Somewhere between the Flatiron District and Williamsburg, a specific kind of sentence started appearing on Hinge profiles with suspicious regularity. "Equally comfortable closing a deal and closing out a rooftop." "Analyst by day, still figuring out my thesis on everything else." "Looking for someone who takes their career seriously but doesn't only talk about their career."
None of these sentences are false, exactly. They are also, increasingly, not written by the analyst at 55 Water Street or the associate at the Midtown firm whose profile they appear on.
New York has always run its dating market a little differently. Hinge's own history makes the point better than any outsider could: the app nearly lost to Tinder in its early years, and the fix — the one that actually worked — was to lean harder into exactly the instincts New Yorkers already had. List where you live. List where you work. List where you went to school. In a city where more than three million adults are single, more than anywhere else in the country, and where the median dater has a résumé before they have a bio, the profile was never just a photo and a joke. It was a soft-launch LinkedIn with better lighting.
Which is precisely why AI has landed here differently than anywhere else. Nationally, roughly three in four singles now say they've used ChatGPT somewhere in their dating life, with usage up more than 300% in a single year. In a market already built around credentialing — school, employer, neighborhood, the unspoken Manhattan-factor math everyone is doing whether they admit it or not — AI hasn't introduced a new behavior so much as it's applied industrial polish to a habit that already existed. New Yorkers were optimizing their profiles like resumes long before a model could do it for them in four seconds flat between subway stops.
The optimised-beige problem, with a Manhattan accent
Call it what it is: everyone from FiDi to Fort Greene is starting to sound like everyone else, only with better grammar and a cleaner em dash.
Ask any sufficiently capable model to write "an engaging, confident dating bio for someone who works in finance/law/media and lives downtown," and it will reach, with impressive consistency, for the same handful of moves — self-deprecating about the industry, sincere about the ambition, a reference to a run along the West Side Highway or a specific bagel opinion deployed as personality shorthand. It is not wrong. It is also, at the scale a single popular app now operates in one five-borough radius, close to indistinguishable from several thousand other profiles reaching for the same moves during the same commute.
This lands harder here than in most cities, for a simple structural reason: New York's dating pool is enormous, but its list of legible status signals is small. Same ten schools. Same twenty employers people actually recognize. Same six neighborhoods worth naming. When the pool of acceptable inputs is that narrow and the tool writing the output is the same tool everyone else is using, the profiles don't just start resembling each other — they start resembling each other along the exact same three axes New Yorkers were already competing on. AI hasn't diversified the performance. It's sanded down the one part of it — the actual writing — that used to require a person to sit down and think for eleven minutes.
And New Yorkers, who are not naive about performance, have started to price this in. Roughly six in ten dating app users nationally now believe they've encountered AI-written messages; a majority say they'd lose interest in a match on learning the profile was AI-generated, even as most of them privately use the same tools. In a city this fast, this competitive, and this fluent in spotting a performance from across a crowded bar, that skepticism arrives early and it arrives sharp. A West Village dater swiping during a 42-minute subway commute — the city's average, and reportedly one of the longest in the country — has neither the time nor the patience for a profile that reads like it was optimized in a prompt window rather than lived in.
What the arms race is actually optimising for
It's worth being precise about the mechanism, because New York's version of the problem sharpens it rather than softens it.
A language model producing a bio for "single, 31, works in finance, lives in Tribeca" is not describing a person. It is predicting the most statistically probable next word given every appealing Tribeca-finance-bio it has ever been trained on — which, at this point, includes thousands of other AI-assisted Tribeca-finance-bios written in the last eighteen months. The output converges toward the center of a distribution that was already unusually narrow to begin with, because New York's professional class self-sorts into a small number of recognizable lanes. The model isn't just averaging humanity. It's averaging a New York archetype that was half-formed before it ever opened the prompt window.
This is precisely backwards from what makes a profile land in a market this saturated. Distinctiveness isn't noise to smooth out — it's the entire signal, and it matters more here than almost anywhere else, because the baseline of competent self-presentation in New York is already unusually high. The detail that's slightly too specific to have been generated — the actual name of the deli, the genuinely strange hobby, the very particular thing about growing up in Bay Ridge or moving here from London for the job — is the detail cutting through a feed of several hundred profiles that all mention the same run club.
What this reveals about the format, not the tool
The instinct is to treat this as a story about artificial intelligence. In New York specifically, it's really a story about what happens when a format that already rewarded credentialed performance gets a tool that can produce that performance at zero marginal effort.
The city's dating culture was built, long before any of this, around a lossy compression: reduce a whole person to their job, their school, their neighborhood, and a few curated photos, then let a stranger judge the compression in the time it takes to finish a swipe on the L train. AI hasn't introduced that failure mode. It's just removed the last bit of friction that used to force some authorship into the process — the eleven minutes of sitting there trying to think of something true to say about yourself instead of something impressive.
What doesn't optimise away
There's a reason this entire problem lives in the profile and dies the moment two people are actually across a table from each other in the same room.
No model has yet learned to draft the specific thing someone says when a real question catches them off guard at a bar in Nolita. Nobody has automated the pause before an honest answer, the actual laugh instead of the typed "haha," the way someone's face changes when the conversation moves off the résumé and onto something they didn't rehearse. That's not an artefact. It's behavior, produced live, and it is the exact information a six-photo, three-prompt Hinge profile was always trying and failing to compress.
We've hosted structured social evenings across Manhattan and Brooklyn as part of more than 19,000 evenings run in 50+ cities since 2014, and if there's one thing New York in particular has confirmed, it's that no version of the profile — AI-assisted or hand-written, credential-forward or carefully casual — has ever substituted for watching how someone actually responds to a real question, in real time, without four seconds to draft the answer. The résumé-profile was always a rough proxy for the person. It becomes close to meaningless once anyone with an app can perfect it without lifting a finger.
The room doesn't have an optimised-beige problem. It can't. There's no prompt for the specific person sitting across the table from you at 7pm on a Wednesday, three neighborhoods from where either of you actually lives, saying something you didn't expect.
Relish hosts structured social evenings for driven professionals in Manhattan and Brooklyn, and across 50+ cities in the US, UK, Canada and Australia, since 2014. Find an evening in New York →