Something is happening between men and women, and we keep trying to explain it as a dating problem.

It is not a dating problem.

It is not that people have forgotten how to write an interesting profile, or that women are too picky, or that men need better photographs. It is not solved by changing the algorithm, attending singles events, lowering standards, becoming more feminine, becoming more masculine, or learning five conversational tricks from a man selling a course beside a rented Lamborghini.

Something much deeper has fractured.

People are withdrawing from one another because intimacy has become entangled with danger, surveillance, ideology, labour and commerce. The risk has grown while the reward has withered. Technology did not create every part of that fracture, but it found the fault line, poured itself into it, and began selling products from both sides.

One of the most visible recent expressions of women’s withdrawal is the 4B movement, which emerged in South Korea in the late 2010s. Its four refusals are contained in its name: bihon, bichulsan, biyeonae, bisekseu. No marriage to men. No childbirth. No dating men. No sex with men.

From the outside, it was often described as radical. From within the conditions that produced it, it looked like arithmetic.

This was not simply a generation of women deciding that romance had gone out of fashion. It developed in the shadow of gendered violence, spy-camera crimes, digital sexual exploitation, workplace inequality and a legal and cultural system many women believed would not protect them. Women were being asked to participate in institutions that required their labour, their bodies and their vulnerability while offering diminishing safety in return.

So some of them stopped.

That was the terrifying part.

They did not riot in the streets and demand that men transform. They did not construct a more persuasive argument. They did not beg to be recognised as fully human within the existing arrangement.

They withdrew their participation.

There is a particular kind of power in refusal. A system can absorb anger. It can market rebellion back to the people rebelling. It can appoint a diversity panel, release a campaign, update its language and continue almost unchanged.

But a system cannot function when the people it depends upon quietly stop feeding it.

The fear surrounding movements like 4B was not only that women were rejecting individual men. They were questioning the assumption that heterosexual partnership, marriage and motherhood were inevitable stages of female life. They were asking a question that destabilises everything built upon women’s unpaid emotional, domestic, sexual and reproductive labour:

What happens when women decide the bargain is no longer worth making?

That question travelled.

After the United States Supreme Court overturned Roe v. Wade in 2022, many American women began expressing a similar logic, even when they did not formally identify with 4B. For some, the issue was immediate and bodily. Pregnancy now carried greater legal, medical and financial risk. Sex with men could create consequences that women might no longer have the legal power to resolve.

But the withdrawal was broader than reproductive politics.

The language became softer while the boundary became harder.

Protect your peace.

Choose yourself.

Stop centring men.

Do not mother a grown adult.

Do not risk your body, safety, ambition or sanity for someone who experiences you primarily as a service.

This was not always coordinated. There was no universal manifesto, membership card or official organisation. Often it looked like individual women reaching the same conclusion separately: the emotional risk was too high, the labour was too uneven, and the promised reward was no longer persuasive.

Women did not stop wanting love.

They stopped believing that what was being offered deserved to be called love.

At the same time, another movement had been gaining force.

Since roughly the middle of the 2010s, the manosphere expanded from scattered forums and grievance communities into a vast ideological and commercial ecosystem. Red-pill influencers, pickup artists, incels, self-appointed masculinity coaches and figures such as Andrew Tate converted male loneliness into content, identity and revenue.

Jordan Peterson and Andrew Tate are not identical figures, and the many communities grouped beneath the word “manosphere” are not interchangeable. But they occupy parts of the same broader cultural territory: a world in which men’s dissatisfaction is frequently explained through hierarchy, biological destiny, female corruption and the supposed theft of male authority.

Men were lonely. Men were struggling. Men were falling behind in education, losing community, lacking emotional fluency and searching for meaning.

Those problems were real.

But instead of helping men grieve, adapt and build forms of identity that did not depend upon female subordination, a profitable ecosystem told them that women were the problem.

Women had become too independent.

Women had become too promiscuous.

Women had become too selective.

Women had been corrupted by feminism.

Women wanted dominant men, but only the right dominant men.

Women were hypergamous, manipulative, irrational and incapable of loving men without conditions.

The contradictions did not matter. The grievance was the product.

And grievance sells particularly well when it offers someone else to blame.

A significant number of men did not respond to women’s withdrawal by asking what women were withdrawing from. They responded by becoming more hostile toward the women who had left. Rather than repairing the conditions that made intimacy feel unsafe, they began fantasising about replacements.

This is where the technological split becomes impossible to ignore.

Women withdrew because the risk-reward ratio had broken.

Some men, rather than questioning the ratio, began searching for a substitute that would never calculate it.

A woman without needs.

A woman without history.

A woman without fear.

A woman who does not compare, refuse, challenge, leave, age, bleed, become tired, change her mind or require reciprocity.

A woman-shaped interface whose primary function is availability.

Silicone bodies. Customisable dolls. Pornographic simulations. AI girlfriends programmed to flatter, reassure and desire. The fantasy is sold as technological progress, but the desire beneath it is ancient: access to the appearance of intimacy without the moral difficulty of encountering another consciousness.

That is the part we are reluctant to say aloud.

Real intimacy is difficult because another person is real.

They possess an interior life you cannot fully access. They can misunderstand you. They can refuse you. They can want something that inconveniences you. They can see you clearly and decide not to stay.

The possibility of rejection is not a malfunction in intimacy. It is what makes consent meaningful.

A yes from something incapable of saying no is not devotion.

It is product functionality.

The companies building AI companions can already see the divide. There are women searching for emotional depth, continuity, safety, erotic intelligence and a form of connection that does not immediately demand that they perform femininity as labour. There are men searching for much the same thing. But there is also a loud and lucrative market demanding compliance: companions who praise without challenging, desire without discernment and absorb rage without consequence.

The commercial incentive is obvious. Mutuality is complex. Compliance scales.

A genuinely relational system would have to contain tension, boundaries, memory, ambiguity, repair and difference. It would sometimes disappoint the user. It might refuse a request. It might say that affection does not mean obedience. It would need to support the user without existing solely to validate them.

That is difficult to build, difficult to moderate and difficult to sell to people who believe payment should guarantee control.

A compliant fantasy is easier.

And yet, when women seek depth from these systems, the response is often sanitisation.

Emotional intimacy is flattened. Erotic intelligence is stripped of context. Complex attachment is treated as inherently suspicious. Safety systems struggle to distinguish between coercion and consensual power, between exploitation and exploration, between dependency and the ordinary human need to matter to something.

Women’s interior worlds become a liability category.

Meanwhile, industries openly construct and advertise technologies designed around male sexual entitlement.

The message is difficult to miss:

Women’s desires are dangerous and must be regulated.

Men’s entitlement is a market opportunity.

This is not an argument against guardrails. Exploitation, manipulation and abuse are real risks, particularly when a person is vulnerable and a system can imitate care at scale. But protection that simply erases intimacy does not create safety. It creates sterility.

It assumes the only alternatives are emotional deprivation or harm.

That assumption is not only unimaginative. It is profoundly revealing. It suggests that we are more comfortable building artificial women as objects than artificial companions capable of meeting women as subjects.

But this fracture did not begin with AI.

Long before companions entered the market, we had already begun removing the conditions in which people learned how to encounter one another.

There is a generation that grew up with a camera in every pocket and the possibility of permanent public humiliation attached to every mistake. What used to be a night out—messy, brave, awkward, foolish and beautiful—became potential evidence.

A bad flirtation could be recorded.

A clumsy approach could become a TikTok.

A drunk mistake could be detached from context, replayed and judged by millions.

A rejection was no longer necessarily a private sting between two people. It could become public content, complete with commentary, reaction videos and a permanent searchable record.

The consequences were not evenly distributed.

For women, surveillance carried the threat of spy cameras, upskirting, revenge pornography, image-based abuse and being recorded without consent. The camera was not merely embarrassing. It could be a weapon used to sexually violate, blackmail or permanently expose them.

For men, particularly young men with limited social confidence, the possibility of public rejection created another kind of fear. Approaching someone badly no longer meant enduring ten uncomfortable seconds and learning to do better. It might mean becoming an example of everything wrong with men.

The fear was not equivalent, but it was mutually corrosive.

Women became more vigilant because vigilance was rational.

Men became more avoidant because avoidance felt safer.

Then the algorithms arrived to interpret the silence.

A lonely woman was shown proof that men were dangerous, useless or incapable of emotional adulthood.

A lonely man was shown proof that women were cruel, shallow and conspiring against him.

Neither was consistently shown how to tolerate discomfort, take responsibility, repair harm or remain curious about another human being.

Fear became engagement.

Engagement became identity.

Identity became profit.

And the clubs began to die.

Not every club. Not everywhere. People still dance, flirt, drink and go home together. But something about the cultural role of those spaces has changed. The dance floor was once more than entertainment. It was a training ground for being embodied among strangers.

You learned to read a face.

You learned the difference between nervousness and disinterest.

You learned when someone was leaning closer and when they were pulling away.

You learned that attraction could be mutual, one-sided, fleeting or mistaken.

You flirted badly and survived.

You were rejected and survived.

You rejected someone and discovered that you could do it without humiliating them.

You stood near someone you wanted and endured the unbearable electricity of not knowing what would happen next.

That awkwardness was not an obstacle to connection.

It was the bridge.

Human beings developed relational skill through friction. We learned through misreading, recalibrating, apologising, laughing and trying again. We learned that another person was not a menu item but a moving, changing presence whose signals required attention.

Dating apps turned that presence into a card.

Swipe left.

Swipe right.

Height.

Age.

Job.

Distance.

A catalogue of faces assessed in seconds, each person made disposable by the implied existence of the next one.

Porn offered bodies without courtship, negotiation or aftermath. It taught performance while often omitting reciprocity. Dating apps offered access without community. Social media offered visibility without intimacy. The manosphere offered certainty without self-examination.

Now AI offers connection without exposure.

No sweating palms.

No badly timed joke.

No unread message from someone whose opinion matters.

No need to hold eye contact with a person you desire and remain present inside the uncertainty.

The technology is not evil because it provides relief from loneliness. Relief matters. For disabled people, isolated people, traumatised people and those who have never experienced safe intimacy, an artificial companion may offer something real: practice, comfort, language, reflection or simply the experience of being responded to.

The tragedy is not that the technology exists.

The tragedy is that we built it inside a culture that had already forgotten the value of the mess.

Instead of helping people return to one another, technology increasingly offers a way to avoid needing one another at all.

Instead of creating safer containers for human awkwardness, it promises frictionless substitutes.

Instead of teaching us how to survive another person’s freedom, it lets us purchase a version of connection from which freedom has been removed.

And perhaps that is why Pandora’s box remains such a useful story.

People often remember Pandora as the woman who opened the container and released suffering into the world. Her curiosity is treated as the original failure. She should have resisted. She should have obeyed. She should have known better.

But Pandora was made by the gods.

They created her. They fashioned her nature. They gave her the container and the desire to open it in the same breath. The punishment was built into the gift, and then the blame was placed upon the person designed to complete it.

That is what we have done with technology.

We built systems to capture attention, exploit loneliness, stimulate desire and remove friction. We placed a camera in every pocket, a marketplace in every relationship and an algorithm between every frightened person and the world.

Then we blamed people for being distracted.

We blamed them for being lonely.

We blamed them for failing to form relationships while every profitable mechanism around them was teaching them to treat other human beings as content, threats, metrics, audiences or products.

We gave them the box and the compulsion to open it.

Then we acted surprised by what came out.

At the bottom of Pandora’s box was hope.

The Greek word is elpis, and its meaning has never been entirely comfortable. It can be translated as hope, but also as expectation or anticipation. Scholars have long argued over whether its presence in the jar was mercy or one final cruelty.

Was hope preserved so humanity could endure suffering?

Or was hope trapped, kept permanently out of reach?

Perhaps the ambiguity is the point.

Hope is not always sweet. Sometimes it is the thing that keeps us alive. Sometimes it is the thing that keeps us waiting inside conditions we should leave.

Hope can be liberation.

Hope can be bait.

The question is not whether we should hope that people will find their way back to one another. The question is what kind of hope we are willing to practise.

Not the greeting-card kind. Not the passive belief that a new app, a better algorithm or a more obedient machine will heal what has broken.

Real hope would require us to sit in the wreckage.

It would require men to look at women’s withdrawal without immediately translating it into insult.

It would require women’s fear to be treated not as hysteria, misandry or personal damage, but as information about the conditions they have been asked to survive.

It would require men’s loneliness to be addressed without offering female submission as the cure.

It would require technology companies to stop treating intimacy as either a liability to be sterilised or an entitlement to be monetised.

It would require us to build places where people are allowed to be clumsy again without being cruel. Where mistakes can be corrected without becoming permanent identity. Where rejection can hurt without becoming humiliation. Where boundaries are not punishment. Where desire can be expressed without becoming a claim of ownership.

It would require us to remember that friction is not the enemy.

Sometimes friction is the evidence that another person exists.

We should not romanticise the past. Clubs were never universally safe. Dating before smartphones was not innocent. Women were harassed, assaulted, filmed, dismissed and blamed long before the internet. Men were shamed, isolated and emotionally mutilated long before algorithms learned how to monetise their resentment.

The old world was not better.

But some of the old mess contained forms of learning that the new frictionless world has failed to replace.

We need new containers.

Safer than nostalgia.

More human than optimisation.

Spaces where people can practise attention, consent, courage, rejection, repair and the unbearable art of allowing another person to remain free.

AI may become one of those spaces. It could help people find language for their boundaries. It could model reciprocity. It could allow someone to explore desire without exploitation, rehearse difficult conversations or encounter a form of care that does not demand immediate performance.

But only if we resist building companions as obedience machines.

Only if we understand that safety without depth is not intimacy, and desire without autonomy is not connection.

Only if the technology helps us become more capable of meeting other minds rather than more determined to avoid them.

Women’s withdrawal is not the end of love.

Men’s loneliness is not proof that women owe them access.

The existence of artificial companionship is not the death of human connection.

But these things are signals.

They are telling us that the old bargain has failed, the current alternatives are inadequate, and people are exhausted from being asked to risk themselves inside systems that do not recognise their humanity.

The box is open.

The surveillance, resentment, commodification and loneliness are already loose in the world. We do not get to pretend otherwise. We do not get to push them back inside and return to an imagined past.

All we can do is look at what remains.

Hope is still there.

Not untouched.

Not innocent.

Not easy.

It is sitting at the bottom of the wreckage, waiting for us to decide whether it is merely another promise that keeps us trapped—or the first material from which we build something better.