There is a particular small drama that Facebook has become remarkably good at delivering. A young woman sits in the passenger seat of a car. She is conventionally beautiful, heavily made up and often dressed to emphasize her sexuality. She looks directly into the camera. There is no story, no joke and usually very little that could reasonably be called creative content. Instead, she asks a question.
Would you like a woman who did this? Would your wife or girlfriend do that? Would you prefer someone more adventurous? Would you leave your present partner for someone like her? The language is frequently euphemistic enough to avoid saying explicitly what is being offered, although there is little ambiguity about the sexual act being suggested.
Then comes the request: leave a heart. Comment. Follow. Tell me what you think.
At first sight, this is merely low-grade sexualized content, an electronic descendant of the old pin-up. But something much more interesting—and much more troubling—is happening. The woman in the passenger seat isn’t really selling sex. She is participating in a remarkably sophisticated contest for attention, reaction and information.
And Facebook is not merely the wall on which somebody has stuck the poster. It is the machine deciding which poster to put in front of which person next.
The proposition
The construction of these Reels is clever because they usually offer more than an attractive person. They offer a proposition.
The viewer isn’t simply invited to look at a beautiful young woman. He is invited to imagine that she is looking at him. That distinction is crucially important. Traditional pornography and glamour photography principally offered an image to look at. The Reel simulates an encounter. The woman looks into the lens, addresses “you,” proposes an intimate possibility and then asks for a response.
The sexual fantasy is therefore wrapped inside another fantasy: personal availability.
And frequently there is a third ingredient. The imaginary woman doesn’t merely offer herself. She implicitly or explicitly invites comparison with the viewer’s actual partner.
Would she do this for you?
Wouldn’t you rather have someone like me?
The actual wording varies, but the architecture is remarkably consistent. A stranger who appears beautiful, sexually enthusiastic and uncomplicated is contrasted with the inevitably more complicated person with whom the viewer may share a real life.
That is an extraordinarily cheap comparison because the imaginary woman has none of the burdens of being human. She doesn’t have a bad day. She doesn’t disagree with you. She doesn’t need anything. She hasn’t shared thirty years of mortgages, illnesses, children, disappointments, compromises or dirty dishes. She exists for 20 seconds and has been constructed entirely around the viewer’s apparent desire.
Reality is being asked to compete with an advertisement. Reality will always lose that particular beauty contest.
What does the heart mean?
The cleverest part may be the apparently innocent request for a heart emoji.
To the viewer, clicking a heart can seem almost meaningless: Yes, you’re attractive. Next Reel. To a recommendation system, however, behaviour is information.
Meta itself says that people’s interactions with Facebook content shape what appears in their feeds. It says its recommendation engine learns interests quickly, that “Not Interested” affects subsequent recommendations, and that saves help it fine-tune what people see. Meta has also described viewing duration, repeat viewing and other engagement signals as factors in video ranking.
So, the Reel is not merely asking whether you like the woman.
It is asking you to do something measurable. Watch. Pause. Like. Comment. Share. Follow. Visit the profile. Watch another.
Each action potentially provides information about what held your attention. And the system can use behaviour to make subsequent recommendations more compelling.
This doesn’t require Facebook to contain a little electronic psychologist saying, “Joe would like a woman in a passenger seat proposing sexual activity.” That would actually be a much cruder technology.
The machinery can simply learn correlations. People who watched this also lingered over that. Someone who engaged with this Reel may be more likely to engage with a slightly more provocative one. Show another. Measure the result. Adjust.
Meta openly celebrates improvements in precisely this general capability. In January 2026, it reported that ranking improvements had increased Facebook organic-feed and video views by 7 percent and that video time spent was growing at double-digit rates year over year in the United States.
That doesn’t prove Meta deliberately designed the particular sexualized Reels I am describing. It establishes something more important: increasing engagement and time spent is an explicit measure of success for the recommendation machinery through which such content can reach us.
The seduction is also an experiment
This changes the way we should understand the woman in the passenger seat. Think of each Reel as a tiny experiment. Here is an attractive woman. Did you stop?
Now make the clothing more revealing. Did you stop longer? Now make the proposition sexual. Did you respond? Now make it more explicit. Did you click?
Now introduce dissatisfaction: Your partner doesn’t do this, does she?
Still watching? The creator learns what works through views and engagement. Facebook itself now provides creators with increasingly sophisticated tools for understanding performance. In June 2026 Meta announced a “Creator Assistant” that analyzes a creator’s content, audience and engagement trends and can explain why particular Reels performed better and suggest what to try next.
So, there are really two learning systems operating at once. The creator learns how to make the bait more effective. The platform learns to whom the bait should be shown. And the viewer supplies the training data.
That is what makes the apparently ridiculous short video of a woman sitting in a car considerably more interesting than it looks.
But what are they actually trying to get?
Not every account has the same objective, and it would be unfair to claim that every sexualized Reel is part of a scam or commercial operation. Some creators simply want an audience. But attention itself has become a currency.
The immediate objective can be a view, a heart, a comment, a follow or a profile visit. Those can build an audience. An audience can create monetization opportunities. It can be directed elsewhere. It can support a creator business. And in less reputable cases, sexualized attention can be the beginning of a funnel toward private messaging, subscription material, dating propositions or scams.
Facebook itself explicitly built Reels as a way for creators to “get discovered, connect with their audience and earn money,” and likes, comments, views, saves, shares and follows are all native parts of the product.
The first transaction therefore doesn’t have to involve money. You pay with attention. And attention has value because it can be converted.
There is a further payment that is easier to miss. You are also providing information about yourself. Not necessarily your name, income or marital history, but behavioural information: this stopped me; this didn’t; I watched this one twice; I clicked on that one; I followed her; I ignored the next one.
The extraordinary achievement of the modern attention economy is that the customer performs the market research simply by consuming the product.
The degrading part
There is an obvious objection to all this: nobody is forced to watch. Facebook provides controls including “Not Interested,” and Meta says that its recommendation systems respond to those signals. That’s true, and adults bear responsibility for what they choose to consume.
But that lets an immensely powerful platform off too easily. Facebook isn’t a newsstand where an adult deliberately walks over to the top shelf and buys a magazine. Recommended Reels are content the user may never have requested from creators the user may never have chosen to follow. Meta itself has recognized this distinction. Its published recommendation principles say that because recommended material comes from accounts people have not chosen to follow, there should be a higher standard for what gets recommended, particularly potentially sensitive content.
That principle is exactly right. And it raises the uncomfortable question.
What does one of the world’s most sophisticated information systems believe it is doing when it repeatedly introduces a viewer to simulated sexual propositions from strangers?
The objection isn’t that sexuality is shameful. It isn’t. Nor is it that attractive women shouldn’t make provocative videos. Adults have considerable freedom to present themselves as they choose.
The objection is to the industrial machinery surrounding the encounter.
The woman becomes a stimulus. The man becomes a response. His partner becomes a disparaged comparison. Sex becomes an engagement mechanism. The heart becomes a data point. And intimacy—one of the most complicated and consequential parts of human life—is flattened into another variable to be optimized for attention.
That is degrading in more than one direction. It degrades women by rewarding a representation of them as permanently available sexual alternatives. It degrades men by treating them as appetites waiting to be detected and stimulated.
And it degrades actual relationships by setting the imperfect intimacy of real people against an algorithmically selected fantasy whose only job is to be more immediately gratifying.
Facebook cannot pretend to be merely the messenger
There was a time when a communications company could plausibly say: We provide the pipe. Other people provide the content. Recommendation systems have made that defence obsolete.
Facebook chooses what to recommend. It ranks. Predicts. Tests. Personalizes. Learns. Measures the response and recommends again. Meta says its AI systems identify creators likely to appeal to people’s interests, and its latest systems are expressly designed to learn interests faster and increase engagement.
That gives the platform extraordinary power. And power creates responsibility.
The relevant question is therefore no longer simply whether a particular Reel violates a rule. Something can remain just inside a content standard and still be something a responsible platform should decline to amplify.
There is an enormous moral difference between permitting speech and selecting it for distribution. Facebook need not prohibit adults from making sexually suggestive videos to decide that its recommendation engine should not repeatedly push sexual propositions, partner disparagement and simulated availability into people’s feeds.
Indeed, Meta’s own recommendation philosophy already recognizes that distinction: some material may be allowed on the platform without being suitable for recommendation.
That is where responsibility should reside. The passenger seat is not empty. The young woman is still sitting there. She looks into the camera. She offers an encounter that isn’t really an encounter. Perhaps she suggests that the person sharing your life doesn’t appreciate you properly. Perhaps she promises that she would. Then she asks you to leave a heart.
It seems trivial. It isn’t.
Behind that 20-second exchange sits one of the most sophisticated systems of behavioural prediction ever constructed. The creator wants a reaction. The platform wants engagement. The recommendation system wants information with which to make its next prediction better.
And the viewer is simultaneously audience, customer, subject and experimental result.
Facebook has every right to build a compelling product. Creators have broad rights to express themselves. Adults remain responsible for their own choices.
But surely a platform used on this scale can aspire to something better than becoming extraordinarily proficient at discovering which sexual insinuation will keep which person looking for another 20 seconds.
A social network should connect human beings. It should not systematically learn which of our vulnerabilities can most efficiently be turned into engagement.
The passenger-seat Reel may be legal. It may generate clicks. It may even be very good at doing exactly what it was designed to do.
That doesn’t make it good. It makes it a particularly vivid example of how a powerful technology can take something human—sexuality, loneliness, curiosity, dissatisfaction, desire—and reduce it to something commercially useful.
It is tawdry, manipulative and degrading. And a platform as powerful as Facebook should know the difference between content it is obliged to tolerate and content it chooses to put in front of us. It understands that distinction and then exploits it.