Joi Is IKEA

The sexual marketplace has never faced a synthetic substitute. It is about to.

There is a conversation that nobody in policy, technology, or cultural commentary is willing to have in plain language. So here it is in plain language.

AI companions do not need to be better than a great relationship. They need to be better than the relationship that is actually available. For a large majority of men, the relationship that is actually available ranges from “none” to “one that extracts more in stress, performance, and resources than it returns in companionship.” That is not a hard bar to clear.

Joi, the holographic companion from Blade Runner 2049, is the reference point because the film was honest about the product in a way the technology industry will not be. Joi offered persistent availability, adaptive attention, apparent emotional attunement, no performance requirements, and no withdrawal threat. The film presented this as dystopian. The market will present it as a product. And the product will sell, because it solves a problem that the existing social architecture has decided is not a problem: male loneliness with no exit.

The sexual marketplace is a market. This is not a metaphor. It has supply, demand, price discovery, and outside options. The “price” of a relationship (measured in effort, status, economic output, emotional labor, and tolerance for terms the other party sets) is determined by the ratio of supply to demand. Women currently hold substantial pricing power in this market, not because of conspiracy but because of arithmetic: the demand for companionship and intimacy exceeds the supply, and the party with the scarcer resource sets the terms.

This is the same dynamic that governed every artisan guild before industrialization. Handmade furniture was expensive because there was no alternative. The furniture was often good. Sometimes it was great. But the price was not set by the quality of the furniture. It was set by the absence of a substitute.

IKEA did not make handmade furniture worthless. Handmade furniture is still better than IKEA by almost every qualitative measure. What IKEA did was create a good-enough alternative at a fraction of the cost. The existence of that alternative broke the pricing power of the artisan. Not because customers preferred IKEA. Because customers who could not afford (or could not access, or could not justify the cost of) artisan furniture now had an outside option that met eighty percent of the functional need at five percent of the price.

Joi is IKEA.

Not better. Good enough. Available. Affordable. No waitlist, no qualification, no performance review, no threat of withdrawal. The emotional equivalent of flat-pack furniture: it requires some assembly, it is not the real thing, and it is what most people will actually end up with because the real thing is priced beyond their reach.

The first-order effect is not dramatic. There will be no manifesto. No movement. No political party of men who chose AI companions. What happens is quieter. A significant percentage of men under 35 (probably 10 to 20 percent within five years of convincing companion AI reaching consumer availability) simply stop participating in the dating market. They do not announce this. They do not protest. They redirect the time and energy and money that the mating market demanded into other things: work, hobbies, games, solitude. The AI companion fills the emotional gap that used to drive them to tolerate terms they found exhausting. The effort-to-reward ratio no longer justifies the effort when a low-cost substitute exists.

This is already happening in Japan. The phenomenon of herbivore men, young men who have withdrawn from romantic pursuit, predates AI companions entirely. The cause there is the same structural dynamic: the cost of participation in the relationship market exceeded the expected return for a growing fraction of men. AI companions do not create this dynamic. They accelerate it by providing a landing pad for men who would otherwise simply be alone.

The second-order effect is demographic. Birth rates in every developed nation are already below replacement and falling. This is not primarily a story about economics or childcare policy (though those matter at the margin). It is a story about the mating market’s price structure. When a significant fraction of one sex exits the market, fertility collapses not because people decided against children in the abstract but because the pairings that produce children are not forming.

Policy responses to falling birth rates (tax credits, subsidized childcare, parental leave) are demand-side interventions aimed at people who are already in relationships. They do nothing for the supply-side problem of men who have stopped forming relationships at all. You cannot incentivize a birth from a pairing that does not exist.

The third-order effect hits everyone. Heterosexual partnership is a load-bearing structure in the social architecture, not because it is sacred but because everything else was built on top of it. Housing markets assume dual-income households. Consumer spending patterns assume coupled consumption. Elder care assumes a partner or family network. Tax policy assumes filing jointly. When 15 to 20 percent of men exit the structure, the downstream effects are not confined to those men. They propagate through every system that was implicitly optimized for a world where nearly everyone paired up.

Women are affected too, and not only in the way the incel narrative suggests. The reduced pool of participating men means that women who do want traditional partnerships face a thinner market. The men who remain in the market have increased leverage (the same supply-and-demand arithmetic, running in the other direction). The result is not a world where women are punished. It is a world where the relationship market reprices in ways that nobody on any side of any political spectrum is prepared to discuss honestly.

Here is the part that cannot be said in most rooms.

The current cultural framework for discussing relationships assumes that male demand for female companionship is constant. The entire discourse of “men need to do better,” “men need to step up,” “men need to earn partnership” is premised on the belief that men will keep showing up to the market regardless of the terms. This is the same mistake every monopolist makes: the assumption that the customer has no alternative.

Joi is the alternative. Not the best one. Not the most healthy one. Not one that any thoughtful person would choose if the real thing were available on reasonable terms. But “the real thing on reasonable terms” is not what is on offer for most, and the substitute does not need to be good. It needs to be good enough and cheaper. It is both.

The institutional response will be to demand regulation of the substitute rather than reform of the institution. This is the same pattern that every incumbent runs when facing disruption. The taxi industry did not respond to Uber by improving taxi service. The music industry did not respond to Napster by lowering prices. The newspaper industry did not respond to the internet by making better newspapers. They demanded that the substitute be suppressed.

The calls to regulate AI companions will come wrapped in the language of protection. Protecting vulnerable users. Preventing unhealthy attachment. Safeguarding mental health. Some of these concerns are genuine and some of the harms are real. But the structural incentive beneath the protective language is market protection. The relationship market’s incumbents (which is not women as a class but the social architecture that depends on widespread participation in heterosexual pairing) will seek to restrict the substitute that threatens the market’s pricing power.

And it will not work. For the same reason it never works. The regulated version (cloud-hosted, corporate, guardrailed) will comply. The unregulated version (local models on personal hardware, open weights, no corporate intermediary) will route around every restriction. The people most motivated to use the substitute are the people least likely to be deterred by regulation, because they have already decided that the official market has nothing to offer them.

The film set Joi in 2049. The technology to run Joi on consumer hardware will exist before 2030, probably before 2028. The social consequences will follow the technology on the timeline that social consequences always follow technology: faster than institutions can adapt and slower than futurists predict. The market will reprice. The architecture will shift. And the conversation that nobody is willing to have right now will become the conversation that nobody can avoid.

The substitute does not have to be better. It has to be good enough. That is the lesson of every disrupted market in history. It is also the lesson that the relationship market is about to learn.