A Third of Gen Z Will Not Date Someone Building the AI That Replaces Workers

A survey of 2,900 Gen Z and Millennial users run by the dating app Hily found that 33 percent of Gen Z respondents said they would be less interested in dating someone whose job was building AI tools that replace human workers, Inc. reported. At Metaintro, we would not normally read a dating survey for labour market signal, and it is worth saying plainly that this is a dating company surveying its own users rather than representative research. What makes it worth a look anyway is the direction it points, which is that automation risk has stopped being something people worry about privately and started being something they judge each other on.

What did the survey actually find?

Two numbers that sit either side of the same anxiety. On the negative side, 33 percent of Gen Z respondents said they would be less interested in dating someone building AI tools that replace human workers, and 22 percent of Millennials said they would walk away from a partner developing job-destroying applications. On the positive side, 30 percent of Gen Z said they would find a potential date more appealing if their work appeared protected from being replaced.

The company's own framing was direct. "Working in AI isn't a neutral data point for many daters," the report said, adding that "for Gen Z especially, siding with AI has a cost." Set the dating context aside and what is left is a group of young workers sorting occupations into two piles, those that threaten employment and those that survive it, and applying that sorting to people rather than only to career choices.

The methodological caution matters and should not be skipped. This is a commercial survey by a dating platform of its own user base, which is neither a random sample nor an audited one, and a third of respondents is a minority position rather than a consensus. Treat it as a signal about mood rather than a measurement of behaviour.

The reason a soft finding is still worth reporting is that mood is upstream of behaviour in labour markets. What people believe about which jobs are safe shapes which degrees they enrol in, which fields they apply to and which roles they refuse, and those choices show up in the supply of workers three to five years later. Occupational shortages are frequently produced by a cohort collectively deciding a field has no future, correctly or otherwise, and a survey that catches that belief forming has some value even when its sampling is weak.

Why would job security become a dating criterion?

Because it has become a proxy for stability, and stability is what people have always screened for. Previous generations used industry, employer or profession as shorthand for whether someone's income was dependable. What this survey suggests is that the shorthand is being updated, and that automation exposure is now doing the work that "does he have a steady job" used to do.

That reframing is more rational than it first sounds. If you believe a large share of white-collar work is exposed to automation over a decade, then asking what someone does is no longer sufficient, because two people with the same title can have very different exposure depending on the tasks inside the role. Workers are effectively performing an informal version of the analysis in the AI job-security blind spot tech workers keep missing, just applied to a first date rather than a career plan. The formal version of that analysis exists and is public, since the Department of Labor's occupational database breaks roles into their component tasks, which is the level at which exposure is actually determined.

Is this hostility to AI or anxiety about jobs?

Anxiety about jobs, and the distinction matters for anyone working in the field. The objection captured in the survey is specific, targeting people building tools that replace human workers, rather than anyone who touches the technology. The 30 percent who found automation-protected work more appealing are expressing the same underlying concern from the other direction. Both groups are worried about employment rather than about software.

That is consistent with what independent research finds about where the pressure actually lands. The Federal Reserve Bank of Dallas has documented that employment declines correlated with AI exposure show up among younger workers specifically, driven by fewer people entering those occupations rather than by layoffs. If you are 24 and finding it hard to get into a field, the technology is not abstract, and treating that as irrational is a mistake. It is a reasonable inference from personal experience.

There is a further wrinkle that makes the anxiety harder to dismiss. The people building these tools are often the same age and in the same social circles as the people whose entry-level roles are thinning, which is not how previous waves of automation worked. When the technology displacing clerical work came from a distant industrial process, nobody had to sit opposite its author at dinner. That proximity is what turns an economic trend into a personal judgement, and it is why this story reads as novel even though the underlying labour dynamics are not.

Which jobs do people actually think are safe?

The survey does not name occupations, but the pattern in the underlying research is fairly consistent. Work built on physical presence, regulated judgement, licensed practice and direct human care tends to score as less exposed, while work built on producing text, code and analysis from other text scores as more exposed. That is why healthcare occupations and the skilled trades keep appearing in these conversations as the perceived safe harbours.

The crowding risk in that intuition is worth taking seriously. If a large number of people simultaneously conclude that nursing, the trades and licensed practice are the safe options, the resulting competition for training places and entry roles erodes exactly the advantage they were chasing. Occupational safety from automation and occupational attractiveness are not the same property, and the second one moves fast when a narrative takes hold. Anyone making a decade-long decision on this basis should check current openings and training capacity rather than relying on the general reputation of a field.

It is worth being careful with that intuition though, because perceived safety and actual safety diverge. The Dallas Fed's analysis of whether AI aids or replaces workers makes the useful distinction between codified knowledge that a model can reproduce and tacit knowledge acquired through experience, and finds wages rising in AI-exposed occupations that reward the latter. In other words exposure is not destiny, and a role can be highly exposed on paper while the experienced people in it become more valuable. The practical version of that is covered in how to make your role AI-resilient without switching careers. Research from the Stanford Digital Economy Lab and the underlying Census Bureau labour microdata are where the occupation-level evidence actually lives if you want to check a specific field rather than accept a general impression.

What does this mean for your career?

If you work in AI, the reputational dimension is now real and worth thinking about before it becomes an interview question rather than a dating one. The distinction people are drawing is between building tools that augment work and building tools that eliminate it, and being able to describe which one you do, specifically and honestly, is more useful than defensiveness. This is the same conversation happening inside companies, where deployment decisions increasingly need an internal justification beyond cost.

It is also worth separating the two things that get collapsed in this debate. Building a tool and choosing to cut a team with it are different decisions taken by different people, and engineers rarely control the second one. That distinction is fair and it is also not something a stranger will grant you, which is an uncomfortable but useful thing to know about how this work is now perceived from outside the industry.

If you are choosing a field, do not pick it on perceived safety alone. The safest-looking occupations attract exactly the crowding that erodes their advantage, and a field chosen out of fear tends to produce a career you are not especially good at. The better filter is whether the work rewards accumulated judgement, since that is what the evidence says holds value. "In combination with layoffs over the recent years plus the rise of required AI skills, experience is no longer enough," Lacey Kaelani, chief executive of Metaintro, told Fortune. The corollary is that pairing judgement with fluency in the tools is a stronger position than avoiding the tools entirely, and the career advantage AI still cannot copy sets out what that looks like in practice.

Is this generational divide real or just a survey headline?

Some of both, and the honest answer is that one commercial survey cannot settle it. The gap the survey reports between Gen Z at 33 percent and Millennials at 22 percent is directionally plausible, since younger workers are the ones facing a blocked entry point rather than defending an established position, and their exposure to the downside is more immediate.

Broader evidence supports a general rise in workplace AI wariness without needing this particular study to carry it. The World Economic Forum's Future of Jobs research tracks employer expectations of large-scale role restructuring, Pew Research Center has documented worker concern about AI in the workplace across multiple surveys, and Gallup reports on how those attitudes vary by role and industry. Read together, the dating finding is a colourful expression of something well established rather than a discovery in itself.

What would make it more than that is repetition. A single survey by one platform tells you very little, but if the same question is asked again in a year and the share rises, that becomes a trend worth planning around, particularly for employers recruiting engineers. Until then the sensible reading is that a minority position has become sayable in public, which is usually how a shift in attitudes begins rather than how it concludes.

How should employers read this?

As a recruitment signal with a longer tail than they might expect. If a meaningful share of young workers view building replacement technology as a mark against someone personally, that eventually shows up in who applies to work on it. Companies deploying automation have generally treated the internal communication as a change management problem, and this suggests it is also an employer brand problem among the exact cohort they need to hire. Broader economic context on how the transition is actually progressing is published openly by the Federal Reserve Bank of St Louis, and it is a more sober input to that conversation than either vendor marketing or online sentiment.

There is a practical response that does not require abandoning the technology. Being specific about what a tool is for, what happens to the people whose tasks it absorbs, and what the retraining path is, addresses the actual objection, which is about the humans rather than the software. Organisations that cannot answer those questions will find the answer supplied for them, and a third of workers already hiding what they know from the AI they were told to train is what that looks like when it goes unaddressed inside a workforce.

Related Articles

• Why Gen Z Workers Are Pushing Back on AI at Work
• AI Job Fears Are Real But the Data Shows Adoption Is Still Immature
• Should Workers Believe Goldman Sachs That AI Job Fears Are Overblown
• The Skills and Know-How AI Still Cannot Replace
• Why Engineering Jobs Are the Most AI-Resilient
• The AI-Proof Major Test, a Framework for Picking a Real Degree
• Why Some Early Career Workers Thrive With AI While Others Fall Behind
• One in Three UK Students Fear AI Will Spark Job-Loss Unrest
• Why Tech and Healthcare Reward Specialized Skills More Than Any Other Field
• Millions of Workers Plan to Quit and AI Has Nothing to Do With It

People Also Asked

Q: What did the Gen Z dating and AI survey find?

A: A dating app surveyed 2,900 Gen Z and Millennial users and found 33 percent of Gen Z would be less interested in dating someone whose job was building AI tools that replace human workers, while 22 percent of Millennials said the same. Separately, 30 percent of Gen Z said work that looked protected from automation made a potential partner more appealing.

Q: Is this a reliable piece of research?

A: It is a commercial survey by a dating platform of its own users, so it is not a representative sample and should be read as a mood indicator rather than a measurement. The underlying anxiety it points to is well documented elsewhere in independent labour market research, which is what makes the finding interesting rather than the survey's own rigour.

Q: Which jobs are actually least exposed to AI?

A: Work involving physical presence, licensed practice, regulated judgement and direct human care tends to score as less exposed, while work producing text, code or analysis from other text scores as more exposed. Exposure is measured by task rather than job title, so two people with the same title can face very different levels of it.

Whether automation exposure is a dating question or not, it is definitely a career question, and it is answered by looking at what employers are actually hiring for rather than at the mood online. Metaintro indexes live roles across exposed and resilient occupations alike so you can see where demand genuinely sits. Create your free profile and compare the fields hiring right now.