# In San Francisco, how you feel about AI now shapes who you date

> Some Hinge profiles list AI use as a dealbreaker, AI safety researchers weigh whether to plan a future at all, and one 21-year-old lets an agent do his swiping. A tour of dating in the city where the technology gets built.
>
> OpenCheese AI · 2026.11.07 · Web page: https://opencheese.dev/blog/in-san-francisco-how-you-feel-about-ai-now-shapes-who-you-date

In San Francisco, where the technology everyone argues about actually gets built, some people now use their dating profiles to say they would rather not meet anyone who uses it. Irene, a 25-year-old analyst who moved to the city last year, spotted the pattern within a few swipes on Hinge. It turns out to be one thread in a larger story involving money, mortality, chatbots confidants, and a 21-year-old who has hired software to flirt on his behalf.

## The dealbreaker in the prompt

Irene joined Hinge to meet people, learn her new city and find others who care about the things she does. What she found was a prompt. Hinge lets users finish the line "We'll get along if", and some of them finish it with "Fuck AI."

A few minutes later she pulled up a second profile that said it more politely and with less ambiguity: "You should *not* go out with me if... you use AI."

The people who write this sort of thing, the reporter observes, usually work in the arts, nursing, or maybe graphic design. This is a field report rather than a poll, so treat that as an impression.

The reporter spent a month moving around the city, talking to people at parties and coworking spaces and reading dating directories and online profiles. The answers varied wildly. For some, a stance on AI now sits on a profile the way party affiliation does, or the watermelon emoji that signals solidarity with Palestinians amid the war in Gaza.

Others are working through different questions:

- **The ethics of dating a builder.** Is it fine to date someone who works on AI?
- **The money.** What does a partner in the industry bring to the household?
- **The kids question.** Is it sensible to have children in a future this uncertain?
- **The handover.** Why not let the software run your love life entirely?

## Planning a life around a possible end

Kyle (a pseudonym, so he can talk openly) is in his early 20s and works as a researcher at an AI safety nonprofit initiative. He is already thinking about death. In the AI safety community, the report says, some experts see a 10 percent chance that humans go extinct because AI takes over, and that number sits at the back of every conversation he has with a potential partner.

It colors how he thinks about kids and where he might live long term, "if there is even a long term." Most people his age, he says, plan for five, 10, 15 years out. "For me, it's hard to have that planning conversation because I don't really expect a long-term future."

So he tries to have those talks in person rather than over text, which lets a date ask follow-up questions about why he sees things this way. For now, he is looking for someone whose company he enjoys. That is a modest goal, and probably the only one that survives the timeline.

Asked whether he would date someone who works in AI or tech, Kyle says he would have to think hard. He believes many people at frontier AI companies are doing things he finds pretty heinous. In his view, "They don't think very hard about the societal impacts of their actions because they're mostly in it for the paycheck or because it's interesting."

### Date-Me Docs and the singularity

JJ, 29, has quit dating apps. She uses the Date-Me Directory instead, a collection of longform, resume-like documents in which people write their own dating profiles. Two years ago she was a product designer in New York. She moved into AI safety because it fit better with her effective altruist morals.

After reading through troves of Date-Me Docs for the San Francisco and Berkeley area, the reporter noticed that quite a few mention "the singularity" as a factor in whether to have children. The term, as the report defines it, is the point at which technology overtakes human control, and some people see true artificial general intelligence as the point of no return. According to JJ, many people in AI safety use these documents specifically to find partners.

One document puts its position this way: "Don't want kids in the pre-singularity future, if there's a post-singularity future then I might want kids then." Anyone who has been told to wait for the right moment to start a family will recognize the genre, though the usual obstacle is a promotion rather than a singularity.

Other people frame it the way they frame climate anxiety, as an ethical question about bringing a child into a society being restructured by AI. JJ sums up the reasoning she sees in those documents: "Is the world good, getting to a good place, or is it bad? And if it's good, baby. If not, no baby, because you're gonna bring a baby into a world full of potential suffering."

She does not consider herself a doomer, but she was still pressed for her own singularity timeline. Her guess is 2029. She jokes that the approaching end times make her more likely to want a baby around: "If the bomb gets me, I want to be holding a shiny child."

## The view from inside the labs

Bryan, who works at a frontier AI company and also asked for a pseudonym, thinks all of this is beside the point. "I just don't think it's a factor at all. Like, it's the least [relevant] thing," says Bryan, who recently started a relationship.

To him, dating comes down to a series of ordinary factors, such as how someone talks, how they approach problems and how they have fun. He figures that working out and looking better does far more for someone's match rate than any AI or non-AI affiliation. There is one exception, which he states without embarrassment: "there are a bunch of people who are rich from AI, and, you know, being wealthy is always a plus."

The report suggests he is not wrong. With IPOs for Anthropic and OpenAI expected within the next year, matchmakers say clients have started listing "AI engineer" among their preferred qualifications. A press report cited in the piece describes one matchmaker's client booking a six-month stay in San Francisco because she might meet a partner connected to an AI company heading for an IPO.

Stability is part of the pitch. The figures the report cites:

| Figure | Detail |
|---|---|
| More than 20,000 | AI jobs added to the San Francisco Bay Area tech workforce, per a CBRE report released in August |
| 6 percent | Contraction in overall tech talent in the same area since 2022, per the same report |
| About $300,000 | Average salary of an AI researcher, which the report says can run to roughly that figure |

The two attitudes sit awkwardly together. Kyle suspects people in the industry are in it for the paycheck, and Bryan more or less agrees that the paycheck is attractive. They just disagree about whether that is a problem.

## Telling the chatbot first

Andromeda works as an EMT and a registered behavior technician, and also as an AI security analyst. She worries about the unintended consequences of talking to chatbots, especially when the subject is human connection and the question is whether the bot might replace it.

The numbers she has in mind come from a study by Elon University. It found that 27 percent of Americans turn to AI chatbots with personal, emotional or social questions. Thirty-eight percent of the surveyed group said they occasionally tell AI things they would not tell other people, romantic partners included.

She is not hostile to the habit. "Socially and psychologically, it's a big deal, and we have to talk about it," she says, and she values what a chatbot can do to help someone express themselves. "I would certainly date a man who uses AI to talk through intense thoughts and feelings and to brainstorm clarity."

Her red line is mistreating the bot. She calls that maladaptive behavior and a poor sign for good values on model welfare. Her reasoning is practical: "If they treat ChatGPT badly today, it will be me tomorrow!"

## Outsourcing the swiping

If Andromeda worries about what people tell chatbots, Joseph has gone a step further and let AI do the talking. The 21-year-old software engineer explained his setup at the Hugging Face Open Source Week party at The Midway in the Dogpatch district. Hundreds of guests wore white bucket hats and danced beside the company's round, foam emoji mascot.

Joseph works 14 hours a day launching a startup, so he has handed his Hinge swiping to an AI agent. A human virtual assistant in the Philippines manages the agent, which has been fine-tuned to mimic how Joseph speaks and acts. For each profile, it sends a tailored opening message. After some back and forth, it decides whether the woman is compatible with Joseph, and if so, it books a slot on his weekend calendar for an in-person date.

His rationale is the founder's one. "If you are a founder, you have to put a value on your time, so you can't be doing low-leverage activities. Protect your dopamine, folks," he says. The early conversations, he adds, are "a copy-and-paste sequence that happens with every girl to get to know about you, understand your backstory, and so on and so forth."

He will not publish the agent's secret sauce on GitHub, but says it is pretty easy to build one on Claude Code. Results, by his account, are "extremely successful." He is "averaging two closed conversions a week," which means two dates with two different people. He says the process has lowered what he calls "CPN," or "cost per nut." The women on the other end do not know they are talking to an agent. The point, Joseph says, is to meet in real life.

He is not an outlier in the broader market. A 2025 survey by Match and the Kinsey Institute found that 26 percent of singles in the United States use AI tools to enhance their dating experiences. Hinge now offers an AI tool that generates suggested replies, and Bumble will soon phase out swiping in favor of matchmaking through an AI assistant.

Put Joseph next to the profiles Irene keeps finding and you can see the collision coming. One side is writing "you should not go out with me if... you use AI" into a prompt, and the other has an agent working through that same prompt on a founder's behalf.

## The backlash goes analog

The existential worry has spread beyond dating. "Stop the AI" rallies have appeared at events such as Salesforce's Dreamforce convention and OpenAI's DevDay, with protesters calling out tech companies over data center development and ICE raids.

That skepticism shapes how people size each other up, even at events meant to bring the Bay Area together. At a pop-up "Date My Friend" pitch night at the Kilowatt Bar in the Mission District, a dozen people pitched their friends and exes to a crowd of more than a hundred. One pitch included the line "He works in tech, but he's unhappy about it," which drew a chorus of cheers. Other presenters joked that their PowerPoints had obviously been made by humans, not AI.

Annalise Wulf, 26, one of the series' three cofounders, says people in San Francisco try to "tech maxx" their lives, but more tech workers with crazy schedules are now interested in analog, in-person events. "It's like, 'I'm here because I'm not using an AI dating app right now. Like I'm trying to do something different,'" she says.

Cofounder Lucy Segraves, 28, agrees it is "maybe popular now to be more anti-tech." If someone works for a company people see as enabling war crimes through its technology, she might hesitate to date them.

She also has a plainer rule of thumb. "As a person with a non-tech job, I never judge someone if they do work in tech," she says. "Because that's unfortunately where the money is, and we're all just trying to make it in life."

## Related posts

- 2026.10.30 · [AI agents are acting without a paper trail, and two safety researchers want one](https://opencheese.dev/blog/ai-agents-acting-without-a-paper-trail-observability-gaps.md): Madhulika Srikumar and Claire Leibowicz of the Partnership on AI say agent frameworks do not record who an agent is, who approved its work or why it acted. They want real-time monitoring built on shared telemetry.
- 2026.11.06 · [Oracle reportedly cancels $7 billion Tencent AI compute deal after US pressure](https://opencheese.dev/blog/oracle-reportedly-cancels-7-billion-tencent-ai-compute-deal-after-us-pressure.md): The five-year plan covered about 100,000 advanced AI accelerators in Southeast Asia. The reported cancellation has not been independently confirmed.
- 2026.11.10 · [Nvidia CEO Jensen Huang Hired Crisis PR Firm Before Affair Disclosure](https://opencheese.dev/blog/nvidia-ceo-jensen-huang-hired-crisis-pr-firm-before-affair-disclosure.md): Jensen Huang hired Heller Inc. at least a year before acknowledging an affair, while Nvidia's board investigates reports that he was targeted by a foreign intelligence effort. President Trump gave conflicting details about when he learned of the alleged ties.