Butterfly effect - what if ChatGPT’s shopping features accidentally births us the next “Vogue”
What if a new class of uber-tastemakers emerges, phoenix-like, from the embers of influencing, torched by ChatGPT's uncanny ability to send you the exact black flat you want?
I was struck—almost moved—by Mo Anwar’s thoughts on fashion, originality, and the slow, deliberate process of actually forming a point of view. Anwar writes about the value of ingesting information deeply enough to digest it, metabolize it, and then—crucially—reimagine it. That’s what tastemakers do. They read, watch, absorb, and then remix. It’s not just knowing what’s “in.” It’s knowing what to do with that information. It’s friction, sure—but it’s mastery of friction. The work is the value. (yes I continue to be obsessed with friction)
Remember when this was the point of Vogue et al? To watch the runway shows, the kids on the street in Tokyo, and the artists coming out of Soho or East London and take all those disparate points and tell us Cerulean was the color of spring? Ah… halcyon days.
Flash forward- at some level, exactly what ChatGPT et al does? It ingests the internet. It digests. It processes. It synthesizes. But the end goal isn’t invention—it’s convergence. Generative AI is, by design, optimized to produce the next most likely word. The most likely black flat. The most likely influencer-approved dupe. It’s excellent at giving you what you expected. What you asked for. What you saw on TikTok and forgot to screenshot.
A good thing!
And for many people, that’s… fine. Great, even. Shopping to be a sheep isn’t a failure of imagination: it’s the point. You’re not trying to stand out, you’re trying not to get it wrong; you’re trying to get the basic right. And honestly, that’s a valuable use case. But it’s not tastemaking.
Tastemaking is the opposite of convergence
Tastemakers, by definition, do the opposite of convergence. They see the thing no one else sees—yet. They pull from the margins and remix the unexpected. Their job is to be left of center, not smack in the algorithmic middle. So if OpenAI is built to flatten toward the most likely, what happens to people whose value is in being unlikely?
Bye bye basic influencers
Here’s one possibility: AI eats the middle. Not the genius. Not the everyday consumer. But the content-producing, brand-collaborating, pseudo-stylist middle. The influencers who built careers curating from the curation. Repurposing trend reports as personal epiphanies. Selling things they were sent for free because they sort of match the feed. AI can do that—faster, cheaper, with better metadata. Why follow someone mid-tier when an AI assistant can recommend the same Amazon dupe based on your browsing history and your best friend’s wedding guest outfit?
Which brings us to a fork in the road—not an unbundling of media, but a bifurcation.
On one side, you have the true tastemakers; the editors, archivists, freaks, and futurists. Instagram/creators’ rise has killed the idea of centralized, hyper-exalted tastemakers. An unbundling as the media folks say. The people who dive deep, see diagonally, and already speak the rare dialects of aesthetic code. With AI tools, they can scale themselves: smaller teams, lower budgets, faster output, smarter reach. They can do more with less and reach further with fewer gatekeepers. Their taste becomes software-enhanced, not software-replaced.
On the other side? The AI that replaces the “just okay” creators. The ones who relied on templates—styling the same 8 outfits in 8 slightly different ways. The ones who followed the algorithm and will now be replaced by it. The mass of creators shilling Temu (or whatever its inevitable post-tariff successor will be) will be outpaced by a bot that can do it all: source the product, write the caption, match it to your data shadow, and deliver you the dopamine hit of being “on trend” without having to scroll a second longer than necessary.
You’ll get your black flats, sure. The ones that perfectly split the difference between your Pinterest board, your favorite influencer, and your budget. All for the low, low price of: (a) the planet’s remaining carbon budget, (b) $145.99, and (c) maybe your soul.
And we might also get an Anna Wintour (probably several) for our generation, types of geniuses at seeing the trends and seductively telling us about them (while selling ads, obvi, I’m not a communist). I look at Laura Reilly 's Magasin which also touched on this friction obliquely last week re: is online shopping TOO easy, and think, imagine where cool shopping content like this can go with AI-driven support.
Convenience isn’t neutral.
Because convenience isn’t neutral. Friction wasn’t the problem. It was the mechanism by which you developed an opinion, refined a sense of self, and encountered something unexpected.
The tastemaker of the future isn’t someone who bypasses friction. It’s someone who masters it—and uses AI to amplify their point of view, not dilute it. They’ll still be rare. But they’ll be louder. The rest? Flattened. Optimized. Monetized. Gone.
Taste was never meant to be easy. And style, like meaning, doesn’t come from convergence. It comes from choice. From risk. From resistance.
And yes, from a little bit of friction.
(also this piece by Kyla Scanlon was also pivotal in my thinking for this piece, highly recommend)
