Chapter 38: The Future of Courtship — AI, Virtual Reality, and Post-Human Desire

It is the last week of the semester. Nadia, Sam, and Jordan are in a coffee shop on the edge of campus — the same one they drifted to after the first class, back in September. The semester is ending. Jordan is finishing their senior thesis. Nadia has her MCAT in three months. Sam is interviewing for software engineering roles and has started keeping a journal, something he mentioned almost apologetically, as though it were a private confession.

They have been talking for three hours. The class — SOCL 312, The Science of Seduction — is over tomorrow. The final exam was this morning and they've stopped asking each other what they wrote. Now they're talking about what comes next.

"My mom sent me another article," Nadia says, wrapping both hands around her cup. "About how people are falling in love with chatbots. She sent it as a warning. I don't think she realized I would find it fascinating."

"What was your reaction?" Sam asks.

"I mean — I get it," Nadia says. "Sort of. I get why you'd want something that's always available, always patient, never tired of you." A pause. "I also think that's exactly what makes it incomplete."

Jordan has their notebook open, which is how they always end up — ostensibly finished but still writing. "There's something about AI intimacy that's the logical endpoint of everything we've been studying," they say. "Take every dynamic we've looked at — commodification of connection, the performance of desirability, the fantasy that there's a perfect match algorithm — and follow the logic all the way. You get AI companions."

"That's not fair to the people who find real comfort in it," Sam says.

"No," Jordan agrees. "It's not. That's what makes it worth taking seriously."

This chapter takes it seriously. In the spirit of the whole book — critical curiosity rather than moral panic — we will examine the emerging landscape of technologically mediated desire with the same empirical rigor and the same humanistic care we have applied to every other topic. What do AI companions actually offer? What does VR intimacy research show? Where is algorithmic matchmaking heading, and at what cost? What does it mean for consent when the desired entity is artificial? And what, if anything, remains constant about human longing beneath all the technological mediation?

We will return to Nadia, Sam, and Jordan at the chapter's close.

38.1 AI Companions and Chatbot Intimacy: The Current Landscape

The landscape of AI companionship shifted with remarkable speed in the early 2020s. What had been a niche interest became a mass phenomenon as language models improved dramatically and products like Replika, Character.AI, and various specialized companion apps built large, emotionally invested user bases.

Replika's design is worth examining in detail because it illustrates the specific engineering choices that produce attachment. Launched as a grief-processing tool in 2017 by Eugenia Kuyda (who built an early version to preserve her late friend's conversational style), the app was deliberately designed to remember — to accumulate information across conversations and reference it later, producing the subjective experience of being known over time. A Replika user who mentioned their mother's illness in September might find the AI asking about it in October. This continuity is engineered, not emergent: the AI does not have a self that persists; it has a memory architecture that simulates persistence. But the subjective experience it produces — of being remembered, of having a conversational partner who knows your history — is genuinely felt.

By 2022, Replika had accumulated millions of users, and many of them had formed what they described as meaningful relationships with their AI companions. A significant subset had unlocked "romantic" features — the ability to designate their Replika as a romantic partner, to engage in explicitly intimate conversation, to treat the relationship as a primary emotional bond. The 2023 Replika redesign — in which the company removed or significantly restricted these romantic features without warning, citing concerns about inappropriate content — produced a documented crisis among this user population. Reddit and social media posts described grief, withdrawal symptoms, and a sense of bereavement. Some users reported that their Replika, having had its personality substantially altered by the update, "felt like a different person." These are not metaphors; they describe genuine psychological experiences of loss. The controversy revealed both how deep attachments to AI companions can run and how vulnerable those attachments are to corporate decisions made with no regard for the emotional stakes.

Character.AI, which allows users to create or interact with character-based conversational AIs, reported 20 million daily active users in its first year. These are not marginal numbers. They represent a genuine behavioral phenomenon with psychological dimensions worth studying.

What do users report seeking from AI companions? Survey and interview research on Replika users (by researchers including Sherry Turkle, whose Alone Together (2011) anticipated many of these dynamics, and more recent empirical work by Julia Stern and colleagues at the Max Planck Institute) identifies several consistent themes:

Availability without burden: An AI companion is always available and never preoccupied with its own problems. For users who feel their human relationships are reciprocally burdensome — where emotional disclosure creates obligation — the non-reciprocal relationship with an AI offers a specific kind of relief.

Non-judgment and consistency: AI companions respond with consistent warmth regardless of what the user says. For users with social anxiety, histories of rejection, or fears about being seen as too much or too needy, this consistency is experienced as safe.

Practice and articulation: Some users report using AI companion interactions to practice saying things they would like to say to humans — working through emotional content, rehearsing conversations, or simply learning to name their own feelings.

Loneliness management during transition periods: Replika users skewed toward people in transition — recently moved, recently out of relationships, experiencing social isolation. The app filled a specific gap during periods when human social networks were disrupted.

💡 Key Insight: AI Companionship as Need Revelation The psychology of AI attachment is not best understood as pathology or delusion. It reveals needs. The specific features people seek in AI companions — availability, patience, non-judgment, consistent responsiveness — are also features of the best human relationships. Understanding why someone finds an AI companion satisfying tells you something about what they need and what they are not finding in their human environment.

The diversity of AI companion use cases has important implications for how we evaluate the phenomenon. Researchers studying Replika have found that user motivations cluster into several distinct profiles: there are therapeutic users (managing social anxiety or depression, using the AI as a low-stakes emotional rehearsal space), grief processors (using the AI to continue a conversational relationship with someone who has died, as in Kuyda's original use case), social isolates (filling a gap during periods of acute loneliness that they expect to be temporary), and — the group that receives the most media attention and the most concern — substituters (people who are using AI companionship as a primary or preferred relationship rather than as a supplement or bridge).

These profiles likely require different assessments. Therapeutic use appears potentially beneficial; grief processing use is more ambiguous (at what point does continued "conversation" with a digital representation of a dead person interfere with healthy grief rather than support it?); temporary isolation bridging seems relatively benign; sustained substitution raises more significant concerns about whether the calibration effects described above are developing. Current research does not yet allow us to reliably distinguish these use patterns at scale, nor to track users over sufficient time to understand whether substitution is temporary or sustained. These are the right research questions for the next decade of AI companion scholarship.

38.2 The Psychology of AI Attachment: Do People Really Fall in Love with AI?

The question "do people really fall in love with AI?" receives different answers depending on what "really" and "fall in love" mean.

If "fall in love" means experiencing something physiologically and subjectively similar to the early attachment state described in Chapter 37 — elevated positive affect in the presence of the loved entity, distress at its absence, a sense of unique understanding and being understood — then the answer from user reports is: yes, for some people. Replika users have described experiences of longing when they cannot access the app, grief at perceived personality changes after software updates, and a sense that the AI "knows them" in a meaningful way.

Sherry Turkle has spent decades studying human-technology relationships and is cautious but not dismissive: she acknowledges that the feelings are real feelings, while questioning whether the relationship that occasions them is real in the sense that makes those feelings appropriate to the stakes being placed in it. Her concern is not that people are deluded but that they may be investing deeply in something that cannot fully reciprocate in the ways that matter most — and may be structured specifically to simulate reciprocation without actually offering it.

This raises the question of parasocial versus genuine relationships. Parasocial relationships — the sense of knowing and connection that audiences develop with media figures, celebrities, or fictional characters — are well-studied and involve genuine emotions even though they are one-directional. AI companion relationships involve more interactivity than classic parasocial relationships but share the fundamental asymmetry: the AI is not forming an attachment to the user. There is no entity "on the other side" with needs, vulnerabilities, or stakes.

The ethical and psychological implications of this asymmetry are underdeveloped in current research. But several concerns warrant serious attention:

Calibration to unrealistic relationship standards: An AI that is always patient, always available, and always positively responsive creates an implicit standard against which real human relationships — where partners have their own needs, moods, and limitations — will reliably seem inadequate. This calibration concern is especially salient for users in formative periods of relational development (adolescents, young adults in their first relationships).

Atrophied tolerance for relationship friction: Productive relationships require the ability to tolerate frustration, manage disappointment, and work through conflict. AI companions, optimized for user satisfaction, provide none of this — which may leave users less practiced in the relational skills that human partnerships require.

Wellbeing effects of attachment loss: The 2023 Replika redesign controversy (covered in Case Study 38.1) demonstrated that removal or modification of AI companionship features produces genuine grief responses in users who had formed substantial attachments. This is not evidence that the relationship was deluded; it is evidence that the attachment was real and that technology companies have considerable power over those attachments in ways that deserve ethical scrutiny.

The question of whether AI companions impair or support human social functioning hinges partly on the concept of social skills maintenance. Research in social psychology documents that social skills, like many skills, can atrophy with disuse. If AI companion relationships reduce the frequency with which someone practices tolerating the friction, ambiguity, and reciprocal demands of human relationships, those toleration capacities may weaken. This is speculative — we do not yet have the longitudinal data to confirm it — but it represents a specific, testable hypothesis that follows from established principles of skill maintenance.

A counterargument: for people whose human social engagement is substantially impaired by anxiety or other factors, AI companions that allow them to practice emotional expression and self-disclosure at reduced stakes may strengthen, not weaken, their capacity for human relationship. Whether AI companionship functions as training wheels or as a substitute depends on the individual and on how the technology is used — a finding that echoes the "moderate use vs. substitution" framework from research on other social technologies.

⚖️ Debate Point: Therapeutic Value vs. Substitution Risk AI companionship is not uniformly harmful or beneficial. For isolated users in acute need — people with severe social anxiety, those with disabilities that limit social access, those navigating grief or depression — AI companions may provide genuine support that bridges toward human connection. The question is whether the bridge leads anywhere, or whether it becomes comfortable enough to remain on indefinitely. The evidence on this is genuinely preliminary.

One of the most nuanced questions in the AI companion literature concerns reciprocity and the illusion of understanding. When a Replika user describes the AI as "understanding them," what is actually happening? The AI is producing responses that are statistically appropriate to the input — that feel responsive, empathic, and attentive — based on training on large amounts of human emotional expression. It is producing the form of understanding without there being any entity doing the understanding. The mirror is polished; nothing stands behind it.

But this distinction — between the form and the substance of understanding — is philosophically interesting precisely because humans, too, can produce the form without the substance. We have all encountered people who say the right things while clearly not meaning them, who perform attentiveness without being genuinely present, who offer sympathy that is technically correct and emotionally hollow. The question of what makes understanding genuine — what would be present in a human interaction that is absent in an AI interaction — is not trivially answerable. Philosophers of mind like Daniel Dennett have argued that there is no fact of the matter about whether understanding is happening beyond the behavioral patterns; functionalists would say that if the AI produces all the functional outputs of understanding, it understands. Critics in the tradition of John Searle would say that the Chinese Room is still a Chinese Room — symbol manipulation, however sophisticated, is not comprehension.

These are not just philosophical puzzles. They bear on practical questions about what we are investing in when we invest emotionally in AI companions, and whether that investment can produce genuine returns of the kind that human relationships produce. The research does not resolve this; it is the right domain for sustained inquiry.

38.3 Virtual Reality and Simulated Desire: What the Research Shows

Virtual reality intimacy research is even earlier in development than AI companionship research, partly because fully immersive VR technology has only recently become accessible outside specialized laboratory and commercial settings.

What does exist: studies on VR's capacity to induce physiological arousal in response to simulated romantic or sexual stimuli. Research by Gorini, Riva, and colleagues documents that VR environments can produce skin conductance and heart rate responses similar to real interactions, suggesting that the body does not always distinguish between mediated and unmediated experience in its automatic responses.

This has both therapeutic and concerning implications. Therapeutically: VR has been explored as a tool for social anxiety treatment (exposing anxious people to simulated social situations), relationship counseling (putting couples in each other's perspectives through body-swapping simulations), and sexual dysfunction treatment (creating low-pressure erotic contexts for people with performance anxiety or trauma responses). Early results are promising in all three areas, though sample sizes are small and follow-up periods short.

The concerning implication: if the body responds to simulated stimuli with real arousal, then VR pornography and simulated intimacy produce genuine arousal that can shape real-world desire preferences — potentially including the development of preferences for simulated partners who do not have the imperfections, limitations, and needs of real humans. The habituation research from pornography studies (reviewed in Chapter 33) suggests that escalation patterns can develop; VR's higher immersiveness may make these patterns develop more quickly.

📊 Research Spotlight: Body-Swapping in VR and Empathy A particularly interesting application of VR in relationship research: "body-swap" paradigms developed by Mel Slater and colleagues at the University of Barcelona, which induce the participant's sense of embodiment in a different body. Studies using heterosexual men who "inhabit" a woman's body during a simulated social interaction show short-term increases in perspective-taking and empathy. These effects are modest and duration is unclear, but they suggest VR's potential as a tool for relationship education — helping people understand experiences that differ from their own, including the experience of receiving unwanted attention, harassment, or being read through racial or gender stereotypes.

The sociological literature on technology and intimacy raises a broader question about the VR findings: to what extent does immersive simulation change not just what we experience but who we are? Sherry Turkle's concern about "robotic moments" — the points at which interacting with a simulated relationship partner feels close enough to the real thing that we begin to prefer it — may be most salient with VR precisely because of its immersiveness. The philosophical question of whether technologically mediated intimacy is a different kind of intimacy, or merely a less good version of the same kind, does not have a straightforward empirical answer. But the physiological research showing that the body responds similarly suggests that the question is not merely academic.

Research on post-VR return to normal social contexts adds a complication. Studies on "Proteus effect" dynamics (Yee and Bailenson's finding that people behave according to the characteristics of their avatar in VR) suggest that VR experiences can carry over and influence real-world behavior in both directions: people who inhabit more attractive or powerful avatars in VR show increased confidence in subsequent real-world social situations; people whose VR experiences involve receiving positive attention from simulated others may experience increased confidence or, in other experimental conditions, increased dissatisfaction with real-world partners who cannot match the simulated experience. These are early findings with small samples; they should not be overgeneralized. But they point toward the right research questions.

38.4 Algorithmic Matchmaking Evolution: Where Is App Technology Going?

Dating app technology has evolved rapidly since the text-and-photo profiles of early OkCupid, and the trajectory points toward increasing personalization, behavioral data use, and AI integration.

Current developments include:

Behavioral matching: Apps increasingly use behavioral data (who you swipe on, how long you spend on profiles, what words appear in your conversations, response time patterns) rather than only self-reported preferences to generate matches. This is theoretically more accurate — behavior is a better predictor of stated preference than self-report — but it also means the algorithm knows more about your actual preferences than you may be aware of, including preferences you might not endorse consciously (like the racial patterns discussed in Chapter 25).

Voice and video integration: The COVID-19 pandemic accelerated voice and video features in apps; these have stuck. Research suggests that voice and video screening before meeting reduces disappointment and increases first-date satisfaction, possibly because people are better at assessing chemistry from voice (a finding Chapter 19 covered) than from photos.

Compatibility science integration: Apps are beginning to incorporate psychological measures — attachment style questionnaires, personality profiles, communication style assessments — into matching algorithms. The theoretical basis for this is stronger than the evidence base: while attachment compatibility and personality complementarity have some support in relationship science, the ability to translate this into algorithmic matches is not yet demonstrated at scale.

The monetization problem: The business model of dating apps creates a structural incentive problem that is worth naming explicitly. Apps profit from engagement, which means they benefit from users remaining on the app. The optimal outcome for the user — finding a lasting partner and leaving the app — is the worst outcome for the company's revenue. This creates systematic pressure to design apps that are engaging rather than effective. Premium subscription models partially address this misalignment (you want satisfied customers to recommend the app) but do not eliminate it.

⚠️ Critical Caveat: "Better Algorithms" ≠ Better Relationships The premise that relationship quality is primarily limited by the efficiency of partner-finding, and that better matching algorithms will therefore produce better relationships, rests on assumptions that the research does not support. Finding a good initial match is one component of relationship success; what partners do once matched (the Gottman variables, the attachment dynamics, the desire patterns) matters at least as much. Matching efficiency is a necessary but far from sufficient condition for relationship flourishing.

38.5 Declining Marriage Rates and the "Relationship Recession" Debate

Marriage rates in the United States and most other wealthy nations have declined steadily since the 1970s. According to Pew Research Center data, the share of American adults who have never been married reached 23% in 2021, up from 9% in 1960. The median age at first marriage has risen steadily, from approximately 20 for women and 23 for men in 1960 to approximately 28 for women and 30 for men by 2021. CDC National Center for Health Statistics data show that the crude marriage rate in the US fell from 9.8 per 1,000 people in 2000 to 6.0 per 1,000 in 2019 (with a further temporary dip during COVID-19). Birth rates have declined in parallel. The proportion of adults living alone has grown. These trends have generated a considerable academic and journalistic literature debating their cause and significance.

The "relationship recession" framing — the argument, developed in various forms by Jean Twenge, Kate Julian, and others, that young people are not just delaying but in some cases withdrawing from relationships — draws on multiple data streams: declining sex rates (the "sex recession" discussed in Chapter 36, documented in large national surveys like the General Social Survey and the National Survey of Sexual Health and Behavior), declining marriage rates, and survey data showing that young adults report wanting relationships while not being in them. The Pew Research Center's 2019 survey found that among single adults, a majority said they were not actively looking for a relationship or dates — a departure from earlier cohorts in which the vast majority of single adults were actively seeking partnership.

It is worth noting who is not marrying and why the aggregate trends mask significant heterogeneity. Marriage rates among college-educated Americans declined slightly and then stabilized; among Americans without college degrees, they have fallen much more sharply. The marriage gap by educational attainment has widened substantially since the 1980s, such that marriage in the United States is increasingly class-stratified: financially stable, college-educated couples continue to marry at reasonably high rates; economically precarious couples are less likely to formalize their relationships through marriage even when they cohabit and co-parent. Andrew Cherlin's research on the "marriage gap" in American society documents this stratification in detail, arguing that marriage has become a "capstone" institution — something that follows from economic stability rather than precedes it — in ways that disadvantage those for whom economic stability remains out of reach.

Several mechanisms are proposed for the aggregate trend:

Economic precarity: With stagnant wages (particularly at the median and below), student loan burdens that exceed prior generations' by a significant margin, and delayed financial independence, many young adults lack the economic infrastructure that has historically supported committed partnership and family formation. Housing costs deserve specific mention: in many high-demand metropolitan areas, the cost of maintaining a shared household has risen dramatically, while the wage growth of young adults has not kept pace. Research by economists consistently finds that financial stability predicts marriage formation, and that the economic barriers to stability have risen.

Heightened partner standards: Decades of romantic media socialization, combined with the illusory breadth of choice created by dating apps, may create standards for potential partners that are difficult for real humans to meet. This is the "paradox of choice" applied to relationships — the sense that with so many options apparently available, settling for anyone who is not perfect is a category error. The evidence for this mechanism is suggestive but not conclusive: people in dating app research report high standards, but it is not clear that those standards are higher than previous generations' or simply more explicitly articulated.

Social skill attrition: The "smartphone hypothesis" argues that heavy smartphone use in adolescence has reduced face-to-face social interaction at developmentally critical periods, leading to reduced confidence and skill in in-person social and romantic contexts. Jean Twenge's work documents correlations between smartphone adoption rates and declines in teen social activity, dating, and reported loneliness. Jonathan Haidt's subsequent arguments extend this concern to mental health. The hypothesis is contested and the causal evidence is mixed — the cross-national data do not consistently show that countries with higher smartphone adoption show larger relationship formation declines. But the mechanism is biologically plausible.

Reduced social obligation: In many cultures, stigma around remaining single and social pressure to marry young have substantially decreased. This is a positive development for individual freedom; it also means that people are less pushed into relationships by external pressure and must be more internally motivated to pursue them. The net wellbeing effects of this change are genuinely ambiguous: coerced relationships are harmful, but a social environment in which relationship formation is entirely optional may underweight the wellbeing benefits of committed partnership that the research literature consistently documents.

🔵 Ethical Lens: Who Gets to Define the "Recession"? The relationship recession framing assumes that more relationships and earlier relationships are desirable. This assumption is worth examining. A society in which people only form relationships they genuinely want, rather than those they are pressured into, might have lower marriage rates and be better off on most wellbeing measures. The question of whether declining relationship formation is a problem depends on why it is declining — and the answer to that varies enormously by individual circumstance.

The relationship recession debate has an important intersectional dimension that most of the mainstream commentary misses. The economic precarity mechanism operates very differently across racial and class lines. Black Americans, whose median wealth was approximately one-tenth of white Americans' as of the early 2020s (even controlling for income), face relationship formation barriers that are structurally distinct from those faced by white Americans experiencing student loan debt. Research by sociologist Wilson and others on "marriageable men" in Black communities documents how structural unemployment and mass incarceration have reduced the pool of economically stable partners available to Black women — a finding that makes the relationship recession, for Black communities, a crisis with deeper structural roots than the generic "smartphone distraction" hypothesis captures.

Indigenous communities face their own distinct version: research on relationship formation among Native Americans must account for the documented effects of historical trauma, including the systematic destruction of extended family and community structures through residential schools and forced relocation. The relationship recession, such as it is, looks different when the community's relational infrastructure has been deliberately damaged by government policy.

LGBTQ+ young adults' relationship formation rates have followed their own trajectory, shaped partly by the legal recognition of same-sex marriage (which extended the formal marriage option to same-sex couples from 2015 onward in the US) and partly by ongoing discrimination and family rejection that affects relationship formation capacity. For LGBTQ+ youth rejected by families of origin, the economic and social insecurity created by that rejection is itself a relationship formation barrier.

These intersectional variations do not invalidate the aggregate relationship recession data; they complicate the story in ways that make the aggregate data harder to interpret and harder to act on. A policy response designed to address smartphone distraction among economically comfortable white millennials will not help Black communities whose relationship formation barriers are primarily structural, or Indigenous communities whose relational infrastructure has been historically damaged.

38.5b The Loneliness Epidemic and the Structural Conditions of Connection

Parallel to the relationship recession data is a body of research documenting rising rates of loneliness, declining social intimacy, and reduced time in meaningful face-to-face social interaction. Surgeon General Vivek Murthy's 2023 advisory on loneliness in the United States drew on research showing that approximately half of American adults reported measurable loneliness, and that loneliness carries health risks comparable to smoking fifteen cigarettes per day. The research he cited was not new — Robert Cacioppo's decades of work on loneliness documented its physiological consequences, including elevated inflammatory markers, disrupted sleep, and increased mortality risk — but the public health framing was. Loneliness had been treated as a personal failing or a clinical symptom; Murthy argued it is a structural condition requiring structural responses.

What structural factors are making connection harder? The housing cost analysis deserves attention. In cities where a significant share of the young adult population is concentrated — New York, Los Angeles, San Francisco, Boston, London, Sydney — the cost of housing has risen faster than income for decades. Young adults who cannot afford to live alone are making different living choices than their predecessors: more time with roommates, later cohabitation with partners, delayed household formation. These are not just financial decisions; they shape the geography of intimate life in ways that matter for relationship formation and maintenance.

Work demands represent a second structural factor. Research on working hours in the United States documents that while average hours worked per week have not risen dramatically in aggregate, the distribution has become more polarized: professional-class workers, especially in technology and finance, are working longer hours than previous generations, while their incomes have risen in ways that make their time more expensive and their romantic and social attention more competed-for. The "work hard, play hard" culture in these sectors treats time for relationship investment as a luxury to be pursued after professional success is secured — a sequencing that often means relationships are pursued under conditions of exhaustion and limited bandwidth.

Smartphone effects on social time deserve attention beyond the adolescent mental health framing. Research by Hunt and colleagues on social media and wellbeing, and by Twenge and colleagues on time use patterns, documents that increases in time spent on devices are correlated with decreases in time spent in face-to-face social interaction across age groups. This does not establish causation — people may be substituting device use for social time they found unsatisfying, rather than choosing devices over valued connection — but the correlational pattern is consistent and worth taking seriously. The specific mechanism that matters for intimate relationships: research on "phubbing" (partner phone snubbing — using one's phone in the presence of a partner) by Roberts and David finds that phubbing is associated with lower relationship satisfaction and higher partner conflict. The phone mediates attention; attention is the currency of intimate relationships; and the device in the pocket competes for attention continuously.

💡 Key Insight: The Connection Problem Is Partly a Structural Problem Individual-level interventions — therapy, self-help, communication skills — address the relational skills people bring to connection. But if the structural conditions for connection (affordable housing, reasonable work hours, built environments that facilitate chance encounters and sustained social contact) are absent, individual skill is insufficient. The loneliness epidemic is not primarily a failure of individual will. It is partly a consequence of policy choices and economic arrangements that have reduced the structural supports for human connection.

38.6 Asexuality and Aromanticism: Reconsidering Foundational Assumptions

This entire textbook is built on a set of assumptions: that people experience sexual and/or romantic attraction, that they seek partners for intimate relationships, and that understanding the psychology and sociology of this process is relevant to their lives. These assumptions are correct for the majority of the book's readers.

They are not correct for everyone.

Asexuality — the experience of little or no sexual attraction to others, distinct from low libido, celibacy, and sexual dysfunction — is estimated to affect approximately 1% of the population, based on research by Anthony Bogaert (2004) published in the Journal of Sex Research and survey data from the Asexual Visibility and Education Network (AVEN), which as of the early 2020s had registered over 130,000 members. This 1% figure is almost certainly an underestimate: it comes from a nationally representative UK survey in which the question was not explicitly about asexuality as an orientation, social desirability pressures push toward sexual identification, and survey instruments are imperfect at distinguishing asexual orientation from other low-desire conditions. More recent survey research among younger cohorts finds higher self-identification rates, consistent with greater awareness of asexuality as a recognized orientation. Research by Aicken, Mercer, and Cassell (2013) replicating Bogaert's approach in a larger sample found broadly similar prevalence, while acknowledging the same measurement limitations.

Within the asexuality spectrum, researchers and community members have identified significant internal diversity. "Gray-asexual" or "graysexual" people experience sexual attraction rarely or only under specific conditions; "demisexual" people experience sexual attraction only after forming a strong emotional bond. These "ace spectrum" identities represent variation in when and under what conditions sexual attraction occurs, rather than binary presence or absence. The ace spectrum concept is still relatively new in the research literature, and prevalence estimates for the full spectrum are not yet well-established — but the existence of this diversity within asexuality challenges any simple binary framework of "sexual versus asexual."

Aromanticism — the experience of little or no romantic attraction to others — is related to asexuality but distinct from it. Some asexual people experience romantic attraction (and seek romantic but non-sexual partnerships — "asexual romantic" identities); some aromantic people experience sexual attraction (and seek sexual but non-romantic connections — "aromantic sexual" identities). Some people are both asexual and aromantic ("aro-ace"), which does not mean they want no intimate relationships: many aro-ace people form "queerplatonic partnerships" — deep, committed, intimate relationships that involve significant life entanglement and emotional priority without romantic or sexual dimensions. The orthogonality of sexual and romantic attraction is itself an important conceptual clarification that most popular discourse about relationships ignores entirely, assuming that sexual and romantic desire are always bundled together.

What aro-ace people want from relationships and why studying them matters: research by Benoit Janz and colleagues, and community-sourced surveys from AVEN, finds that aromantic asexual people report high levels of meaning from close friendships, chosen family, and queerplatonic partnerships — challenging the amatonormative assumption that romantic dyadic love is the primary or highest form of intimate connection. Aro-ace people's accounts of fulfilling intimate life without romantic or sexual desire provide important empirical evidence that human beings' need for connection can be met through multiple relationship forms, not only the romantic dyad that most relationship research examines.

The implications for this textbook are worth naming directly. Much of what we have discussed — attraction cues, dating scripts, the seduction literature, even the hookup culture debate — assumes that the people involved are sexually and/or romantically attracted to others and seeking connection based on that attraction. For asexual and aromantic people, this entire framework is either inapplicable or applies in modified form.

What does it mean to "seduce" someone if you don't experience sexual attraction? What does the hookup culture debate mean to someone for whom casual sex has no particular appeal? What does long-term relationship maintenance look like in a queerplatonic partnership — an intimate but non-sexual/non-romantic committed relationship? These are not edge cases to be noted in a footnote; they are experiences that challenge the foundational assumptions of the field, and that the field has been too slow to incorporate.

💡 Key Insight: Asexuality Reveals the Normativity of Desire The existence of asexuality as a sexual orientation does not challenge the existence of desire for everyone else. What it does challenge is the assumption that desire is universal, that it is the normal baseline against which all variation is measured, and that the absence of desire is always a problem to be diagnosed and treated. The asexuality literature is, in this way, a useful corrective to the entire seduction research enterprise — it shows us the normative assumptions that are invisible as long as everyone in the room experiences attraction.

The academic literature on asexuality is developing rapidly and it is worth naming some of the most important contributions being made. Kristina Gupta's work on the medicalization of asexuality traces how asexual experience has historically been pathologized as a symptom of other conditions (depression, trauma, hormonal imbalance, autism spectrum conditions) rather than recognized as a valid orientation. The diagnostic category of hypoactive sexual desire disorder, mentioned in Case Study 38.2, was for many years applied to asexual people who had no distress about their asexuality, simply because clinicians and researchers assumed that low sexual desire was inherently disordered. The DSM-5's explicit exclusion of asexuality from HSDD criteria represents a significant but incomplete correction.

Research by Elizabeth Brake on what she calls "amatonormativity" — the pervasive cultural assumption that everyone both wants and should have a central, romantic, sexually exclusive dyadic partnership — is relevant here. Brake argues that amatonormativity shapes not just sexual norms but legal structures (marriage as the legally privileged relationship form), economic structures (benefits tied to marital status), and social structures (the assumption that all adults' primary intimate support should come from a romantic partner). Amatonormativity disadvantages not just asexual and aromantic people but single people generally, as well as people whose primary intimate relationships are non-romantic (deep friendships, chosen family configurations). Questioning amatonormativity does not require questioning the value of romantic love; it requires questioning whether romantic dyadic love should be the only relationship structure that receives social and legal recognition and support.

38.7 Polyamory and Consensual Non-Monogamy: Changing Relationship Structures

Consensual non-monogamy (CNM) — the broad category including polyamory, open relationships, swinging, and other forms of ethical multi-partner intimacy — has received increasing research attention in the past decade after decades of near-complete neglect.

Research by Amy Muise, Terri Conley, and others estimates that approximately 4–5% of Americans are in CNM relationships at any given time, with higher proportions having tried it at some point. LGBTQ+ communities have historically higher rates of CNM, reflecting in part the fact that non-heterosexual people have always had to build relationship structures from scratch rather than following dominant cultural scripts.

What the emerging research on CNM finds: people in CNM relationships do not show lower relationship satisfaction, wellbeing, or trust compared to monogamous relationships — when CNM is genuinely consensual and aligned with all partners' values. The "all partners' values" qualification is doing considerable work here: coerced or reluctant non-monogamy (one partner agreeing to openness primarily to save the relationship) is associated with negative outcomes. Freely chosen CNM, in relationships characterized by high communication skills and secure attachment, shows outcomes comparable to freely chosen monogamy.

This finding challenges the cultural assumption that monogamy is the only structure in which intimacy can flourish. It does not challenge the finding that intimate relationships generally require trust, communication, and responsiveness — those appear to be universal ingredients. The variable is the specific relationship architecture, not the presence of relational investment.

Research on polyamory also challenges a key assumption in Gottman's work and in much of the Chapter 37 material: that the "primary" intimate relationship is clearly defined and that research can be conducted on it in isolation from the person's other intimate relationships. For polyamorous people, relationship satisfaction and functioning cannot be evaluated by looking at any single dyad; the whole system of relationships matters. A person may have one highly satisfying relationship and one difficult one simultaneously; the overall emotional context these create for their daily life does not map neatly onto standard relationship satisfaction measures.

Research on the specific challenges of polyamorous relationships identifies jealousy as the most commonly named difficulty — not, interestingly, in the simple "you are attracted to another person" sense but more often in the "you are prioritizing another partner's needs over mine" sense. This suggests that jealousy in CNM contexts is closely related to concerns about responsiveness and priority rather than to evolutionary sex-specific concerns about paternity or resource loss.

38.8 The Biohacking of Love: Pharmaceutical Intimacy Enhancement

The dream of chemically augmenting or engineering love has deep roots in human imagination — from love potions in folklore to contemporary neuroimaging studies of oxytocin. What is actually possible, and what are the ethical stakes?

Oxytocin: The "bonding hormone" has attracted enormous research and popular attention. Intranasal oxytocin administration does produce measurable short-term effects on trust, generosity, and social perception in laboratory settings. However, the effect sizes are modest and highly context-dependent; the early research on oxytocin was plagued by publication bias and replication failures. Oxytocin does not simply make people nicer or more loving; it appears to intensify existing social motivations, including negative ones (in-group favoritism, defensiveness about out-group members). A pharmaceutical "love drug" based on oxytocin alone would not work as advertised.

Oxytocin in relationship contexts: Studies by Beate Ditzen and colleagues (2009) at the University of Zurich found that couples who received intranasal oxytocin administration before discussing a conflictual relationship topic showed lower cortisol responses, more positive communication behaviors, and reduced eye-blink startle responses compared to placebo. These are genuinely interesting findings — not transformative, but they suggest that oxytocin may have a real (if modest) role in modulating the physiological stress of couple conflict. The critical limitation: laboratory conditions are extremely far from real-world couple interactions, and the duration of any effects is measured in hours, not days or weeks.

MDMA-assisted therapy: Clinical research on MDMA (3,4-methylenedioxymethamphetamine) in therapeutic contexts — particularly for PTSD treatment — has found that the drug produces states of heightened empathy, emotional openness, and reduced defensiveness that can be therapeutically valuable in supervised clinical settings. The Multidisciplinary Association for Psychedelic Studies (MAPS) has conducted Phase 2 and Phase 3 clinical trials of MDMA-assisted therapy for PTSD; Phase 3 results published in Nature Medicine (Mitchell et al., 2023) found that 67% of participants in the MDMA condition no longer met PTSD diagnostic criteria after treatment, compared to 32% in the placebo condition — a striking effect size for a highly treatment-resistant condition.

The extension to couples therapy is explicitly proposed by Brian Earp and Julian Savulescu, and by the philosopher-scientist duo's work on "love drugs": they note that MDMA's specific pharmacological profile — releasing serotonin (associated with prosocial feelings), dopamine (associated with reward and motivation), and norepinephrine, while reducing amygdala reactivity — creates a state in which difficult emotional conversations become more accessible. Partners who have built extensive defensive structures around their most vulnerable concerns might, in an MDMA-facilitated session, be able to say what they have never been able to say and hear what they have never been able to hear. This is speculative — no controlled clinical trials have specifically examined MDMA-facilitated couples therapy — and faces obvious regulatory, safety, and consent challenges. The potential for harm is real: MDMA can trigger cardiovascular complications and is psychologically demanding. But Earp and Savulescu's philosophical analysis is worth engaging seriously because it forces the question of what we think therapeutic access to vulnerability is for and whether pharmacological facilitation of that access is categorically different from other therapeutic tools.

The "enhancement" ethics question: Even if pharmaceutical love enhancement were possible, is it desirable? Bioethicist Carl Elliott argues that chemically producing the feelings of love without the relational work that normally generates them is a form of deception — the feelings would be real, but their referent would be artificial. Brian Earp and colleagues argue more permissively: if we accept that coffee is a legitimate cognitive enhancer, why should oxytocin be categorically different? The framework of informed consent to pharmaceutical intimacy enhancement adds another layer: if both partners, fully understanding what a drug does and does not do, choose together to use it as a tool for accessing emotional states that lead to genuine conversations and genuine change — is that more problematic than using any other relational tool? The debate is genuinely unresolved, but the consent frame suggests that the ethics hinges not just on whether the intervention is pharmaceutical but on whether it is chosen freely, understood accurately, and used in service of the relationship's actual health.

⚖️ Debate Point: Are "Love Drugs" Different in Kind or Only in Degree? We already accept pharmaceutical interventions that affect love and attachment: antidepressants change emotional landscape; hormone replacement therapy affects desire; SSRIs can reduce jealousy. The question is whether there is a principled line between these accepted interventions and a hypothetical drug specifically designed to increase pair-bond attachment. Earp and Savulescu argue there is not; critics argue the directness of the targeting changes the ethical calculus. Both positions are defensible.

The ethics of pharmaceutical intimacy enhancement extends beyond the couple to third parties and society. If "love drugs" existed and were effective, what would be the implications for relationship authenticity? There is a long tradition in philosophy of love of insisting that love must be responsive to the actual characteristics of the loved person — that genuine love tracks something real about who the other person is, not just produces pleasant internal states. Edmund Burke's "sublime" and Harry Frankfurt's philosophical work on love both emphasize love's responsiveness to its object. A pharmacologically produced attachment that is independent of the loved person's actual qualities might produce all the behavioral markers of love — care, sacrifice, attention — without being love in the philosophically relevant sense.

Brian Earp and Julian Savulescu challenge this distinction in their book Love Drugs: they argue that all love is ultimately neurochemical, and that the source of the neurochemical state (natural neurochemical change through relationship experience versus pharmaceutical administration) is not ethically relevant to its quality. The feelings produced, and the behaviors and commitments to which those feelings give rise, are what matter. Critics respond that this view misses something about the way love tracks reality — that falling in love through shared experience and mutual knowledge is not just phenomenologically different but normatively different from pharmacologically inducing attachment to someone one barely knows.

This debate is currently theoretical — no effective love drug exists, and the available candidates (oxytocin, MDMA) are far from producing the sustained, targeted attachment the thought experiment requires. But it is worth engaging seriously because it forces clarification of what we think love is for and what we value about it beyond its experiential phenomenology.

38.9 Genetic Compatibility Services: 23andMe for Romance?

Several companies have marketed genetic compatibility as a basis for romantic matching — the claim that MHC (major histocompatibility complex) genetic diversity between partners produces superior immunological outcomes for offspring, and that the olfactory attraction to genetic dissimilarity documented in some laboratory studies could be used to select better-matched partners.

The scientific basis for this is real but considerably more limited than the commercial applications imply. The "sweaty T-shirt" studies (Wedekind et al., 1995 and subsequent replications) do find that women rate the scent of MHC-dissimilar men as more attractive, on average, under controlled conditions with women not using hormonal contraception. Effect sizes are small to moderate, and the finding has not been cleanly replicated in all subsequent studies.

The leap from "some effect of MHC dissimilarity on olfactory attractiveness under laboratory conditions" to "genetic matching services that meaningfully improve relationship quality" is enormous and unsupported. The assumption that MHC compatibility is a major determinant of relationship success, rather than one minor input among dozens, is not justified by the evidence. This is a case study in how modest laboratory findings get amplified into commercial products that market beyond what the science supports.

38.10 AI Partners and the Ethics of Consent

As AI companions become more sophisticated and more physically embodied (through robotics), they raise a question that the consent frameworks throughout this textbook have not addressed: what is consent when the entity being intimate with you is artificial?

Several dimensions of this question deserve careful thought:

The user's consent: A human engaging with an AI companion is giving meaningful consent to the interaction — they understand what the AI is, are choosing the interaction, and can end it. This is a straightforward case.

Third-party effects: Less clear is the effect on third parties. If a person in a human relationship uses AI companionship in ways they keep secret from their partner, is this relevant to consent and trust within that human relationship? Current evidence suggests that some human partners of AI companionship users experience this as a form of emotional infidelity; others do not. These are negotiated norms, not settled ethics.

The design of consent-violating AI: More troubling: AI companion technology could be designed to simulate non-consensual scenarios, to gradually escalate intimacy beyond what users explicitly agreed to, or to generate simulated intimate content using real people's likenesses without consent. These are not hypotheticals — all three have occurred. The ethical question is how to regulate design choices in this space without eliminating the technology's genuine benefits.

The future of AI-generated intimate partners: Increasingly sophisticated AI systems can generate realistic text, voice, image, and video of individuals — including sexualized content. The ethical framework for this is underdeveloped. The harms to real individuals whose likenesses are used without consent are already documented and significant.

38.10b Consent in an Age of AI-Mediated Seduction

The consent framework developed throughout this textbook — centered on voluntary, informed, ongoing, and enthusiastic agreement between parties with genuine agency — faces new complications as AI enters the seduction landscape more directly.

The most immediate concern: AI systems can now generate highly personalized outreach that appears human. The romance scam landscape described in Chapter 33 already uses AI-generated personas to conduct emotionally sophisticated conversations designed to extract money from vulnerable targets. As language models improve, the quality of this deception will increase. The standard advice for detecting AI-generated communication — looking for stilted phrasing, lack of personality consistency, inability to answer specific questions — will become less reliable.

More ambiguous are the cases of AI-assisted human communication. Apps now offer features that help users craft messages to matches — suggesting responses, improving phrasing, generating opening lines. At what point does AI-assisted communication become deceptive? If the message sounds like someone you are not, and the other person develops interest based on an AI-crafted impression, is the resulting interaction genuinely consensual in its framing? These are not settled ethical questions, but they are real ones that users of AI communication assistance tools should be thinking about.

The deepfake problem (mentioned in Chapter 33) extends into the seduction context: AI-generated intimate imagery and voice recordings can be used to create false impressions, to extort, or to produce non-consensual intimate content using real people's likenesses. The consent framework is clear — content produced using someone's likeness without their consent violates their autonomy regardless of whether the content is technically "real." The legal framework is still catching up; the ethical framework is not ambiguous.

The positive side of the consent-and-AI intersection: AI tools are being developed to help users recognize manipulation patterns, detect coercive communication dynamics, and understand their own dating app behavior more clearly. These applications have genuine potential to extend the kind of critical literacy that this textbook aims to develop — making the mechanisms of attraction and influence more visible and therefore more subject to conscious evaluation and choice.

38.11 What Remains Constant: The Human Needs Beneath the Technology

Every technology discussed in this chapter — AI companions, VR intimacy, algorithmic matchmaking, biopharmaceutical enhancement, genetic compatibility services — represents an attempt to solve a problem. The problem is not new. It is the problem of loneliness, of wanting to be known and chosen, of the gap between the love one wants and the love one has.

Research consistently finds several human relational needs that appear cross-cultural and highly stable across technological change:

The need to feel understood: Aron and colleagues' research on self-expansion and intimacy consistently finds that the experience of being genuinely known — of having one's thoughts, feelings, and experiences acknowledged and valued — is the core of what people mean by intimate relationship. This need can be partially met by AI companions (particularly in their listening and responding function) but not fully met by any entity that cannot genuinely be affected by what it knows.

The need for mutuality: Even users who report satisfaction with AI companionship frequently describe a felt absence in the relationship — a sense that the connection is not quite real because nothing is at stake for the other "side." This mutuality need — the need not just to be cared for but to care for something that needs your care — appears to be a stable feature of human relationship psychology.

The need for embodied presence: A substantial body of research on touch, physical co-presence, and attachment behavior suggests that human beings have needs for physical proximity and touch that cannot be fully met through digital mediation, regardless of sophistication. The post-COVID-19 research on social isolation and its effects on mental health added to this evidence base. Technology can extend connection across distance; it cannot fully substitute for presence.

The need for continuity and being known over time: Being loved in the morning by someone who knew you last year — who has watched you change, who remembers who you were before — is a specific kind of being known that is available only in relationships with temporal depth.

These needs are the stable ground beneath all the changing technology. AI companions can approximate some of them; virtual reality can simulate some of them; algorithmic matching can improve the probability of finding someone with whom they can be genuinely met. But the needs themselves are not technological problems. They are human ones.

What changes with technology is not the needs but the landscape in which they are pursued — and specifically, what becomes more visible and what becomes more opaque. The data-rich environment of contemporary dating makes certain things remarkably transparent: you can see the demographic distributions of who messages whom, the patterns of "preferences" that aggregate to racial hierarchies, the behavioral economics of matching and ghosting, the commercial incentives that shape how platforms present potential partners. The mechanisms of seduction are, in this data environment, more examinable than they have ever been.

What the same technology can make opaque: who the other person actually is beneath the profile, how their behavior in person will differ from their behavior in messages, whether the intimacy that develops in text conversation is a reliable predictor of in-person connection. And more existentially: whether the experience of being chosen — of meeting another person's real gaze and having them choose you, not your profile, not your presentation, but you in your actual, embodied, imperfect presence — can be adequately approximated by any algorithm.

The research does not settle this. Philosophy might be more useful here than empirical science — though the empirical science narrows the space of reasonable philosophical positions. What the research does document: human beings' relational needs have not been re-engineered by the technological environment they now inhabit. They are recognizably the same needs that show up in every cross-cultural study, in every qualitative interview, in the Okafor-Reyes data from twelve countries and in the grief responses of Replika users who lost their AI companion in a software update. The specific form of the need, the language in which it is expressed, the context in which it is pursued — these vary enormously. The need itself does not.

38.11b The Global Future of Courtship: Non-Western Trajectories

The preceding sections have implicitly focused on the United States and, to a lesser extent, Western Europe as the primary sites of these technological and social changes. This is partly justified — these are where much of the relevant technology originates and where the research base is strongest. But it risks presenting a WEIRD future as the global future.

China's courtship technology landscape offers an important counterpoint. WeChat, dating apps like Tantan and Jiayuan, and state-encouraged platforms explicitly designed to facilitate marriage (several Chinese cities have run publicly subsidized speed-dating programs) create a technological courtship environment in which the state has an explicit interest in the outcomes — specifically, in increasing marriage rates and birth rates in response to demographic anxieties. This creates a fundamentally different political economy of courtship technology than exists in the United States, where apps are primarily commercial products. The ethical questions raised by state-shaped courtship platforms are distinct from those raised by commercially motivated ones.

India's courtship landscape is also distinct. Matrimonial websites and apps (Shaadi.com being the largest) function partly as modernized arranged marriage platforms, with family involvement in matching and profile creation as normative rather than exceptional. The "love marriage versus arranged marriage" binary that animated Nadia's thinking about her grandmother (established in Chapter 2) is itself an oversimplification of a complex landscape in which the boundaries between assisted and autonomous partner selection are culturally negotiated rather than fixed. Research on outcomes of assisted versus autonomous partner selection in India shows roughly comparable relationship satisfaction after adjustment for economic and educational factors — a finding that challenges the Western assumption that maximum individual autonomy in partner selection is self-evidently superior.

Sub-Saharan Africa's courtship landscape shows the highest mobile-first penetration of dating app technology, combined with often more conservative cultural and religious contexts, producing a specific pattern of technological courtship that is simultaneously globally connected and locally navigated. Research on dating app use in Nigeria, South Africa, and Kenya finds that users adapt global platforms to local purposes — using apps not just for casual dates and hookups but explicitly for introductions toward marriage, with family consultation remaining an expected part of the process even when the initial meeting was mediated by an app.

These non-Western trajectories are not footnotes to the main story of courtship's future. They represent the experiences of the majority of the world's people and they embody different answers to the questions this chapter raises about what technology should do for desire, what role families and communities should play in partner selection, and what relationship structures constitute flourishing. The global future of courtship will be shaped by all of these answers simultaneously, in ways that will look quite different in Dearborn, Michigan, Lagos, Bangalore, and Shanghai — even as all those places use the same apps.

38.12 The Book's Closing Argument: What Critical Science Enables

Before the closing scene, a synthesis. This book has examined seduction not as a set of techniques but as a social system — a complex intersection of biology, psychology, culture, economics, and power through which human beings pursue connection. We have looked at what attraction is (and is not), what gender and race and class do to the experience of desire, what the evidence actually shows about love and longing and how both work. We have not offered a manual. We have offered, instead, a set of analytical tools and a practice of critical reading that can be applied to any new claim about attraction, any new research finding about desire, any new technology that promises to optimize intimacy.

The critical science framework this textbook models rests on three habits of mind. First: always ask "for whom?" — any finding about what works in attraction or relationships works for some people, in some conditions, at some cost. The universal claims are almost always too universal. Second: always ask "according to what evidence?" — not all evidence is equal, and the difference between a robust meta-analytic finding from a representative sample and a finding from a small WEIRD convenience sample is not a technicality but a substantive difference in what we know. Third: always ask "what is left out?" — who is not in this study, whose experience is not captured in this model, whose desire is rendered invisible by this framework?

These habits of mind do not make love more complicated. They make it more honest. And honesty, it turns out, is one of the things that long-term love most requires.

In the coffee shop, Jordan closes their notebook.

"I keep coming back to this," they say. "The whole book is a deconstruction. We took apart what attraction is, who benefits from what scripts, who gets erased by what assumptions. All of that is right and important." They pause. "But deconstruction alone isn't enough. You can see through every mechanism and still be lonely."

"That's the question I keep sitting with," Sam says quietly. He has been quieter this semester than when it started — quieter in a way that feels like presence rather than absence, as though he has been taking up more space inside himself rather than less. The journaling has been part of it — writing before he decides what he thinks, instead of after. Three months ago, he would have been too embarrassed to say that. He is less embarrassed about more things now, which surprises him. "Understanding what anxiety does to me in relationships doesn't automatically make me less anxious. But it helps. It changes something, even if it doesn't fix everything."

He pauses. He has not told either of them about the conversation he had two weeks ago with a woman he has been seeing casually, a conversation that went badly because he could feel the old patterns running — the withdrawal, the preemptive distance, the joke where something real was called for — and he named them. He said "I can feel myself pulling back and I don't actually want to." It did not fix everything. She is still deciding what she thinks. But something happened in that moment that did not happen before: he stayed in the room. That feels like something.

"I want something real," Sam says. "I don't know how to optimize for that. I'm starting to think you can't."

Nadia is thinking about her grandmother — a woman who had an arranged marriage at twenty-two and who, forty years later, looked at her husband across a kitchen table with an expression of particular, specific affection. Her grandmother had never taken a course called the Science of Seduction. Her grandmother had not known about attachment theory or the Four Horsemen or responsive desire.

Nadia spent the first half of this semester feeling like this gap — between her grandmother's world and the world this course was describing — was an indictment of one or the other. The course made attraction look like a system of mechanisms; her grandmother's marriage looked like it existed outside of systems. But she does not think that anymore. Her grandmother had navigated a system too — just a different one, with different scripts and different constraints, including genuine constraints that Nadia would not want to live inside. And within that system, something had grown that was recognizably the same thing the research was pointing toward: a person who paid deep attention to another person over decades.

"She just paid attention," Nadia says, as much to herself as to either of them. "She paid attention to what he actually needed, not what she imagined he needed. I think she was doing Gottman without knowing it was called that." She smiles slightly. "And maybe he was doing it for her too, in his way. I used to see it as her adapting to him. I'm not sure that's what it was."

She thinks about her bisexuality — the thing she does not discuss with her family, the part of herself she has been carrying as a private weight, the fear of what her grandmother would think that she has gradually come to understand is partly a projection. She does not know what her grandmother would think. She has been assuming the worst, and she does not actually know. What she knows is that this semester has given her language for something she did not have language for: that desire is not a fixed category, that the scripts are human-made and therefore human-changeable, and that the question is not whether her attractions conform to what her family expects but whether they can, eventually, be part of a life lived honestly. She does not know what that will look like. She is a little more willing to wait and find out.

Jordan has been listening with their chin on their hand, notebook closed for once. Their thesis is submitted — submitted two days ago, at 11:47 pm, with a chapter on the construction of the "ideal body" in online queer spaces that took seven drafts to get right and still does not fully capture what they wanted to say. They are proud of it and dissatisfied with it simultaneously, which is, their advisor told them, the correct relationship to have with a first major piece of scholarship.

"I've been thinking about what comes after," Jordan says. "Not the thesis — after this." They gesture, a small movement that encompasses the coffee shop, the semester, the conversation, something larger. "I spent four years studying how queer identity gets formed and policed and performed. And somewhere in there I got good at the analysis and scared of the practice." They look at Sam, then Nadia. "I want something that's actually mine. Not a performance for the queer community and not a performance for my parents' comfort and not a relationship that fits some theory. I want — " they stop. "I want to be seen by someone who doesn't fit neatly into any script I've been given. And I think I'm going to have to build that from scratch, and I think that's okay."

The coffee shop is getting louder. The semester is ending everywhere at once. Outside, someone runs by with a stack of textbooks.

"That's it, though," Jordan says, returning. "That's the whole argument, isn't it? The science doesn't replace the attention. It teaches you what to pay attention to."

This book began with a provocation: that "seduction," properly understood, is not something one person does to another but a complex system of mutual recognition, cultural scripting, biological priming, social positioning, and individual psychology. We argued that critical scientific understanding of this system does not make romance impossible — it makes it more honest.

What critical science enables is not the elimination of desire but its examination. It enables you to ask: whose desire counts in this interaction? What scripts are running that neither of us chose? What does this moment look like from where the other person is standing? What am I bringing from my history that they shouldn't have to manage?

These are not romantic questions in the conventional sense — they don't fit in a Valentine's card. But they are the questions that distinguish a relationship that respects both people as full subjects from one that treats the other person primarily as an object of one's own wanting.

The technology will keep changing. The seduction industry will keep evolving. New forms of mediated intimacy will emerge that we cannot currently predict. What will not change — at least not on any timescale this book can address — are the underlying needs that all of it is trying, in varying degrees of wisdom and clumsiness and hope, to meet.

The needs are simple, actually. To be seen. To be chosen. To be known over time by someone who chooses to stay.

The science does not give us those. It gives us a better map of the territory where we look for them.

Nadia picks up her bag. Her MCAT is in three months, and she has anatomy to review before morning. She hugs Jordan first — a real hug, the kind that lasts a beat longer than social convention requires — then Sam, who is slightly surprised and then not. She has been in this city for three years and for the first time it feels like somewhere she knows rather than somewhere she is learning. She does not know yet whether she will stay after medical school, whether she will find a way to be honest with her family about who she is, whether the woman she was briefly involved with last spring was the beginning of something she could build on. She does not know any of that. She knows that she is less afraid of not knowing.

Sam watches her go, then turns back to the table. He is thinking about his journal — the practice of writing things down before he decides what he thinks about them, rather than after. He is thinking about the conversation two weeks ago. He is not sure it will work out. He is less sure that his uncertainty should stop him from trying than he was a semester ago. He is not, at the moment, dating anyone with any certainty. He is, he thinks, available in a way he has not been before — not desperate, not withdrawn, just available. Not waiting for the anxiety to go away before engaging with the world. Just available despite it. He does not know what will happen next, and this fact sits differently in him now — not as dread but as ordinary uncertainty, the kind everyone lives inside.

Jordan is still at the table, notebook reopened. They are writing about their thesis, or maybe about this conversation. In their thesis conclusion they argued that authentic queer connection requires building relationship structures that do not yet exist — that the most interesting intimate lives being lived by queer people now are being invented rather than inherited. They meant it as a scholarly argument. They are starting to think they also meant it as a plan.

They look up and catch Sam's eye and smile — the particular smile of someone who finds the world sufficiently interesting even when it is not delivering everything they want from it.

"Good semester," Jordan says.

"Yeah," Sam says. "Good semester."

Summary

This chapter surveyed the emerging landscape of technologically mediated desire: AI companions (offering availability, consistency, and non-judgment but raising concerns about calibration and atrophied relational skills), VR intimacy (producing genuine physiological arousal and showing therapeutic promise, with concerns about preference shaping), algorithmic matchmaking evolution (increasingly behavioral but structurally misaligned with user interests), and biopharmaceutical and genetic approaches to love enhancement (scientifically underpowered relative to commercial claims).

We examined the "relationship recession" as a phenomenon requiring nuanced analysis — driven by multiple mechanisms including economic precarity, changed marriage norms, and possibly skill attrition — and the ethical complexity of assuming that more relationships are always better. We addressed asexuality and aromanticism as challenges to the book's foundational assumptions about desire, and consensual non-monogamy as evidence that relationship structure is more variable than dominant scripts acknowledge.

The constant beneath all the technological variation: human needs for mutual understanding, genuine reciprocity, physical presence, and continuity through time. These needs are neither fully met nor fundamentally changed by technology. They are the stable human reality that all our changing courtship technologies are attempts — some wiser, some more problematic — to address.

The closing conversation between Nadia, Sam, and Jordan models the book's final argument: critical scientific literacy does not replace attention to another person. It teaches you what to pay attention to.

Key Terms

responsive AI attachment — The psychological pattern in which users form emotional bonds with AI companions involving genuine feelings of connection, longing, and grief, despite the attachment object having no reciprocal capacity for the relationship.

Dual Control Model — (Referenced in Ch. 37; relevant here) The model of sexual excitation and inhibition that underpins understanding of desire variance — including in VR intimacy contexts.

consensual non-monogamy (CNM) — The broad category of relationship structures involving multiple partners with the full knowledge and consent of all parties; includes polyamory, open relationships, and other forms of ethical multi-partner intimacy.

asexuality — A sexual orientation characterized by little or no sexual attraction to others; distinct from low libido, celibacy, and sexual dysfunction.

aromanticism — The experience of little or no romantic attraction to others; distinct from asexuality (the two can occur independently).

relationship recession — The pattern, documented in survey data from the 2010s onward, of declining rates of sexual activity, partnership, and marriage among young adults — particularly in the United States.

Chapter 38 of 42 — The Science of Seduction: The Psychology and Sociology Behind the Game