“If dating apps privatized how we meet, AI is beginning to privatize why we need another person at all.”
What problem Marriage was solving
For most of history, marriage was never just about romance.
It was an economic institution. A survival institution. A social institution.
People married to combine land, wealth, labor, family alliances, and social status. Love often emerged after marriage rather than serving as its prerequisite.
Industrialization slowly changed that equation.
The welfare state reduced dependence on extended families. Women entered higher education and the workforce. Urbanization weakened community ties. Individual freedom became a defining social value.
Marriage gradually shifted from being an economic necessity to becoming an emotional choice.
Today, we expect one relationship to satisfy needs that were once distributed across an entire community. A spouse is expected to be:
– a best friend,
– a romantic partner,
– a financial equal,
– a therapist,
– a travel companion,
– a co-parent,
– an intellectual peer,
– an emotional safe space,
– an exciting lover.
Never before have humans expected so much from one person. Ironically, as marriage became freer, it also became more demanding.
The Optimization Trap
Technology amplified this transformation.
Dating apps didn’t merely give us more people to choose from. They encouraged us to believe that there exists an optimal partner.
The idea of an optimal partner entered the equation.
Height. Income. Education. Career. Religion. Politics. Fitness. Lifestyle. Travel. Hobbies. Location. Ambition. Each preference seems reasonable on its own. Together, they create an increasingly complex optimization problem.
But optimization works beautifully for products. Human relationships don’t.
A product has specifications. A person has contradictions.
A product’s performance can be measured before purchase. Compatibility between two people often emerges only after they spend time together.
Relationships are dynamic systems. Two people create compatibility through shared experiences, compromise, forgiveness, conflict, and growth.
Modern dating reverses that sequence. We increasingly seek certainty before commitment. But relationships often create certainty through commitment. We want to know whether someone is compatible before investing in them, even though some of the most important dimensions of compatibility cannot be observed until we actually invest.
The result is a strange paradox: The more precisely we try to optimize for compatibility, the more we risk filtering out the very relationships in which compatibility could have developed.
The optimization mindset becomes even more complicated when two highly independent individuals enter a relationship. Modern relationships increasingly resemble negotiations between two complete economic units.
Both partners may have:
– careers
– financial independence
– personal ambitions
– separate social circles
– separate identities
– individual goals
These are remarkable achievements of modern society. They change the economics of relationships. In economics, every decision involves opportunity cost. Time spent nurturing a relationship competes with:
– Career growth.
– Fitness. Education.
– Travel.
– Entrepreneurship.
– Personal branding.
– Mental wellness.
– Social media.
Relationships are no longer competing only against other relationships. They are competing against every alternative use of our time. This makes emotional labor one of the scarcest resources in modern life. And that changes the perceived economics of commitment. We increasingly expect certainty before we are willing to make the investment.
The Unbundling of Marriage
Perhaps the most profound transformation of the twenty-first century is that marriage is losing its monopoly over human needs.
Historically, marriage bundled together several essential functions:
– Companionship
– Sexual intimacy
– Economic partnership
– Child-rearing
– Emotional support
– Social recognition
Technology and modern institutions have gradually begun to unbundle these functions.
Food arrives through delivery apps.
Entertainment comes from streaming platforms.
Friendship exists in online communities.
Professional networking happens on LinkedIn.
Children can be conceived through IVF or donor programs.
Sex increasingly exists outside long-term commitment.
Each service replaces one small reason people once depended on marriage. None of these developments eliminate relationships. But together they reduce the functional necessity of having one. This is an important distinction. Marriage increasingly survives not because society requires it, but because two individuals continue choosing it. That makes relationships simultaneously more authentic and more fragile.
When a relationship no longer provides essential economic or social functions, its survival depends increasingly on whether it continues to deliver emotional value.
And this brings us to an interesting product question:
What exactly are people hiring relationships to do?
AI Is Solving a Different Problem – Complicated Relationships
This is where the Jobs-to-be-Done framework provides a useful lens.
In product development, customers don’t simply buy products. They hire them to make progress in a particular circumstance. A JTBD approach asks us to look beyond the product itself and understand the underlying functional, emotional, and social job a customer is trying to accomplish.
Let’s try to apply the same framework to relationships.
Instead of asking: “Why does someone want a partner?”
ask: “What job is someone hiring a partner to do?”
The answer is not one job. It is a bundle.
The Functional Job
A relationship can provide practical companionship.
– Someone to share a household with.
– Someone to raise children with.
– Someone to travel with.
– Someone to help make decisions.
– Someone to share responsibilities.
Historically, marriage bundled many of these functions together. But technology has gradually created alternatives for several of them.
The Emotional Job
This may be the most important one.
When someone feels lonely, stressed, anxious, rejected, bored, or emotionally exhausted, they may be hiring another person to provide:
“Help me feel understood, valued, and less alone.”
This job doesn’t necessarily require marriage.
It doesn’t even necessarily require dating. And increasingly, it doesn’t require another human.
The Social Job
Relationships also perform social functions.
People want to feel desirable. Wanted. Accepted. Successful.
They want someone to share experiences with and someone who validates their identity within society.
Again, technology can increasingly perform parts of this job. Social media provides recognition. Online communities provide belonging. Dating apps provide validation. AI companions can provide personalized attention.
The interesting question is therefore not whether AI can replace relationships. It is:
Which jobs traditionally performed by relationships can AI perform well enough to become an alternative?
From Dating Apps to AI: The Job Is Changing
Dating apps emerged because people had a problem:
“Help me find someone I might want to have a relationship with.”
Dating platform solved the discovery job. But discovery was never the only job. People ultimately wanted companionship. They wanted someone to talk to at the end of the day. Someone who remembered their stories. Someone who cared about what happened at work. Someone who could celebrate their successes. Someone who would be there when they were struggling.
Dating apps solved the problem of access. They didn’t necessarily solve the problem of connection.
And this is where AI enters.
An AI companion can increasingly perform several of the emotional and social jobs that people traditionally hired relationships to perform.
When I’m lonely, I can talk to it.
When I’m stressed, it can listen.
When I’m bored, it can entertain me.
When I need validation, it can provide it.
When I want to talk at 2 a.m., it is available.
The circumstance is different from traditional dating, but the underlying job is surprisingly similar.
The customer is still trying to make emotional progress. The product has simply changed.
Frictionless Relationships – Why AI Feels Better Than People
This raises an uncomfortable question. If humans are naturally social creatures, why would anyone choose an artificial relationship? The answer is surprisingly simple. Human relationships are emotionally expensive. They demand:
– Patience.
– Compromise.
– Forgiveness.
– Listening.
– Growth.
– Disappointment.
– Conflict.
– Repair.
AI asks for none of these things. It delivers emotional validation without emotional risk.
In behavioral economics, humans consistently prefer lower-friction experiences. One-click purchasing replaced physical shopping. Streaming replaced DVDs. Food delivery replaced cooking. Remote work replaced commuting. AI companionship applies the same principle to emotional life. It dramatically reduces friction.
From a product-design perspective, this is extraordinary. From a sociological perspective, it is much more consequential.AI isn’t simply creating a new category of software. It is beginning to compete for the same jobs traditionally performed by human relationships.
The danger isn’t that AI becomes conscious. The danger is that it becomes convenient. Convenience has reshaped nearly every human behavior over the past fifty years. Love may not be an exception.
The Market Doesn’t Care Whether Love Is Human
Capital follows unmet demand. And loneliness is rapidly becoming one of the largest unmet emotional needs in developed societies. Whenever markets encounter unmet demand, they search for a product that can satisfy it.
Dating apps monetized introductions.
Social platforms monetized attention and belonging.
Therapy platforms monetized emotional support.
AI platforms are beginning to monetize companionship itself.
The progression feels almost inevitable. The customer has a job. The market identifies the job. Technology finds a cheaper, faster, or more scalable way to perform it
It can scale intimacy without scaling another human being.
This is where the economics become radically different. The marginal cost of emotional availability approaches zero. And once something becomes abundant, the market can transform it from a scarce human experience into a consumer product.
But What Happens When the Job Is Too Easy?
There is, however, a deeper problem. JTBD helps us understand why someone might hire an AI companion. But it doesn’t tell us whether the product is actually good for the customer in the long run.
A product can solve the immediate job while creating a larger problem.
A cigarette can solve the job of taking a break, calming down, and socializing.
A food-delivery app can solve the job of getting dinner without cooking.
A social network can solve the job of staying connected.
But satisfying the immediate job does not necessarily mean creating the best long-term outcome. The same question applies to AI companionship.
The question is no longer whether AI companions will improve. They will.
The more important question is whether human relationships can continue competing against experiences explicitly designed to maximize comfort and minimize discomfort. Because that has never been what love was. Love was never optimized. It was negotiated. Messy. Slow. Unpredictable. Wonderfully inefficient. And perhaps that inefficiency was never the bug. It was the feature.
The future of love is therefore not really a question about whether AI will replace humans. That framing is too simplistic.
The question we should be asking:
Which human needs will we choose to satisfy through machines, and which ones will we continue to satisfy through other people?
That is a product decision. It is also a cultural decision. And eventually, it becomes a societal one.
We may reach a world where an AI can remember every conversation, anticipate every emotional need, adapt perfectly to our personality, and provide companionship whenever we want it.
It may become the perfect partner by every measurable definition.
But there is one thing it cannot give us:
The more technology optimizes relationships for convenience, the more valuable the unoptimized parts of love may become.
The future may give us perfect digital companionship.
But perfect companionship is not necessarily the same as intimacy.
And perhaps, in an age of algorithms that can give us almost everything we ask for, the rarest human experience will be something much simpler:
being loved by someone who doesn’t have to.
Comments