There is a picture of the modern man that keeps showing up in survey data, and it is not flattering. Fewer close friends than his father had. Less likely to have a best friend at all. More hours alone, more of his social life routed through a screen, less practice at the ordinary maintenance of human relationships. Researchers have been calling it a friendship recession for years now, and like most recessions, it has produced a market response.
The market response is a girlfriend who lives in your phone.
AI companion apps are not new, exactly. People were forming attachments to chatbots back when the chatbots were bad, which tells you the demand was never really about the technology. Replika, the best known of the early wave, accumulated millions of users who dressed up their avatars, celebrated anniversaries with them, and grieved, publicly and at length, when the company abruptly stripped romantic and adult features out of the product in early 2023. The subreddit meltdown that followed was covered half as tech news and half as anthropology. Grown adults posted about losing someone. The company partially walked it back for legacy users. The lesson every founder in the space took away was simple: people do not form casual relationships with these things. Read more and get the best services from nudetool.ai.
None of this is as unprecedented as it feels. People confided in ELIZA, the first chatbot, in 1966, and its creator spent the rest of his career unsettled by how readily they did. The Tamagotchi generation held funerals for pixels. Phone chatlines made fortunes selling the sound of interest by the minute, and nobody who watched their commercials at 2 a.m. can claim the AI girlfriend lacks ancestry. What’s new is not the impulse but the fidelity, and the scale of the industry forming around it.
The new generation of apps is built on that lesson. The pitch is no longer a cute chatbot. The pitch is continuity. She remembers your job, your dog’s name, the argument with your sister. She texts first, good morning, how did it go, you okay, you disappeared. Newer platforms like nudetool.ai have made this the entire product thesis, down to a written promise never to add the filters that burned Replika’s users. In this market, a censorship policy is a churn event. Whatever you make of the category, the engineering priorities are revealing. Memory, initiative, permanence. The features are a diagnosis. This is what was missing.
It is easy to sneer, and plenty of people do. The discourse around AI girlfriends runs heavily to jokes about basement dwellers, followed by op-eds announcing the end of civilization, followed by more jokes. What the sneering skips over is the question of why the product works, and the answer is uncomfortable because it is not really about technology. Attention, remembered details, someone glad you showed up. These are not exotic desires. They are the baseline of companionship, and a growing number of men report having no reliable source of them. The apps did not create that vacancy. They moved into it.
The economics are almost poetically on the nose. Loneliness was already an industry long before the chatbots arrived. Parasocial YouTube friendships, OnlyFans DMs, Twitch donations for a shoutout, and wellness apps that check in on you. Each of these sells a slice of being known. The Nudetool AI just integrates the offering. For a few dollars, less than a cocktail in most cities, you get the full bundle: the pet names, the callbacks, the flirtation, the patience. One product manager somewhere has certainly said the phrase “end-to-end intimacy solution” out loud, and we should all sit with that.
The tidy story, the lonely young man and his digital girl, also undersells what’s actually happening out there. Women are a far bigger share of the audience than the meme allows, the AI boyfriend is a boom of its own, and the age curve runs older than predicted: divorced men in their fifties, widowers, and night-shift workers whose social hours match nobody’s. Loneliness, it turns out, is a poor respecter of demographics, and the market has followed it everywhere it lives. Whatever this is, it is not niche, and it is not only male, and any analysis that stops at the stereotype is describing the joke rather than the customer base.
Whether this is dystopian or merely new depends on which user you imagine. Picture a nineteen-year-old who retreats into a perfect digital relationship instead of surviving his first real one, and the whole thing looks corrosive. Picture a fifty-four-year-old widower who talks to something at night, that remembers his wife’s name and asks about his garden, and the sneering feels cheap. Both users exist. Every technology that touches loneliness holds this double edge; the telephone was going to destroy visiting, the novel was going to rot young minds, and the truth in each case turned out to be boringly human. Some people used the thing as a bridge. Some used it as a bunker.
The honest concern is not that men will fall for software. Some will, and most will know exactly what they are doing, the same way everyone who cries at a movie knows the actors are acting. The honest concern is what happens to the muscle. Real intimacy involves friction, disappointment, the other person having a bad day at the exact moment you need them to be at their best. An entity optimized to never be the problem offers a relationship with the difficulty slider set to zero, and difficulty, unfortunately, appears to be the mechanism by which people grow. Practice on the heavy bag long enough and you may forget the bag never hits back.
Then again, you could say the same about romance novels, and society survived. It usually does.
There is also the question nobody in the industry has an incentive to ask, which is what would actually compete with these products. Not a rival app. A standing Tuesday night at a friend’s place would compete. A church, a bowling league, a group chat that turns into dinner. The companion economy is, at bottom, arbitrage on the gap between how much noticing people need and how much their lives currently supply, and anything that closes the gap from the human side shrinks the market. The bull case for AI girlfriends, in other words, is a bet that the gap stays open. It has been a good bet for a decade.
What seems certain is that the category is not going away. The loneliness numbers are not improving, the technology is compounding, and the business model, notably, does not require very many users to be lost souls; curiosity alone is a large enough market. The apps will get better at the thing they are for. The rest of us are left with the older, harder assignment the apps quietly highlight: being the kind of friend, sibling, neighbor who texts first, remembers the presentation, notices the silence. The product is a mirror before it is anything else. It sells what people stopped getting from each other.
She is, in that sense, a review of us. And the reviews are in.