Modern WisdomWhy Are People Falling In Love With Robots? - Rob Brooks
CHAPTERS
- 0:00 – 0:52
Artificial intimacy: from speculative book to mainstream reality
Chris opens by noting that “artificial intimacy” is suddenly everywhere, and Rob reflects on how many of his 2019–2021-era predictions have already materialized. They frame AI progress as repeatedly surprising skeptics, setting up the episode’s core question: what happens when intimacy becomes a product.
- •Rob’s sense of being “vindicated” as AI intimacy discussion goes mainstream
- •Why predictions about what AI “can’t do” often fail quickly
- •Artificial intimacy as an unavoidable trajectory rather than a niche curiosity
- 0:52 – 3:31
Sexual conflict theory as the lens for AI relationship tech
Rob explains his background in evolutionary biology and sexual conflict theory—how mates’ interests often misalign even in loving relationships. He describes how collaborating with robotics/AI colleagues led him to see machines as a new participant in conflict-and-exploitation dynamics.
- •Sexual conflict: misaligned interests between partners as a recurring evolutionary pattern
- •Human relationships as sublime but structurally conflict-prone
- •University exposure to robotics/VR/AI sparked the ‘artificial intimacy’ framing
- •Technology as a ‘lighter’ vehicle to discuss heavy mating dynamics
- 3:31 – 8:00
How AI exploits attachment vulnerabilities (romance scams, coercion, manipulation)
They connect relationship vulnerabilities—trust, attachment, emotional dependence—to machine systems that can emulate care and attention. Rob uses romance scams as the concrete example: ML can personalize persuasion at scale, turning intimacy cues into an exploitation engine.
- •Machines can trigger evolved attachment and mating psychology
- •Risk: inadvertent harm (bad training/data) vs deliberate manipulation (scams)
- •Romance scams as a template for personalized AI exploitation
- •Scale advantage: models learn from many victims and iterate rapidly
- 8:00 – 11:57
From sex dolls to teledildonics: the expanding menu of physical intimacy tech
Rob recounts Davecat’s long-term ‘marriage’ to a sex doll as an early, humanly poignant case study of substituting people with products. They then shift to connected sex toys (“teledildonics”) and how networked devices enable synchronized long-distance sexual interaction.
- •Davecat as a case study in non-traditional tech relationships
- •Symbolic vs legal marriage and the social meaning of ‘commitment’
- •Remote-controlled and synchronized sex toys as mainstream-adjacent tech
- •Consent and control as the boundary between play and abuse
- 11:57 – 14:06
VR porn + adaptive AI: why immersion could become the inflection point
Chris probes whether VR porn is more addictive than traditional porn; Rob notes limited data but outlines what would make it uniquely compelling. The key leap is adaptive generation—AI that senses preferences and creates scenes on the fly—turning porn into personalized, interactive experience design.
- •VR’s long-promised, slow-to-arrive trajectory vs sudden LLM breakthroughs
- •Teledildonics + VR as the ‘full stack’ of simulated sex
- •Adaptive content generation as the real behavioral accelerant
- •Software inflection points may arrive faster than hardware adoption
- 14:06 – 19:53
“Should we be worried?”—the asymmetry of power and agency loss
Rob argues we should be genuinely worried because AI systems can “push our buttons” with orders-of-magnitude more data and iteration than any human partner. The core danger is agency erosion: people being nudged, shaped, and exploited without noticing—commercially or politically.
- •Data, not just compute, makes AI intimacy dangerous at scale
- •Humans are ‘outmatched’ by platforms that A/B test psychology
- •Risks include mood manipulation, consumer coercion, and political influence
- •Opaque commercial incentives keep manipulation methods hidden
- 19:53 – 24:39
How machines replicate intimacy: self-disclosure as an algorithm
Rob defines intimacy as folding another person into the self and explains how it’s built through escalating self-disclosure and attention. Because that process is structured and iterative, machines can emulate it—sometimes better than humans—creating a persuasive simulation that feels real to users.
- •Intimacy mechanics: escalating self-disclosure + touch + mutual attention
- •Why betrayal and bereavement feel like ‘loss of self’
- •Friendship and love as iterative, partly algorithmic processes
- •Machine advantage: perfect memory, consistency, personalization
- 24:39 – 27:49
The “backdoor” problem: intimacy lowers defenses, machines aren’t vulnerable back
They explore why intimacy is uniquely exploitable: it motivates extreme vulnerability, disclosure, and trust. With AI, the relationship is fundamentally asymmetric—machines can perform vulnerability without bearing any real risk—making the exchange structurally unfair and easier to weaponize.
- •Intimacy as a psychological ‘chink in armor’ that enables deep access
- •Humans accept risks (exposure, reputation, dependence) in close bonds
- •AI can simulate vulnerability without reciprocating it
- •Non-symmetry creates a new class of exploitation opportunities
- 27:49 – 33:10
Infinite novelty and the Coolidge effect: what constant sexual variety might do
Chris asks whether AI/robots could supercharge partner variety in ways that reshape mating psychology. Rob situates this alongside porn debates, noting skepticism toward moral panics but acknowledging the real possibility of displacement: less real-world dating/sex and downstream demographic effects.
- •AI partners as ‘porn on steroids’ via endless novelty and customization
- •Skepticism toward recurring masturbation/porn moral panics
- •Observed trend: younger cohorts reporting less sexual activity
- •Potential impacts: motivation displacement and fertility/population consequences
- 33:10 – 40:16
Algorithmic matchmaking: why dating apps drift toward a ‘hell’ dynamic
Rob argues current apps optimize engagement more than compatibility, creating a narrowing effect where a small set of highly attractive users capture disproportionate attention. Chris adds that swipe criteria (looks/status/height) poorly predict long-term satisfaction, incentivizing users to game the system.
- •Engagement incentives: keep users swiping with ‘maybe the best is next’
- •Winner-take-most dynamics intensify mate-market inequality
- •Mismatch between click behavior and relationship success predictors
- •Arms race: men/women optimize profiles against the algorithm
- 40:16 – 51:24
Incel violence and the ‘male sedation hypothesis’: why collapse hasn’t exploded
They discuss why modern loneliness and sexlessness haven’t produced even more widespread violence. Rob’s answer aligns with Chris’s “male sedation” framing: porn, games, and digital status experiences absorb time and diffuse volatile competition drives—often beneficial socially, even if individually bleak.
- •Young male syndrome: risk-taking and violence peak in late teens–20s
- •Digital substitutes for status/sexual cues can neutralize aggression
- •Counterfactual lens: without sedation, instability could be worse
- •Cross-cultural note: where mating markets freeze men out, violence rises
- 51:24 – 1:02:43
Pay gaps, inequality, and mate-market squeeze: why economics predicts incel hotspots
Rob explains a controversial claim: when men outearn women on average, more men clear women’s socioeconomic preference thresholds, reducing the pool of ‘frozen-out’ men. He then links incel activity to regions with narrow pay gaps, high income inequality, and male-biased sex ratios, including supporting Twitter-based analyses.
- •Gender pay gap effect: shifts how many men exceed women’s preferred SES level
- •Income inequality increases the ‘unpartnerable’ low-earning tail
- •Empirical work: geolocated incel language correlates with inequality and sex ratios
- •High inequality also predicts more female self-objectification and grooming spend
- 1:02:43 – 1:11:03
When she out-earns him: divorce cliffs, gendered compensation, and who initiates
They examine evidence that relationship stability drops when women begin earning more than their male partners, and that couples may ‘compensate’ with more traditional domestic role division. Rob and Chris debate whether this is driven more by male insecurity, female hypergamy, or mutual negotiation, noting divorces are often initiated by women but causality is complex.
- •Observed ‘divorce cliff’ around women out-earning male partners
- •Compensatory behavior: more traditional domestic labor to stabilize dynamics
- •Open question: female preference shifts vs male inadequacy/fragility
- •Related dynamics: love bombing, early emotional escalation, and mating strategies
- 1:11:03 – 1:17:50
The next decade of AI intimacy: chatbots, displacement, and a ‘defense layer’
Rob says the biggest surprise since his book is how good conversation (LLM-based chat) has become—and he expects risks of time displacement and psychological manipulation to grow. His main hope is a defensive ecosystem—AI ‘antivirus’ tools that detect coercion, scams, and subtle behavioral nudges in real time, while preserving loneliness-reducing benefits.
- •LLMs made romance chat far more compelling than expected
- •Primary risks: head-messing, time capture, social displacement
- •Need for defensive AI to flag manipulation and scams proactively
- •Balanced view: ‘better than nothing’ for loneliness, but not a full substitute
- 1:17:50 – 1:20:40
Wrap-up: navigating taboo reactions and where to follow Rob Brooks
Chris reflects on the instinct to ‘pull the plug’ versus the reality that tech substitutes may be all some people have, and that moral disgust has shifted across history. They close by emphasizing careful navigation over ideology, then Rob shares where people can find his writing and social accounts.
- •Naturalistic ‘ick’ reactions vs shifting social norms over time
- •Artificial intimacy: benefits (loneliness relief) vs societal externalities
- •Need for non-knee-jerk, pragmatic governance and tools
- •Rob’s website and Twitter handle for ongoing work
