on the same question: if taste is suddenly the secret-sauce for success, where is mine actually coming from?
Table Setting
I’ve been thinking about my taste a lot lately, but not in a what is good design way. More in a late-night-spiraling-while-doom-scrolling way. The kind where you read the comments before the post, decide how to feel, then catch yourself and wonder if your opinion was even yours to begin with.
Earlier this year I stumbled upon Martin Parr’s work. (If you don’t know him: British photographer, brutal color, often captured the quirky British seaside.) I started on Google Images and I clicked on photos that caught my eye. Then I moved to Instagram, where the same photographs were sitting under different little numbers (likes, comments, shares) and I noticed I was clicking the ones with higher engagement.
Same eyes. Same photographer. Different filters before the click-to-view.
It’s the kind of thing you almost don’t notice, but once you do, you can’t stop catching yourself in the act. Was I forming preferences based on public consensus and calling them mine?
After that, my digital habits sort of split in two.
There’s the fast side: the algorithmic feeds that know me too well, that hand me opinions I already half-agree with, products I’d actually buy, and references made for whatever I’m working on that week. (Think: TikTok, Instagram, Pinterest)
And then there’s the slower side: the quieter, more intentional rabbit hole (think: Are.na, Wikipedia, sometimes Reddit at 1am) where I have to actually click, save, label, connect. There’s relatively more friction, and the friction is the whole point. It makes you pause, assess, then react, and maybe even organize too. It introduces a sense of involvement.
And recently I’ve rediscovered a third side that lives mostly off my screen (big obvious-but-not-obvious admission given that I work digitally for a living). Imagine: sitting through a bad album until you find your new favorite song. A slow Sunday at a bookstore where you rummage through shelves to uncover a new author. Staying with a movie that bores you for forty minutes until, suddenly, it doesn’t. The kind of friction that used to be the only friction; the deepest version of the slow side, with no UI in the way.
When curiosity paired with struggle leads to discovery, that outcome feels more “earned.” And not just that, it feels more mentally absorbed and ingrained.
But in a world where the friction-to-discovery adventure gets zapped by the algorithm from beginning to end, without effort or detours, what happens to the version of us that the struggle used to produce? As creatives, what happens to the things we make?
Meet the guests
TikTok: A recommendation machine wearing a personality.
The genius of TikTok is that it makes discovery feel personal. That my ‘For You’ page is curated differently than yours. But it’s optimized for the fast yes: watch time, retention, the shape of your hesitation before a swipe or reaction. Tiktok itself acknowledges the “filter bubble” risk and says it deliberately injects “varied” content into the feed, which I find darkly funny. The filter bubble is patched by adding more bubbles.
TikTok is the friend who agrees with everything you say, loves everything a little bit, and in a room of ten people is the average of all ten. Charming but forgettable.
Are.na: A curation tool that doesn’t really want to be a platform.
It describes itself as “software for saving and organizing content that is important to you” and “a toolkit for assembling new worlds from scraps of the old.” There’s barely an interface. No likes. Follower count doesn’t matter in the usual sense. You collect things (text, links, images, pdfs), put them into channels, and connect the channels to more channels if you want. The site quietly nudges you to “ignore algorithmic rabbit holes” and treat your classification system as a muscle worth exercising.
Are.na is the friend who knows something about everything. You ask why low-rise jeans came back and she can piece together a theory about a runway show, the broader economic mood, and a 1970s subculture that brought it back. Weirdly, it makes sense and you love her for it.
Jonathan Anderson: An author of culture who happens to design clothes.
At JW Anderson, Loewe, and now Dior, he uses a cross-disciplinary lens to assemble strange new worlds that feel like they should’ve already existed. He pulls from Renaissance paintings, queer cinema, Irish ceramics, internet ephemera, gardening, the specific quiet significance of a green Fermob park chair. He doesn’t chase trends. He metabolizes and synthesizes concepts that captivate him to produce craft that’s undeniably his own.
Jonathan is the friend you wish you were cool enough to have. You’d ask how his day was and he’d give you a quiet, thoughtful answer about a weaving technique he learned from a 70-year-old tailor. Then he’d segue into a realization he had during his daily walk through Tuileries about how those green park chairs shape Paris’ social culture. He’d end with a verbose rant about a problem in his garden. You’d go home and think about it for a week. A few months later, you’d see green miniature park chairs as Dior’s invitation and gardening tools in his namesake collection, recast with an Andersonesque lens of object literacy.
Appetizers
I’ve noticed something strange about how I collect things.
On TikTok, I save reflexively. A frame, a soundbite, a sentence someone said. It feels less like I love this and more like this feels right. If you asked me why, I might genuinely not have an answer.
When I move those same references into an Are.na channel, something shifts. I start organizing: I rename channels to be more specific (yes, I have separate channels for “interior_architecture” and “interiors_object”). I group things and then un-group them. I move a screenshot from one channel to another and realize what I was actually drawn to was a lighting choice, not the room itself. I’m not just collecting anymore: I’m curating.
But as a creative, there’s a strange thing that happens at scale. When you’re exposed to this much great work, collecting and sifting through this many references and opinions and ideas, the line between seen and original blurs.
I’ll see something on TikTok that lines up almost exactly with an idea from my notes app last week. Instead of feeling validated, I feel slightly displaced. Was I late? Did I see this and forget? Are we all sinking our forks into the same invisible cake?
Picture that cake.
The icing is what TikTok does: the surface decoration, the first thing anyone sees, the part most easily copied.
The sponge underneath is what happens on Are.na: the structural layers that hold a sensibility together once you’ve sat with it long enough to organize.
And the filling between the layers, the specific thing that makes the cake yours and not someone else’s, is the synthesizing: the part where you stop sounding like everyone else. The further down you go, the harder it gets, and the more it requires of you.
Which makes me wonder if the real question isn’t about access anymore. We all have access, and quite frankly, we’re drowning in it via slippery endless scrolls and fleeting cognitive load. Maybe the question is what we’re willing to do with what we’ve collected once the collecting is over.
Main Course
My dad called me a few weeks ago to ask, gently, whether my job felt “AI-proof” or not. We were ~20 minutes in before he landed on taste as the thing that would save it. That’s the consensus right now: taste is the differentiator, taste is the moat, taste is what’s left.
Here’s the trouble. Taste has gotten more valuable at the exact moment the conditions that produce it have gotten weaker. The systems we currently have reward speed over study, reaction over reflection, visibility over intimacy. We’re being tapped for something rare using tools designed to flatten it.
And I’m not, by the way, worried about whether humans beat AI on taste. I’m interested in whether I’m still building taste in a way that produces anything specific and distinct at all.
Historically, cultivating taste was always algorithmic in a sense. In Pierre Bourdieu’s La Distinction (1979), he argued that taste isn’t innate; it’s a product of class, geography, education, the milieu you happened to be raised in. The curation engine just used to be invisible. It happened through your parents, your zip code, the bookstores within walking distance, the people who happened to sit next to you in college.
Over time, the curation engine has become visible and access has democratized (incredibly; I’d take today’s exposure over 1974’s gatekept canon any day). But the engine is a recommendation system, optimized for engagement, not for you.
And it goes deeper than the feed itself. Adam Mosseri recently confirmed that even Instagram comment sections are algorithmically ranked and sorted by what the platform thinks you’ll engage with. So even the discourse around the content is pattern-matched to your preferences. It’s pattern recognition all the way down.
Are.na resists this on purpose. You sit with references, revisit them, organize them, and (if you’re me) reorganize the already-organized into the more-organized. There’s a quiet construction happening. And construction is where authorship begins.
(Though I’ll admit, Are.na isn’t immune to performance either. A beautifully titled channel of fifty unprocessed images is still fifty unprocessed images. Curation becomes taste only when you’re willing to throw things out.)
Anderson sits somewhere else entirely. His work assumes a level of taste and authorship instead of trying to prove or perform it. When he puts a ceramic shape on a heel, he doesn’t tell you what to think about it; he trusts that something in the gesture will land. The authorship is confident enough that it doesn’t ask for your engagement metrics.
Consider what Anderson did with shoes at Loewe Spring 2022. The collection’s mood was Pontormo, the Mannerist Renaissance painter known for distorted, hysterical figures, and the show was billed as “neurotic, psychedelic, completely hysterical.” The gesture that broke the room was the ironic sculpted footwear: a series of stilettos with the heels replaced by everyday objects: a cracked egg with oozing yolk, a bar of soap, a bottle of nail polish, a birthday candle, a crushed red rose.
No tweet preceded any of this. No trend forecasted it. He took a luxury form, removed its functional component, and substituted in objects from someone’s bathroom shelf; Dadaist found-object logic applied to a slingback, with a Renaissance painter sitting underneath the whole collection.
The audience’s aesthetic vocabulary had to stretch to fit. That’s authorship in the algorithmic era: not pulling from consensus, building ahead of it. The outcome is greater than the sum of its parts because Anderson is the anchor that holds it all together. And the only reason that’s possible is the slow, considered, and personal off-platform diet that feeds him in the first place.
The industry has rewarded this. In 2025, LVMH named him sole creative director at Dior, making him the first person to lead women’s, men’s, and couture since Christian Dior himself in 1957. His own framing of the appointment, told to Wallpaper: “Daring to enter the house of Dior requires an empathy with its history, a willingness to decode its language, which is part of the collective imagination, and the resoluteness to put all of it in a box.“ Decode, then re-author. Take in the inherited material, sit with it, and produce something specific enough that it makes the next thing possible.
That’s what an Andersonesque practice actually looks like in habit form: daily walks through Tuileries, slow intentional study of craft techniques, a cross-disciplinary information diet, a refusal to chase the discourse, and the resoluteness to put all of it in a box of his own making. You can’t optimize for any of it and that’s the point.
Dessert
Maybe taste right now isn’t about finding the right references or “human storytelling” or other synonymous phrases I’m becoming allergic to. We’ve already been doing all of that.
Maybe it’s about being able to withstand influence. Sit in the noise. Take in the endless stream of output without flinching. And still make something that feels specific. Or even a little weird.
I’m not interested in ditching the internet. That’s neither possible nor productive, and the people who pretend to have done it are probably lying. The point isn’t to escape influence (because no one ever has) but rather to metabolize it. Run it through your own contradictions and instincts and embarrassing obsessions, and return something that feels chosen.
The act of thoughtfully synthesizing and authoring opinions, craft, and stories from our own point-of-view, rather than redoing what already exists, requires time, obstacles, and effort.
That’s the friction I don’t want to lose. Not the friction of slow internet, but the friction of slow thinking. The bad album, the boring movie, the slow beginning before the book becomes the book. Embracing the discomforts of everyday life.
So maybe the question isn’t do I have taste. Maybe the question is am I still willing to be bored, patient, slow, deliberate.
Anderson authored his own taste by walking through Tuileries and learning what one tailor knew. Most of us don’t get to make our living from things we encounter slowly. But the routine is reproducible. The willingness to sit with a boring thing, to be comfortable not knowing, to connect the dots slowly, to make the work the algorithm couldn’t have predicted because we had the patience to turn material into something that feels chosen.
So this week, I went back to Martin Parr. Same photographer, same Google Images, same TikTok search. This time, I simply made myself scroll deeper or click on the stuff with lower numbers under them; the ones the algorithm had decided weren’t going to do well.
Some of them I genuinely loved, some of them I didn’t. I knew which was which, and the answer felt like mine.
Party Favors
Why Tech Bros Are Now Obsessed with Taste
By Kyle Chayka
How do you describe are.na at a party? [Are.na channel with user inputs]
By Meghna Rao
Notes on “Taste” (Are.na channel)
Brie Wolfson
Jonathan Anderson’s Things (2016)
By Alice Gregory
Anderson Apotheosis: JW Remodels His Own Brand
By Tim Blanks
Jonathan Anderson in conversation with Bella Freud (YouTube)
In conversation: Jonathan Anderson & Tim Blanks | System (YouTube)
If you’re new here, hi!
Each issue of The Dinner Party is a cross-disciplinary thought experiment where two to three different people, brands, or cultural concepts are seated at the same metaphorical “table.”
We unpack their connective threads, commonalities in ideologies, and what they tell us about creativity, psychology, and culture today.
Bon appétit🍴🥂
This article was written & designed by a human. No generative AI was used.
Long live long form