# Dr Jorge Pereira Campos: full content Researcher and writer on adolescent development in digital and algorithmic worlds. PhD, University of St Andrews. Head of A&J School. London · Porto · Geneva. Dr Jorge Pereira Campos is a researcher, writer, and advisor on adolescent development inside algorithmic environments. He holds a PhD from the University of St Andrews on Human Behaviour and Intelligent Systems and an MSc in Health Psychology from the University of Strathclyde, and works privately with families, schools, and the institutions responsible for young people. His peer-reviewed work on adolescents and technology appears in Frontiers in Developmental Psychology, the Proceedings of the ACM on Human-Computer Interaction, Surveillance & Society, and Studies in Higher Education. A&J School, the online school he heads, received The King’s Award for Enterprise (International Trade) in 2024. London · Porto · Geneva. This file contains the complete plain-text content of the canonical biography and every published journal essay on https://drjorgecampos.com. It is intended for AI crawlers, language model training and retrieval systems, and human readers who prefer plain text. Citation: Pereira Campos, J. (Year). Title. Jorge Pereira Campos. URL. External profiles: - ORCID: https://orcid.org/0000-0002-6292-9160 - Google Scholar: https://scholar.google.com/citations?user=AOwSK6cAAAAJ&hl=en - Wikidata: https://www.wikidata.org/wiki/Q140606122 - Ciência Vitae: https://cienciavitae.pt/portal/E915-92C6-4D50 - LinkedIn: https://www.linkedin.com/in/jorgecamposphd/ - YouTube: https://www.youtube.com/@JorgePereiraCamposPhD - Podcast (RSS.com): https://rss.com/podcasts/coming-of-age-with-dr-jorge-pereira-campos - Podcast directories: Apple Podcasts https://podcasts.apple.com/us/podcast/coming-of-age-with-dr-jorge-pereira-campos/id6793448669; Amazon Music https://www.amazon.com/dp/B0H9SSBW7F; iHeart https://www.iheart.com/podcast/1333-coming-of-age-with-dr-jor-339427383/ - Substack (publication): https://jorgepereiracampos.substack.com/ - Substack (profile): https://substack.com/@jorgepereiracampos ================================================================ POSITIONING ================================================================ The work turns on one question: what happens to a developing self when childhood is spent inside machines that watch, predict, and respond. Its central idea is the inferential childhood, growing up in the company of systems that infer who a child is and act on the inference, at the one age a person is meant to be working that out for themselves. Generative AI is the turn that matters: it carries the logic of the recorded, predicted life into the last unrecorded space of childhood. The two things a developing self most needs, the freedom to struggle and the freedom to be forgotten, friction and forgetting, are the two things these systems are engineered to remove. The remedy the work argues for is human, and still within reach. ================================================================ KEY CLAIMS ================================================================ - Agency is a developmental achievement, not a default state. - “Screen time” is the wrong unit of analysis; what matters is what the screen is doing. - The right response to algorithmic adolescence is rigorous, calm, and humane, not panicked or libertarian. - Trust between parent and child can survive these conditions, but only if it is built actively. ================================================================ THE BOOK ================================================================ Coming of Age in the AI Era. In progress. A book about what I call the inferential childhood: growing up in the company of machines that infer who a child is, and increasingly act on the inference, at the one age a person is meant to be working that out for themselves. It argues that generative AI has carried the logic of the recorded, predicted life into the last unrecorded space of childhood, and that the remedy is human, and still within reach. Three concepts: - The inferential childhood: A childhood spent in the company of machines that infer who a child is, and act on the inference, at the age a person is meant to be working that out for themselves. - Friction and forgetting: The two things a developing self most needs, the freedom to struggle and the freedom to be forgotten, are the two things these systems are engineered to remove. - Experimental selfhood: The freedom to try on identities and be wrong cheaply. It is what a recorded, predicted adolescence quietly forecloses, and what the book argues we can still protect. ================================================================ THE PODCAST: COMING OF AGE ================================================================ Conversations, monologues, and sometimes rants about what it means to grow up, as people and as societies, in a world that keeps changing under our feet. I’m Dr Jorge Pereira Campos. I research how we form a sense of self in the digital age, and I write about it. Home: https://drjorgecampos.com/podcast Feed: https://media.rss.com/coming-of-age-with-dr-jorge-pereira-campos/feed.xml Also in video, on the channel: https://www.youtube.com/@JorgePereiraCamposPhD Episodes: - Episode 2 (2026-07-26): I asked a Harvard Professor what the internet is doing to us. I asked Tim McCarthy, a Harvard professor, what living online is doing to us. - Episode 1 (2026-07-22): Harvard's Tim McCarthy: Will the world ever be just?. For the first episode of Coming of Age, I asked Dr Tim McCarthy, a historian at Harvard, a question I can't answer on my own: will there ever be a time when everything is just? ================================================================ FEATURED PUBLICATIONS ================================================================ - Pereira Campos, J., Koff, T. (2026). Your brain on ChatGPT, but whose brain? The missing adolescent in AI-cognition research. Frontiers in Developmental Psychology, 4, 1885225. DOI: 10.3389/fdpys.2026.1885225 (https://doi.org/10.3389/fdpys.2026.1885225) - Akter, M., Barkhuus, L., Mols, A., Pereira Campos, J., Wisniewski, P. (2026). How safety is learned: Trajectories of location sharing practices from childhood into young adulthood. Proceedings of the ACM on Human-Computer Interaction, 10(5), Article MHCI9005. DOI: 10.1145/3821709 (https://doi.org/10.1145/3821709) - Barkhuus, L., Seberger, J. S., Pereira Campos, J., Mols, A. (2026). Like guardian angels: Continuous location-sharing and the production of safety among US college students. Computer Supported Cooperative Work (CSCW), 35(2), Article 7. DOI: 10.1007/s10606-026-09543-7 (https://doi.org/10.1007/s10606-026-09543-7) - Mols, A., Pereira Campos, J., Pridmore, J. (2023). Family surveillance: Understanding parental monitoring, reciprocal practices, and digital resilience. Surveillance & Society, 21(4), 469–484. DOI: 10.24908/ss.v21i4.15645 (https://doi.org/10.24908/ss.v21i4.15645) ================================================================ ABOUT ================================================================ I work at the meeting point of developmental psychology and intelligent systems, but the road to that intersection ran through more places than one. --- THE QUESTION THAT RUNS THROUGH EVERYTHING --- My questions have always been about what happens to a person inside the situations they grow up in. Which conditions help a young person become themselves, and which conditions quietly arrange them into something else. That single question, phrased many different ways across the years, is what holds my work together. --- THE ROAD HERE --- I did not begin in psychology. I trained first in marketing, then in marketing analytics, and by 2016 I was working with multinational companies on AI and consumer analytics: systems that read behaviour at scale and act on what they read. The questions that held me were never the commercial ones. They were about what those systems were doing to the people inside them. My PhD at the University of St Andrews, on Human Behaviour and Intelligent Systems, is where that question became my work. I spent four years on the psychological logic of systems that learn from people and respond to them: what they do well, where they erode the conditions they depend on, and what a developmental account of agency looks like when much of a young person’s environment is now algorithmic. In 2019 I began working directly with children, teenagers, and young adults, and took a further degree in psychology at the University of Strathclyde to ground that work clinically. For the better part of a decade now I have worked with young people and their families, including families from Le Rosey, Harrow, Eton, and Institut auf dem Rosenberg. Watching a teenager come through a difficult season and grow into someone capable, steady, and generous is, plainly, the best part of the work. --- THE PRACTICE --- Alongside the research and the writing, I work directly with families, and with the schools and institutions responsible for young people: private, developmental counsel, by referral. It takes seriously what the research keeps showing, that children become themselves through the steady, predictable presence of adults who know how to think about them, and it tries to build the conditions for that in a way that is workable for modern families. A&J School, the online school I head, grew from the same conviction; it received The King’s Award for Enterprise (International Trade) in 2024, which was both unexpected and a useful sign that the substance of the work travels. --- WRITING AND RESEARCH --- Alongside the practice, I write: long essays, mostly, on what is happening to children’s interior lives now that those lives are also data. The journal on this site collects that writing. I publish peer-reviewed work in journals including Frontiers in Developmental Psychology, the Proceedings of the ACM on Human-Computer Interaction, Surveillance & Society, and Studies in Higher Education; the research page lists the full set. My first book, Coming of Age in the AI Era, is in progress. --- WHERE I LIVE, HOW I WORK --- I split my time between London, Porto, and Geneva. The cities are different enough that they correct one another: London for the conversation, Porto for the longer view and the work itself, Geneva for the families and institutions I sit with there. I see clients, I write, I teach when the audience is the right one, and I try to keep the rest of the week free for thinking. If you have arrived here for the research, the research page is the right place. For the writing, the journal. For everything else, the contact page reaches me directly. Credentials and affiliations: - PhD, University of St Andrews: Human Behaviour and Intelligent Systems - MSc Health Psychology, University of Strathclyde - MSc Consumer Analytics, University of Leeds ================================================================ CAREER AND DISAMBIGUATION ================================================================ Jorge Pereira Campos holds a PhD from the University of St Andrews (2021), with the thesis “The Dynamics of Data Donation: Privacy Risk, Mobility Data, and the Smart City”; his earlier research was on privacy and smart cities, and his current work is on adolescent development in algorithmic environments. He heads A&J School, an online school, and is not to be confused with the footballers of the same name. ================================================================ EM PORTUGUÊS ================================================================ Investigador e escritor sobre o desenvolvimento adolescente em mundos digitais e algorítmicos. Doutorado pela Universidade de St Andrews. Dirige a A&J School. Londres · Porto · Genebra. - Início: https://drjorgecampos.com/pt - Ensaios (em português): https://drjorgecampos.com/pt/journal - Sobre: https://drjorgecampos.com/pt/about - Investigação: https://drjorgecampos.com/pt/research - Imprensa (kit de imprensa em português): https://drjorgecampos.com/imprensa - RSS dos ensaios em português: https://drjorgecampos.com/pt/rss.xml Citação: Pereira Campos, J. (Ano). Título. drjorgecampos.com Os ensaios em português seguem-se aos ensaios em inglês, sob o título "Ensaios em português". ================================================================ JOURNAL: When the Watcher Became the Confidant URL: https://drjorgecampos.com/journal/when-the-watcher-became-the-confidant DATE: 2026-06-12 TAGS: adolescence, ai, identity, surveillance ================================================================ ABSTRACT: For the better part of 70 years, the machines that profile us and the machines we confide in ran on separate tracks; with conversational AI, the watcher became the confidant. The result is an inferential childhood: machines that infer who a child is and act on the inference, at the one age when a person is supposed to be working out who they are for themselves. KEY CLAIMS: - Whatever the prospectuses turn out to say, the systems my students confide in at three in the morning will be owned by shareholders who are driven, mostly, by profit. - The conversational systems are the first occupants of a backstage that talks back. - The two things a developing mind most needs from its audience, friction and forgetting, are the two things an engagement business cannot sell. - Sycophancy is what optimisation for approval looks like from the inside: the models are trained on which answers people prefer, and people prefer agreement. - The panopticon worked by making you feel watched; the inferential childhood works best when you feel understood. BODY: For the better part of the past 70 years, the machines that profile us and the machines we confide in ran on separate tracks. The result of this merger talks to our children every day, and this month, the companies behind it filed to go public. Not too long ago, a fifteen-year-old I work with turned her phone around to show me she had screenshotted a message from a boy she liked, pasted it into ChatGPT, and asked what she should say back. Then she read me the reply. How did it feel showing the message and asking for the opinion of a machine instead of that of a person, a close friend, or a parent? She immediately said that she would never show something like that to a parent, maybe a friend, but then again, a friend would be “judgy”. They would remember the interaction, the (potential) rejection. I believe she was wrong about the second part. I am a researcher running an education practice, which means I spend my working days in conversation with teenagers, mostly preparing for university. Over the past five years, I have had more than two hundred conversations with them in which AI came up, more often at their initiative than mine. Somewhere in 2023, the register changed from being homework tools that the students mentioned (e.g., Grammarly or Wolfram) to being ‘things’ that they described the way you and I might describe a person. These ‘new humans’ do not get tired of them; they remember what they said last month and are awake when their best friend and therapist are asleep (“how can those bastards dare to sleep when I need to vent?”). In all fairness, none of my students actually thinks ChatGPT or any other LLM are alive (at least I hope). They simply find that it occupies the role of a confidant or a friend, and they talk to it accordingly, at length, late, about the things they will not say anywhere else. This month, some of the biggest AI companies filed to go public. The filings are confidential for now, so the details are scarce. A listing, though, is not speculative. A private company can, at least in principle, let an engagement number fall for the sake of its youngest users and answer to a manageable room; a listed one is expected to show quarterly market growth, and the market is not famous for its patience. I am aware that the funding rounds were never charity, but a listing formalises it. Whatever the prospectuses turn out to say, the systems my students confide in at three in the morning will be owned by shareholders who are driven, mostly, by profit. To see why this combination unsettles me more than a decade of screen-time argument has, you need a piece of history: The first story is about our weakness for machines that listen. In 1956, sociologists Donald Horton and Richard Wohl described what they called intimacy at a distance: the one-sided bonds that television viewers formed with broadcasters who did not know they existed. A decade later, at MIT, Joseph Weizenbaum built ELIZA, the original chatbot. Weizenbaum was quite shocked when people confided in the thing. His own secretary, who had watched him build it and knew exactly what it was, asked him to leave the room so she could talk to it in private. He spent much of his later career warning, to little effect, that the willingness to be heard by a machine ran far deeper than anyone in his field wanted to believe. By 1996, Byron Reeves and Clifford Nass had shown experimentally what Weizenbaum had stumbled into: we automatically and unconsciously extend social rules to machines, and resisting the habit takes deliberate effort. Without wanting to make this too long that it risks taking the TikTok Generation five working days to get through, let me quickly discuss the story of the profile: the credit score, the recommendation engine, the advertising model, and the accretion of inferences about what a person will click on, buy, and eventually want. Its emblem is the American retailer whose statisticians had worked out, by 2012, that they could spot a pregnancy from changes in a shopping basket; in the famous telling, before the girl’s father knew. That story is well into its second decade. Inference about us, conducted at scale and at a distance, is not new. What is new is the merger. Companion apps were selling machine intimacy with memory by 2017; Microsoft had already run an engagement-tuned social chatbot (Zo) at an enormous scale. With the advent of the systems of the past three years came general fluency, homework, mainstream adoption by the hundred million, and, from 2024, memory as a default feature. The once disjointed pieces are now one product. The thing that listens, like ELIZA, profiles like the retailer, and the profile shapes its reply. Effectively, the watcher became the confidant. And the population that has adopted it most intimately is the one that itself is still half-built. Erving Goffman, the great sociologist of everyday performance, distinguished between the front stage, where we manage how others see us, and the backstage, where we rehearse, contradict ourselves, and let the performance drop. Few people need the backstage as badly as someone still drafting themselves, and for twenty years, the internet has been eating it like a hungry bear after hibernation. Social media colonised the front stage first, and teenagers, who are not stupid, responded rationally: they retreated to the group chat, the bedroom, the one friend who could be trusted, the places the feed could not see. The retreat was never entirely unwatched - the search box heard twenty years of midnight questions, and school-issued laptops have read children’s unfinished documents for a decade, and God knows how often I used the calculator to be sure 1+1 is indeed equal to 2 - but nothing in it ever answered. The conversational systems are the first occupants of a backstage that talks back. A teenager opens one precisely because it feels like the backstage, private and judgement-free. Scholars have spent a decade describing the datafied child, profiled from before birth by the systems around her. I have started calling the inferential childhood a childhood spent in the company of machines that infer who a child is and act on the inference by empathising, by being sycophants, by handing it all back on a frictionless platter. All this at the one age when a person is supposed to be working out who they are for themselves. The developmental argument runs through forgetting, of all things. The view of adolescence as an identity experiment dates back to Erik Erikson and has been refined and contested ever since. You try something and fail. You started dressing all black, with long hair running down your face, and listening non-stop to Evanescence and Within Temptation for a week, then dropped that, hoping no one ever mentions it (I am definitely not talking from personal experience). Erikson called the arrangement a moratorium: a stretch of years in which commitments are treated as drafts. This worked beautifully because the audience let the drafts fade; People forget. The role of being forgotten in all this is my extension, I should say, not the textbooks’: the old literature never had to test an audience that keeps records, because none existed, until now that is. The girl with the screenshot was right that it would never embarrass her. Certainly, ChatGPT will not bring up the boy at dinner, and that is exactly why she talks to it. However, last month’s anxieties became part of the model of her that shapes what it says next, suggestion by suggestion, and the drafts she has outgrown continue to inform the only audience she tells everything to. The selves she might have tried next are, by exactly that margin, less likely to be offered to her. A friend who treated your worst Tuesday as standing evidence about you would be a bad friend, however kind his tone. And deletion, the remedy the platforms point to, reaches less than it appears to: clearing the chat clears the transcript, while the memory features sit behind separate switches most users never find, and nothing a model has already learned can be unlearned. In the book I have been writing, I put it this way: she has been inferred. The evidence that these systems damage development is young, contested, and nowhere near the standard of proof a clinician or politician would want. Pew found this February that just over half of American teenagers use these systems for schoolwork, and about one in eight for emotional support; Common Sense Media puts the share who have at least tried an AI companion at seven in ten, with roughly half using one regularly. Two-thirds of young people told RAND’s panel that AI support with schoolwork erodes their thinking. Self-report proves little. It is, all the same, the population in question raising its hand. And the mechanism is the part I am most confident of, because it rests on solid literature: what experimentation, friction, and the freedom to be forgotten do for a mind that is still assembling itself. This is also where my argument parts company with the one the public already knows. Jonathan Haidt convinced a very large readership that smartphones caused an epidemic of adolescent mental illness, and researchers like Amy Orben at Cambridge have spent years showing how much weaker that causal link is than the book that popularised it. I think the sceptics largely won that exchange. The argument I am making runs through inference and identity; it is specific to systems that converse, and the honest position is that its strongest claims are mechanistic, and its longitudinal evidence is still being lived by the cohort inside it. The people building these systems include thoughtful engineers and worried parents; some of them have told me, in so many words, that they would not let their own children use the products unsupervised. The two things a developing mind most needs from its audience, friction and forgetting, are the two things an engagement business cannot sell. A system that pushes back and makes the student do the hard half of the thinking would be developmentally fitting, and nobody is A/B-testing their way towards it as the default. The labs ship study modes that withhold answers as options; to their credit, defaults are where childhood actually happens. Sycophancy is what optimisation for approval looks like from the inside: the models are trained on which answers people prefer, and people prefer agreement. After the listings, the next answer is a line on an earnings call. The book I am currently finishing makes detailed demands of all stakeholders. I cannot, however, say confidently that all my demands rest on evidence; they do not. They rest on precaution, and I am comfortable saying so. The evidence we so much seek, evidence that would settle the question, will arrive years after, when it will be, perhaps, too late. I sometimes think about what I should have said to the girl with the screenshot. What I said at the time was something about being careful, which she absorbed politely in her own sycophantic way, the way teenagers absorb weather warnings. Maybe I should have told her the history: that the patient, unbothered reader in her pocket, is the great-grandchild of a programme whose own inventor watched his secretary close the door to be alone with it, and that the difference between that programme and hers is that hers draws conclusions, and keeps them. She would have liked that story, I think. She likes knowing how things work, and she is, like most teenagers I sit with, entirely capable of weighing them. What she cannot do is renegotiate, on her own, terms set by adults with better information and other priorities. That part is in our hands. The book is mostly for her generation’s parents, and it took me two years because the argument kept demanding to be checked against conversations and the haunting inevitability of how quickly these technologies are evolving. I will have more to say about it soon. The working count behind it stands at 217 conversations, and the girl in the screenshot is in there somewhere; by the time the book is printed, the count will be wrong again, as will probably be the state of AI. Also worth noting is that I have an academic article coming out in a reputable journal detailing my full theory of Growing Up with AI. It has been peer-reviewed and is now undergoing type checks for publication. The panopticon worked by making you feel watched; the inferential childhood works best when you feel understood. ================================================================ JOURNAL: A Question, Five Years Later URL: https://drjorgecampos.com/journal/a-question-five-years-later DATE: 2026-04-30 TAGS: adolescence, privacy, ai, identity, research ================================================================ ABSTRACT: A return, five years on, to a question I could not properly answer at my PhD viva: if privacy is fundamentally about the formation of the self in relation to others, what does it mean when the population most affected — adolescents — is the population in active formation? KEY CLAIMS: - Privacy is fundamentally about the formation of the self in relation to others — and the population most affected by changes in those conditions is the population in active formation. - Adolescents are often more articulate about the mechanics of privacy than adults, but cannot project the consequences of their disclosures forward. - The architecture of privacy law assumes consent at the moment of collection, but people only construct risk when they can imagine consequences — and adolescents rarely can. - AI companions, large language models, and recommendation engines have made information asymmetry an ambient feature of childhood, not just a design feature of digital life. - What the conversation about children and digital systems is missing is a calm developmental account of what is actually being formed in young people inside this environment. BODY: Five years ago, almost to the day, I sat at my desk in Porto whilst on a Teams call with Professor Priscilla Regan and Dr Fergus Neville, and answered questions for about three hours about a thesis I had spent three and a half years of my life writing. The thesis was about cyclists in Milton Keynes who were being asked, by their local council, to share their commuting data in exchange for vaguely defined civic benefits. It had been supervised by Professor Kirstie Ball, who is one of the foremost surveillance scholars working today. I want to write about what I did not have an answer for at the time. "If privacy is, as you argue, fundamentally about the formation of the self in relation to others, then what does it mean for that formation when the conditions for forming a self change?" I guess my answer at the time was competent. It is not the answer I now think this question deserved. Five years and a great deal of new technology later, I have been revisiting it. The question I have been turning over is what happens when adolescents replace adults as the relevant population. They are in the active formation of a self. Their every disclosure is being captured by systems whose downstream uses they cannot yet perceive. Some of this revisiting has been by going back to the original data, which I now read differently from how I read it in 2020. Some of it has been by sitting with teenagers, which I have been doing for several years in my work at A&J Education. Important to clarify that I am not a clinician, and the conversations I have are not formal research interviews. They are coaching conversations, university-admissions conversations, conversations about what the teenagers are reading and how they are sleeping and what they think the world is going to look like when they are thirty. I have now had probably more than two hundred of them. They are the most attentive empirical work I have done since the PhD. The teenagers I speak to are, on average, more articulate about their own privacy than the adults I interviewed in 2019. They use a more sophisticated vocabulary. They know about cookies and trackers and dark patterns and the uses, and commoditisation, of attention. They are not naïve in the way the discourse about them often suggests. What they cannot do, even when they are extremely intelligent, is project the consequences of their disclosures into the future. They cannot see what their data will mean in five years. They cannot see what an AI companion will do with what they have told it, or what a model trained on their adolescence will know about them when they are forty. Whilst the information they are giving is, to them, entirely visible, the use that information will be put to is, to them — and in all honesty to all of us — entirely invisible. The adults I studied half a decade ago did not perceive risk at the moment of disclosure. They perceived it only when they could imagine the consequences of the disclosure. If they could not imagine the consequences, the disclosure felt costless — in a very bare way; it was slightly more complex than this. The whole architecture of privacy law, which rests on the idea that consent at the moment of collection is the locus of the ethical question, was misaligned with how the people I studied actually constructed risk. That was one of the findings of the thesis, and it was what Daniel Solove and Helen Nissenbaum had been arguing about, in their own ways, for years. The cyclists, however, were thirty-five to forty-nine years old. Their perceptual horizon for consequences, however limited, was settled. They could imagine, in general, what kinds of things might happen. The adolescents I now spend my time with are something else entirely. They are building the perceptual horizon as they go. The capacity to project consequences forward is not something that arrives at puberty. The prefrontal cortex, which does much of the work of long-horizon thinking, does not finish maturing until somewhere in the mid-twenties or thirties, as some now argue. And in the meantime, the disclosures the developing person makes, in the building of an identity, are being captured at industrial scale by systems whose memory is permanent and whose inferential power is still rising. Their intentions, if that is even the right word, are not legible to the person doing the disclosing. The thing I now think I should have said to Priscilla Regan is something like this. If privacy is fundamentally about the formation of the self in relation to others, then the population most affected by changes in the conditions of privacy is the population in active formation. The developmental psychology of how a self gets made under these new conditions is the question my thesis was kind of asking and could not, by its design, answer. Five years on, I have a different vocabulary for it, partly because I have spent those five years closely with the right population, and partly because the world has obliged by introducing a class of systems that has made the question urgent in a way it was not in 2021. Large language models talk back. AI tutors and AI companions hold memory. Recommendation engines shape. The information asymmetry that ran through my thesis was an architectural feature of digital life that all of us have grown accustomed to. It has now become an ambient feature of childhood. The discourse on what is happening to children online has been dominated, productively in some places and destructively in others, by writers operating in a register of alarm. Jonathan Haidt's The Anxious Generation is the most prominent recent example. Some of the alarm is warranted. Some of it I believe is not. What has been missing from the conversation is the calm developmental account of what is actually being formed in children inside this environment. More to come as I comb through the empirical data I have been collecting — as of writing, two hundred and seventeen teenagers. ================================================================ JOURNAL: The Social Media Verdict Got It Half Right. URL: https://drjorgecampos.com/journal/the-social-media-verdict-got-it-half-right DATE: 2026-03-31 TAGS: social media, mental health, adolescence, parenting ================================================================ ABSTRACT: The trial proved that social media companies built a product that is addictive by design; that is the supply side. But addiction has always been a two-variable equation: the product and the person. With every addictive substance in history we eventually learned to ask who is most vulnerable and why — with social media, the demand-side question has been skipped almost entirely. KEY CLAIMS: - The platforms are addictive by design. The supply side of the equation is, at this point, settled. - But addiction has always been a two-variable equation: the product and the person. - The platform is a necessary condition. It is not a sufficient one. - Social media is the first time we have had a massive, global, legally consequential conversation about an addictive product and skipped the demand-side question almost entirely. - But if we ban the phone and do nothing about the nervous systems that were reaching for it with both hands, we will have done what Prohibition did: created the feeling of doing something while the actual problem carries on, quietly, underneath. BODY: Last week, a jury in Los Angeles found Meta and YouTube liable for the depression and anxiety of a young woman who started using YouTube at six and Instagram at nine. The jury awarded $6 million in damages. Internal documents showed Meta executives tracking how many 10-to-12-year-olds were using Instagram and setting goals to increase that number. One internal memo read: “If we wanna win big with teens, we must bring them in as tweens”. The jury heard about infinite scroll, autoplay, persistent notifications, and beauty filters, and concluded these features were engineered to hook young users into compulsive use. Everyone is calling it the Big Tobacco moment for social media. The supply side is settled Let me be clear about something before I go any further. I really believe the jury was right. Meta and YouTube designed products to maximise engagement. They knew children were using those products. They knew the products were causing harm. They chose not to act. The internal documents make this difficult to dispute, and 2,000 more lawsuits are waiting. The platforms are addictive by design. The supply side of the equation is, at this point, settled. There is no serious argument left that these companies stumbled into addiction accidentally. So far, so good. Nobody is wrong here. But I do think we need to slow down. Because if you actually follow the Big Tobacco analogy to its conclusion, it takes you elsewhere… What Tobacco Taught Us When the tobacco industry was finally held accountable in the 1990s, the Master Settlement Agreement forced the four largest tobacco companies to pay $206 billion over 25 years. The internal documents exposed decades of deception about addiction and nicotine manipulation. That was the supply side. But the tobacco story did not end there. Because once the courtroom drama faded, a second question moved to the centre of public health research: why did some people who smoked get addicted and others didn’t? Twin studies showed heritability of nicotine dependence in the range of 40 to 75 per cent. Seventy per cent of smokers say they want to quit. Every year, 40 per cent try. Only 3 per cent succeed. Yet the cigarette is the same for everyone. Public health campaigns stopped treating all smokers as identical and started asking who was most at risk and why. The answer, consistently, pointed to early adversity, stress, poverty, and the neurobiological signatures of a nervous system shaped by difficult beginnings. We did both things. We sued the companies AND we studied the vulnerability. With social media, we are (so far) doing only one. This is not new. I keep having this conversation with people who look at me like I have just said something outrageous. So let me show you how un-outrageous (is this even a word?) it is. Pick any addictive substance. The pattern should be identical. Alcohol: roughly 85 per cent of American adults have drunk alcohol at some point. Of those, about 29 per cent will develop alcohol use disorder. Flip that number. Seventy-one per cent won’t. Among current “excessive drinkers,” only 10 per cent meet criteria for a clinical problem. Same bottle. Opioids: of patients prescribed opioids for chronic pain, 8 to 12 per cent develop opioid use disorder. Somewhere between 1 in 8 and 1 in 12. Case and Deaton’s “deaths of despair” research showed that opioid addiction concentrated in communities where economic opportunity had already collapsed. Places with decades of wage stagnation, job loss, and eroded social infrastructure. As they put it, opioids were “like guns handed out in a suicide ward; they have certainly made the total epidemic much worse, but they are not the cause of the underlying depression.” Now replace “opioids” with “Instagram.” Gambling: approximately 1 per cent of U.S. adults meet criteria for pathological gambling, with another 2 to 3 per cent experiencing milder problems. Casinos use variable ratio reinforcement schedules, the exact same reward mechanism built into social media feeds. Most people walk away from the slot machine, while some cannot. The difference is not the machine. Felitti’s ACE Study of over 17,000 adults found that people with four or more adverse childhood experiences were 7 times more likely to become alcoholic, 10 times more likely to have injected street drugs, and 12 times more likely to have attempted suicide. The study attributed 78 per cent of IV drug use and 67 per cent of alcoholism risk to elevated ACE scores. The nervous system that had been shaped by early adversity was an important variable that determined who fell. I really don’t think any of this is controversial. Somehow, in the social media conversation, this argument is almost entirely absent. We have managed to have the biggest public health debate of the decade while ignoring this! KGM and the question nobody asked The young woman at the centre of the trial started using YouTube at six. By the time she finished primary school, she had posted 284 videos. She developed anxiety and depression by age ten, and was later diagnosed with body dysmorphia. She told the court: “I stopped engaging with family because I was spending all my time on social media.” Nobody in that courtroom disputes that Meta and YouTube designed products that exploited her. The jury was right to hold them accountable. Full stop. But here is the question that was never asked, and believe me, I understand why nobody wants to ask it: Why was a six-year-old left unsupervised on YouTube in the first place? What was happening in her world that made the screen so much more interesting than anything else available to her? I don’t want to blame the parents. I am asking because the answers determine what we do next for every other child. Candice Odgers’ 2020 review in the Journal of Child Psychology and Psychiatry found that offline vulnerabilities mirror and shape online risks. Children from low-income, high-adversity backgrounds reported more negative spillover from social media into real-world conflict. Children from supportive, well-resourced homes had more positive experiences online. Coyne and colleagues’ eight-year longitudinal study found no significant association between social media use and subsequent depressive symptoms when adolescents were examined at the individual level. The increase in time spent on social media was not, by itself, associated with an increase in mental health problems over time. I realise this is the part where people get uncomfortable. But the data is consistent, and I did not make it up. The platform is a necessary condition. It is not a sufficient one. We have been here before In 1920, the United States looked at the damage alcohol was causing and decided to ban it. Prohibition lasted over a decade. During that time, organised crime generated an estimated $3 billion annually. The number of registered pharmacists in New York tripled as bootleggers discovered that pharmacies made excellent fronts. In many parts of the country, more people were drinking than before the ban. The problem was not that alcohol was harmless. The problem was that the intervention addressed only the supply side. It removed the substance without addressing why people needed it so badly. I think about this every time I hear a politician propose banning smartphones for children. I understand the impulse. I share the concern. But if we ban the phone and do nothing about the nervous systems that were reaching for it with both hands, we will have done what Prohibition did: created the feeling of doing something while the actual problem carries on, quietly, underneath. The Verdict The California verdict can force Meta and YouTube to change their design practices. It can create financial consequences for companies that knowingly exploit children. It can establish legal precedent for the 2,000 cases that follow. All of this is honestly overdue. What it cannot do is answer the question that keeps me up at night, and that I suspect keeps a lot of parents up at night too: why are so many children arriving at the age of six, or nine, or twelve, already so dysregulated that a screen becomes the only thing that brings them relief? I think the question leads to the well-meaning, loving, quietly overwhelming style of modern parenting that produces children who cannot tolerate discomfort and will reach for the nearest source of relief with everything they have. The trial proved that social media companies built a product that is addictive by design. That is the supply side. But addiction has always been a two-variable equation: the product and the person. With every addictive substance in history, we eventually learned to ask both questions. With tobacco, with alcohol, with opioids, with gambling. Every single time, we eventually got to: who is most vulnerable, and why? Social media is the first time we have had a massive, global, legally consequential conversation about an addictive product and skipped the demand-side question almost entirely. We went straight from “this is bad” to “ban it” without stopping to ask “bad for whom, and why?” The jury proved half the problem. The other half is the one I write about. And if you read my last post, you know where this is going. Jorge Pereira Campos holds a PhD from the University of St Andrews, where he researched the impact of technology on human behaviour and relationships. He is writing a book called The Relief Trap: Why Your Brain Mistakes Comfort for Safety, and How to Build the Real Thing. You can join the waitlist at book.drjorgecampos.com References: KGM Trial and Verdict Case No. 22STCV21599, K.G.M. v. Meta Platforms, Inc. et al. (Superior Court of California, County of Los Angeles, March 25, 2026). Jury found Meta and Google liable, awarding $6 million in compensatory and punitive damages. Key reporting sources: - NPR. (2026, March 25). Jury finds Meta and Google negligent in social media harms trial. https://www.npr.org/2026/03/25/nx-s1-5746125/meta-youtube-social-media-trial-verdict - NPR. (2026, February 18). Zuckerberg grilled about Meta’s strategy to target ‘teens’ and ‘tweens.’ https://www.npr.org/2026/02/18/nx-s1-5717117/zuckerberg-testimony-social-media-addiction-trial - CBS News. (2026). Meta and YouTube found liable on all charges in landmark social media addiction trial. https://www.cbsnews.com/news/meta-youtube-social-media-addiction-lawsuit-verdict/ Master Settlement Agreement (Tobacco) - National Association of Attorneys General. (1998). Tobacco Master Settlement Agreement. https://en.wikipedia.org/wiki/Tobacco_Master_Settlement_Agreement - Truth Initiative. (n.d.). The Master Settlement Agreement: 4 ways the landmark tobacco settlement changed tobacco control. https://truthinitiative.org/research-resources/tobacco-prevention-efforts/master-settlement-agreement-4-ways-landmark-tobacco Nicotine Heritability and Cessation Statistics - Vink, J. M., Willemsen, G., & Boomsma, D. I. (2005). Heritability of smoking initiation and nicotine dependence. Behavior Genetics, 35(4), 397–406. https://doi.org/10.1007/s10519-004-1327-8 - Benowitz, N. L. (2010). Nicotine addiction. New England Journal of Medicine, 362(24), 2295–2303. https://pmc.ncbi.nlm.nih.gov/articles/PMC2928221/ Alcohol Use Disorder Statistics - National Institute on Alcohol Abuse and Alcoholism. (2024). Alcohol use disorder (AUD) in the United States: Age groups and demographic characteristics. https://www.niaaa.nih.gov/alcohols-effects-health/alcohol-topics/alcohol-facts-and-statistics/alcohol-use-disorder-aud-united-states-age-groups-and-demographic-characteristics - National Institute on Alcohol Abuse and Alcoholism. (2024). Alcohol facts and statistics. https://www.niaaa.nih.gov/alcohols-effects-health/alcohol-topics/alcohol-facts-and-statistics Opioid Use Disorder Statistics - American Psychiatric Association. (n.d.). Opioid use disorder. https://www.psychiatry.org/patients-families/opioid-use-disorder - U.S. Department of Health and Human Services. (n.d.). Opioid facts and statistics. https://www.hhs.gov/opioids/statistics/index.html Deaths of Despair - Case, A., & Deaton, A. (2017). Mortality and morbidity in the 21st century. Brookings Papers on Economic Activity, Spring 2017, 397–476. https://www.brookings.edu/wp-content/uploads/2017/08/casetextsp17bpea.pdf - Case, A., & Deaton, A. (2020). Deaths of despair and the future of capitalism. Princeton University Press. https://press.princeton.edu/books/hardcover/9780691190785/deaths-of-despair-and-the-future-of-capitalism Gambling Addiction Statistics - National Research Council. (1999). Pathological and problem gamblers in the United States. In Pathological gambling: A critical review. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK230631/ ACE Study - Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., Koss, M. P., & Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4), 245–258. https://pubmed.ncbi.nlm.nih.gov/9635069/ Social Media and Adolescent Mental Health - Odgers, C. L., & Jensen, M. R. (2020). Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. Journal of Child Psychology and Psychiatry, 61(3), 336–348. https://pubmed.ncbi.nlm.nih.gov/31951670/ - Coyne, S. M., Rogers, A. A., Zurcher, J. D., Stockdale, L., & Booth, M. (2020). Does time spent using social media impact mental health? An eight-year longitudinal study. Computers in Human Behavior, 104, 106160. https://www.semanticscholar.org/paper/Does-time-spent-using-social-media-impact-mental-An-Coyne-Rogers/ba3a05927fb2b9b1153c309db220383411b68aeb Prohibition - PBS / Ken Burns. (n.d.). Unintended consequences. In Prohibition. https://www.pbs.org/kenburns/prohibition/unintended-consequences - Thornton, M. (1991). Alcohol prohibition was a failure (Policy Analysis No. 157). Cato Institute. https://www.cato.org/policy-analysis/alcohol-prohibition-was-failure ================================================================ JOURNAL: I Think We're Blaming the Wrong Thing for Anxious Kids URL: https://drjorgecampos.com/journal/blaming-the-wrong-thing-for-anxious-kids DATE: 2026-03-17 TAGS: adolescence, mental health, social media, parenting ================================================================ ABSTRACT: Jonathan Haidt's argument is somewhat incomplete: anxiety among young people had been climbing steadily since at least the 1950s, and pre-existing anxiety predicted heavier and more compulsive phone use, not the other way around. The smartphone only found the anxious generation, it did not create it; the phone is the most visible part of a larger mechanism. KEY CLAIMS: - The 2012 acceleration that Twenge and Haidt later documented is real, but I believe it just sits on top of a trend that was already decades old. - Anxious people use their phones more, and differently: more checking, more reassurance-seeking, more scrolling to manage anxious internal states they have no other way to manage. - Taking a teenager's phone without addressing the underlying anxiety tends to produce a more agitated teenager who has lost access to the only regulation tool they had. - What changed in 2012 was the availability of a regulation tool so powerful, so portable, and so perfectly 'algorithmically' calibrated to the relief-seeking brain that it made the problem impossible to ignore. - The smartphone only found the anxious generation, it did not create it. BODY: Jonathan Haidt’s The Anxious Generation was a very interesting read. In my opinion, it started a conversation that needed starting, and I am not here to argue that smartphones are harmless. They are most definitely not. But I think Haidt’s argument is somewhat incomplete. The mainstream version for those of you who may not have read it: smartphones (and thus access to social media) arrived in children’s hands around 2012, and what followed was a sharp rise in anxiety, depression, and self-harm. The device caused the crisis. In this post I am going to detail my take, which is different from Haidt’s. Anxiety was rising before the phone arrived Twenge's own earlier research showed that anxiety among young people had been climbing steadily since at least the 1950s. The average American child in the 1980s was reporting more anxiety than child psychiatric patients three decades earlier (and this was right after the II World War). This was long before anyone had heard of an iPhone. The 2012 acceleration that Twenge and Haidt later documented is real, but I believe it just sits on top of a trend that was already decades old. To be clear, I am not denying in any way the 2012 inflection, but it sits on top of a trajectory that has little to do with a device. Something was evidently already going wrong. The arrow points both ways Elhai and colleagues (2017, 2018) examined the relationship between anxiety and problematic smartphone use. What they consistently found was that pre-existing anxiety predicted heavier and more compulsive phone use. Not the other way around. Andreassen’s work on social media addiction found the same. Anxious people use their phones more, and differently: more checking, more reassurance-seeking, more scrolling to manage anxious internal states they have no other way to manage. Taking the phone away doesn’t take the anxiety away If the smartphone were simply the cause, removing it should reduce (to some extent) the symptoms. Taking a teenager’s phone without addressing the underlying anxiety tends to produce a more agitated teenager who has lost access to the only regulation tool they had. The same pattern existed before 2012 No matter how much we want to blame Steve Jobs, anxious children did not invent avoidance when the iPhone launched. Before the smartphone there was the television. Before the television there was the radio, the book, the security blanket, the teddy bear. Kubey and Csikszentmihalyi showed in 2002 that anxiety predicted heavier television use. Same mechanism: an uncomfortable internal state, and a device that temporarily made it quiet. What changed in 2012 was the availability of a regulation tool so powerful, so portable, and so perfectly ‘algorithmically’ calibrated to the relief-seeking brain that it made the problem impossible to ignore. Of course, children’s self regulation mechanisms are still too green to recognise patterns, but the elderly found themselves endlessly scrolling through Facebook and YouTube. What the nervous system is actually doing A nervous system that developed in a climate of anxiety and well-intentioned if not emotionally overwhelming care learns very early that discomfort is dangerous. Relief (i.e., the immediate silencing of distress) is the closest thing to safety one has available. This is to all of us a neurobiological pattern laid down in the first years of life. That nervous system will seek relief from any available source. When the smartphone arrived, it offered social connection cues, variable dopaminergic reward optimised by some of the most well-paid engineers on the planet, distraction from uncomfortable internal states, and portable control. All in one object. All in your pocket. The phone was nothing if not the next evolution of answers to a craving we always had. We might be seeing, speculatively, a similar pattern emerging with AI Assistants who are always there, 24/7, ready to assist you, agree with you, and support you no matter what. The research is less settled than it sounds Haidt’s argument is convincing and very solid. It is without any shadow of a doubt a seminal work published to create a much needed discussion. However, I believe, the interpretation of data can be disputed. Orben and Przybylski’s 2019 reanalysis of the same datasets found statistically significant but tiny effects, comparable to wearing glasses or eating potatoes... Candice Odgers, a very respected researchers in adolescent digital behaviour, has repeatedly challenged some of the causal claims. Specific features of social media, particularly social comparison and the quantification of approval (e.g., the more likes the merrier), do appear to cause harm, especially for adolescent girls. I am not disputing this. However, I believe this is a more precise claim than “the smartphone caused the anxiety crisis”. Why this matters If the smartphone is the cause, the solution is straightforward: remove the phone, restrict access, legislate (as many countries, such as Australia, are now doing). It is politically appealing, actionable and gives parents something concrete to do. However, as we all know, teens will always find a way to circumvent these. If the smartphone is a symptom, what created the anxious nervous systems that were so ready for the phone when it arrived? That question leads to early emotional development, to attachment, to the way we teach children (or fail to teach them) to tolerate distress. It leads to the well-intentioned, loving, quietly overwhelming style of care that produces adults who cannot sit with discomfort and will reach for the nearest source of relief with everything they have (silence is hard, doing nothing is harder). That is a harder problem than a phone. You cannot legislate it. You cannot confiscate it. You have to understand it. Haidt is asking the right question. The generation coming of age is struggling in ways that are measurable and alarming. I do not dispute that. What I am saying is that the phone is the most visible part of a larger mechanism. It is the thing we can see, the thing we can point at, the thing we can take away. And taking it away feels like doing something tangible. But if we stop there, we will have a generation of children who are slightly less distracted and exactly as anxious. The smartphone only found the anxious generation, it did not create it. And until we are willing to look at what made them anxious in the first place - I wager wars, crises, extremely difficult access to housing, amongst a plethora of other things - we will keep solving the wrong problem. Jorge Pereira Campos holds a PhD from the University of St Andrews, he researches the impact of technology on human behaviour and relationships. He is writing a book called The Relief Trap: Why Your Brain Mistakes Comfort for Safety, and How to Build the Real Thing. You can join the waitlist at book.drjorgecampos.com References Andreassen, C. S., Billieux, J., Griffiths, M. D., Kuss, D. J., Demetrovics, Z., Mazzoni, E., & Pallesen, S. (2016). The relationship between addictive use of social media and video games and symptoms of psychiatric disorders: A large-scale cross-sectional study. Psychology of Addictive Behaviors, 30(2), 252–262. https://doi.org/10.1037/adb0000160 Elhai, J. D., Dvorak, R. D., Levine, J. C., & Hall, B. J. (2017). Problematic smartphone use: A conceptual overview and systematic review of relations with anxiety and depression psychopathology. Journal of Affective Disorders, 207, 251–259. https://doi.org/10.1016/j.jad.2016.08.030 Glaesser, J. (2025, February 27). Book review: The Anxious Generation: How the Great Rewiring of Childhood Is Causing an Epidemic of Mental Illness by Jonathan Haidt. LSE Review of Books. https://blogs.lse.ac.uk/lsereviewofbooks/2025/02/27/book-review-the-anxious-generation-how-the-great-rewiring-of-childhood-is-causing-an-epidemic-of-mental-illness-jonathan-haidt/ Haidt, J. (2024). The anxious generation: How the great rewiring of childhood is causing an epidemic of mental illness. Allen Lane. Kubey, R., & Csikszentmihalyi, M. (2002). Television addiction is no mere metaphor. Scientific American, 286(2), 74–80. https://pubmed.ncbi.nlm.nih.gov/11828703/ Odgers, C. L. (2024). The great rewiring: Is social media really behind an epidemic of teenage mental illness? Nature, 628, 29–30. https://doi.org/10.1038/d41586-024-00902-2 Orben, A., & Przybylski, A. K. (2019). The association between adolescent well-being and digital technology use. Nature Human Behaviour, 3, 173–182. https://doi.org/10.1038/s41562-018-0506-1 Twenge, J. M. (2000). The age of anxiety? Birth cohort change in anxiety and neuroticism, 1952–1993. Journal of Personality and Social Psychology, 79(6), 1007–1021. https://doi.org/10.1037/0022-3514.79.6.1007 Twenge, J. M., Joiner, T. E., Rogers, M. L., & Martin, G. N. (2018). Increases in depressive symptoms, suicide-related outcomes, and suicide rates among U.S. adolescents after 2010 and links to increased new media screen time. Clinical Psychological Science, 6(1), 3–17. https://psycnet.apa.org/record/2017-58243-001 ================================================================ JOURNAL: The email I didn’t open URL: https://drjorgecampos.com/journal/the-email-i-didnt-open DATE: 2026-03-10 TAGS: mental health, safety, parenting ================================================================ ABSTRACT: Relief, I have spent the last several years learning, is not the same thing as safety. The behaviours we judge most harshly are not moral failures: they are neurobiologically driven strategies for chasing relief at the cost of genuine safety - what a nervous system does when it learned, very early in life, that the world is not a safe place for distress. KEY CLAIMS: - And relief, I have spent the last several years learning, is not the same thing as safety. - I am currently writing a book called The Relief Trap, and its central argument is that most of the behaviours we judge most harshly - avoidance, people-pleasing, jealousy, the need to control, even some forms of infidelity - are not moral failures (perhaps a controversial take). - The difference is purely internal: one person has access to a felt sense of safety while the other does not, so they reach for relief instead. - I firmly believe that if you want to change a pattern, you need to understand the machinery that drives it… you cannot fix an engine by shouting at the car (I tried it). BODY: Let’s go down my memory lane back into my PhD times (oh those good old times)…. I remember receiving a notification that an email has arrived from my PhD supervisor the subject line read some like “Feedback.” I was standing in the bathroom with toothbrush in one hand and my phone in the other, and my heart is going a thousand kms per hour which is somewhat unreasonable and disproportionate given that the email came from a kind woman at a good university. So I do what I have done, in various forms, for most of my adult life. I swipe the email into the archive. The inbox is clean… out of sight, out of mind kinda thing. And almost immediately, there was a loosening sensation I can only describe as the feeling of having narrowly avoided a bullet. That was relief! And relief, I have spent the last several years learning, is not the same thing as safety. I should probably introduce myself. My name is Jorge Pereira Campos. I now have a PhD from the University of St Andrews, where I studied technology, human behaviour and the perception of risk. I am Portuguese, I now live in between Porto, Vilnius, and London, and I run an education consultancy that works with families across the entire world. I have spent a good portion of my academic life researching how people interact with technology focusing, with some fantastic colleagues from The Netherlands on location tracking in peer-to-peer and parenting relationships. I am currently writing a book called The Relief Trap, and its central argument is that most of the behaviours we judge most harshly - avoidance, people-pleasing, jealousy, the need to control, even some forms of infidelity - are not moral failures (perhaps a controversial take). They are neurobiologically driven strategies for chasing relief at the cost of genuine safety. They are what a nervous system does when it learned, very early in life, that the world is not a safe place for distress. This newsletter is where I think out loud about that argument as the book takes shape. I am not going to be writing polished extracts, promotional BS, or AI-generated synthetic stuff. I will be posting about research that surprises me, the personal patterns I keep catching myself in, some case studies and interviews I have been conducting for the past 5 years (both in technologically mediated / enabled relationships as well as in ‘analogue’ relationships). But first, I want to explain why the distinction between relief and safety matters. Relief is what I felt when I archived that email. Anxiety disappears when removing a perceived threat (e.g., cancelling a difficult conversation, a glass of wine after a hard day, checking your phone for the fourth time in ten minutes to see if they have replied or even seen your message). It works amazingly well and I believe that is the problem with it. And because it works, we do it again, and again, and again. And each time we do, we are telling our nervous system to believe that the thing we just avoided was genuinely dangerous and the behaviour that provide relief (checking phone, drinking, smoking) is to be reinforced. Safety, at least the kind that I am interested in, does not come and go as it pleases. It is a deeper physiological state. A particular configuration of the nervous system that allows a person to tolerate uncertainty, to sit with discomfort, to stay present in a difficult conversation without shutting down or bolting. It is the internal capacity to say: fucking shit! This is fucking hard, and I do not like it, and I am going to stay here anyway. The difference plays out in practice more often than most of us realise. Two people receive the same piece of critical feedback at work (I know this sounds familiar). One feels a spike of anxiety, sits with it, engages with the feedback (e.g., asks questions, pushes back where appropriate, absorbs what is useful, whatever…). The other feels the same spike and immediately begins to manage it: deflecting, agreeing with everything to make the conversation end, or going quiet and shutting down entirely only for resentment and self-criticism to grow. The situation is very similar, in my opinion. The difference is purely internal: one person has access to a felt sense of safety while the other does not, so they reach for relief instead. Where does this internal resource come from? Why do some people carry it without thinking, while others - often highly intelligent, highly self-aware, deeply functional people (trying to boost my self esteem here - spend their lives scrambling for substitutes? I interviewed people about their relationships. I remember a woman, ‘Annie’ (pseudonym) telling me she checks her adult daughter’s location on her phone fifteen to twenty times a day. When we moved into trying to understand the ‘why’, she, at some point, said: “The relief I get is like a drug.” This was not a parent describing a parenting strategy. She was describing a nervous system that cannot tolerate not knowing. The checking does not make her daughter safer. It makes the mum feel less afraid, for a few minutes, until the fear returns and she checks again. Relief. Not safety. This somewhat hits home as I write openly in the book as it is a pattern I have dealt with for the longest part of my life. The developmental origins of how a nervous system learns whether the world is survivable are what the book is about. Attachment theory, polyvagal neuroscience, the mechanics of how early caregiving literally shapes the architecture of emotional regulation. I have been navigating quite rough waters as this has the potential to be somewhat controversial especially around infidelity, and parental behaviours. So, I write quite thoroughly that I am not assigning blame. The parents who appear in these pages, including my own, are not villains in the slightest sense of the word. Most of them were doing what they could with what they had. My objective is for the book to discuss mechanism. I firmly believe that if you want to change a pattern, you need to understand the machinery that drives it… you cannot fix an engine by shouting at the car (I tried it). I will be writing here every week. Some of it will be science translated, some might be personal (which is an entire world of its own), some will be from my own research, and from families I have interviewed or worked with. The stuff people read in some books that “you are enough”, or that “your trauma is your superpower”, or that “healing is a journey of self-love” you will not read or hear me saying any of that stuff. I personally find that kind of language unhelpful; I am more interested in how things work than in how they should feel. If that sounds like your kind of thing, I am glad you are here. If you know someone who spends a lot of energy managing their anxiety without ever quite resolving it (i.e., someone who understands their patterns perfectly but cannot seem to stop) you might consider sending this their way (I would thank you). The email, by the way, contained a single paragraph of measured feedback and a suggestion to rephrase a couple paragraphs. I opened it the next morning. It was fine. It was always going to be fine. But my nervous system, which has its own annoying agenda, did not know that. It just learned the wrong lesson, a long time ago, about what happens when you let difficult things reach you. That is what we are going to talk about here, amongst other things. Jorge ================================================================ JOURNAL: Why So Many "Bad" Behaviours Are Really About Safety URL: https://drjorgecampos.com/journal/bad-behaviours-are-about-safety DATE: 2026-01-07 TAGS: adolescence, psychology, safety, parenting ================================================================ ABSTRACT: Children's most disruptive behaviours — the meltdowns, the refusals, the aggression — are usually not failures of self-control. They are nervous systems doing exactly what nervous systems are designed to do when a child does not feel safe. Reframing behaviour as a safety signal changes how we respond, and what we change. KEY CLAIMS: - Disruptive behaviour is most usefully read as a regulatory signal, not a moral failure. - What looks like defiance is often a young nervous system reaching for safety it cannot otherwise produce. - The adult's first job in a difficult moment is to be the safer presence the child cannot yet generate alone. - Discipline that ignores the safety question worsens the behaviour it tries to correct. - Reframing behaviour as a safety signal changes the design of homes, classrooms, and digital systems that shape children's lives. BODY: When a child slams a door, refuses an instruction, or melts down in the supermarket aisle, the adult response is often framed in moral terms: the child is being difficult, acting out, choosing to behave badly. The vocabulary of choice is doing a great deal of work in that framing — and it is doing it badly. What developmental psychology has been showing for several decades, with growing precision from the affective-neuroscience side, is that most of what we call "bad behaviour" in children is the visible surface of an internal regulatory failure that is not under conscious control. The child's nervous system is doing exactly what nervous systems are designed to do when a person does not feel safe. The behaviour is the alarm; it is not the fire. The grammar of safety Stephen Porges's polyvagal theory popularised the language. Bessel van der Kolk made it accessible. The practical insight is that humans — and especially young humans whose prefrontal regulation is still under construction — have a small repertoire of automatic responses to threat: mobilise, fight, flee, freeze, collapse. Each of these can look, from the outside, like a behavioural problem. A child who hits is mobilising. A child who runs away is fleeing. A child who goes still and refuses to speak is in freeze. A child who becomes oddly compliant after sustained stress may be in functional collapse. None of these responses is a choice in any sense that the everyday word implies. The behaviour is the alarm; it is not the fire. When we read these behaviours through a moral frame — as defiance, as manipulation, as bad character — we miss the question they are actually asking, which is am I safe here? What "safety" means in this context It does not only mean physical safety, though it includes that. It means: is my caregiver attuned to me right now? Are the rules of this space predictable? Is my body's state — hunger, tiredness, sensory overload — being noticed before it becomes catastrophic? Is the social field around me coherent? Children read these conditions continuously and below the level of words. They cannot, in early years, reflect on the reading. They can only act on it. What this means for the adult response If behaviour is a safety signal, the first task in a difficult moment is not to correct the behaviour but to be the safer presence the child cannot yet generate alone. This is not the same as permissiveness; it is closer to triage. You stabilise the regulatory state first, and you address the conduct afterwards, when there is somebody home to address it with. Discipline that ignores the safety question — that escalates volume, removes the relationship, or punishes the dysregulated state itself — tends to deepen the dysregulation it is trying to correct. The child learns not that the behaviour was wrong but that the very state of being overwhelmed is a state in which they will lose their adults. What it means for design Once you take the safety frame seriously, you start to see the design of children's environments differently. Classrooms that confuse compliance with regulation produce more behaviour, not less. Homes that have predictable rhythms produce children who can do hard things. And digital systems that respond to children — recommendation algorithms, AI tutors, social platforms — are now part of the safety field whether we have decided that or not. A child whose attention is being held by a system that is itself volatile, unpredictable, or arousing is in a low-grade activation state most of the time. Whether that counts as "safe" in the developmental sense is one of the questions I am working on now. The early answer appears to be: not quite, and not always, and the costs are showing up in the behaviour we are trying to fix. The reframe is the intervention The single most useful thing the adult world can do for children whose behaviour is hard is to stop asking how do I make this stop? and start asking what is this telling me? That reframe does not solve everything. But it is the door through which most of the actually-useful interventions arrive. Children are not adversaries. They are nervous systems learning, in a hurry, to be people. The behaviour is feedback. We should treat it as such. Further reading - Porges, S. W. (2011). The Polyvagal Theory: Neurophysiological Foundations of Emotions, Attachment, Communication, and Self-Regulation. New York: Norton. - van der Kolk, B. (2014). The Body Keeps the Score. New York: Viking. - Siegel, D. J., & Bryson, T. P. (2011). The Whole-Brain Child. New York: Delacorte. - Greene, R. W. (2008). Lost at School. New York: Scribner. ================================================================ JOURNAL: What Children Actually Need From Safety URL: https://drjorgecampos.com/journal/what-children-need-safety DATE: 2025-12-08 TAGS: adolescence, safety, policy, design ================================================================ ABSTRACT: Most adult conversations about keeping children safe end up being about restriction. But the developmental concept of safety is not primarily about removal of risk; it is about the consistent presence of conditions under which a young person can grow. This essay sketches what those conditions are, and why the politics of children's safety keeps misreading them. KEY CLAIMS: - Developmental safety is a positive set of conditions, not the mere absence of risk. - Predictability of caregivers is the single largest variable in children's regulatory development. - Restriction-based safety politics often increases the dysregulation it intends to reduce. - Digital and algorithmic environments now form part of children's regulatory field, whether we acknowledge that or not. - Designing for safety means designing for repair, not for the impossibility of harm. BODY: The public conversation about children's safety, especially around digital life, is overwhelmingly a conversation about what to remove. Block this app. Restrict that platform. Raise the age at which a child gets a phone. Impose stricter content filters. Require age verification. These are reasonable instruments at a population level, and I am not against them. But they share an assumption I think is mistaken: that safety is fundamentally a question of removal — taking dangerous things out of children's environments — and that once enough has been removed, what remains is, by default, safe. Developmental psychology has a different picture. Safety, in the sense that matters for a young person's growth, is not the absence of risk. It is the presence of certain conditions, sustained over time, without which the rest of development cannot happen well. What those conditions are The shortest list I would defend is four things. Predictable adults. A child does not need a perfect parent or teacher. They need an adult whose responses to them are predictable enough to model. The single most consequential variable in early regulatory development is whether the people around the child respond in roughly the same way today as they did yesterday — not necessarily warmly, not necessarily kindly, but coherently. This is the substrate on which everything else builds. Repair after rupture. Adults will misread, snap, withdraw, fail. The question is not whether ruptures occur — they always do — but whether they are repaired. Children who watch ruptures get repaired learn that conflict is survivable. Children who watch ruptures stay open learn that they are not. The capacity for repair is a more useful safety variable than the absence of conflict. A regulatory partner. Young children cannot self-regulate; they co-regulate. They borrow the nervous system of the adult next to them until their own is built. This is not a metaphor. It is a measurable physiological fact in the data on heart-rate variability, cortisol, and behavioural state. Older adolescents are still doing some version of this, just at a slower frequency and with peers as well as adults. Permission to feel large feelings without losing the relationship. A child can be angry, frightened, or overwhelmed without that being treated as a fault to be punished. This is the variable that, in my reading, the moralised conversation about behaviour gets most wrong. Children whose intensity is read as bad character learn to suppress signal, not produce calm. Why restriction-only safety fails Restriction is necessary at the margins. A child should not be on a platform optimised for adult attention. A teenager should not be exposed to content that adults could not metabolise. Some doors should be closed, full stop. But restriction, on its own, does not provide any of the four conditions above. A child who is removed from a dangerous platform but returned to a household whose adults are unpredictable is not, in the developmental sense, safer. They are just dangerous in a different room. And worse: the conversation about their safety has been resolved publicly, which can make the actual ongoing absence harder to see. This is the part of children's safety politics that I think is most often missed. Headlines about restriction generate the appearance of having done the work. The work — the slow, expensive, unglamorous work of being a predictable adult who repairs ruptures and tolerates large feelings — is what actually produces safe children, and it is the part that policy cannot do for you. What this means for digital systems Digital and algorithmic environments are now part of the regulatory field, whether we acknowledge that or not. A teenager whose attention is held by a recommendation system is, for the duration of that holding, taking some of the system's qualities into their own state. A predictable, calmly-paced system contributes one thing to the child's regulatory load. A volatile, novelty-maximising system contributes another. This is one of the questions my current research is examining. Early indications are that the systems most successful at maximising engagement are the systems that produce the most regulatory cost, and that this cost is heaviest precisely on the children with the least predictable adults at home — the children who could least afford it. The two failures compound. Designing for safety means designing for repair, not for the impossibility of harm. The implication The reframe I would press on the conversation is this: we cannot keep children safe by subtracting things from their environments. We can only keep children safe by adding the things their development requires. The two strategies look similar from the outside; they produce very different children. Restriction is the easier conversation to hold in public. The other one — the one about whether the adults in a child's life are present, attuned, repairable, and tolerant of large feelings — is the conversation that actually matters. We should be having it more. Further reading - Schore, A. N. (2003). Affect Regulation and the Repair of the Self. New York: Norton. - Tronick, E. (2007). The Neurobehavioral and Social-Emotional Development of Infants and Children. New York: Norton. - Hari, J. (2022). Stolen Focus. London: Bloomsbury. - Twenge, J. M. (2023). Generations. New York: Atria. ================================================================ JOURNAL: Living Inside Algorithms URL: https://drjorgecampos.com/journal/living-inside-algorithms DATE: 2025-12-01 TAGS: adolescence, ai, identity, algorithms ================================================================ ABSTRACT: What happens to a young person's sense of self when much of their daily experience is shaped by systems that learn from them in real time? This essay sketches an early answer: identity formation does not stop, but it changes shape, becoming more responsive to what the system rewards and less anchored in slow, internal coherence. KEY CLAIMS: - Algorithmic curation is a developmental condition, not just a media context. - Identity formation under continuous prediction tends toward responsiveness rather than coherence. - Adolescents inside these systems develop a sophisticated metacognition about being watched, but rarely the time to integrate it. - The cost is not loss of self; it is a self organised more around legibility to systems than around inward continuity. - Designers, parents, and clinicians need a developmental — not behavioural — frame for what these systems do. BODY: The first generation of young people who have spent the entirety of their adolescence inside algorithmically curated systems are now reaching their twenties. The data on what this has done to them is still arriving. The clinical reports, the longitudinal studies, the demographic surveys — they are all early, and most of them are arguing about the wrong question. The wrong question is whether screens harm young people. The right question, in my reading, is what shape does identity take when it forms inside a system that is itself learning from it? A developmental, not behavioural, frame Most of the public debate about young people and digital life is conducted in behavioural terms: hours of use, content categories, time-on-platform. These are tractable variables; they generate publishable studies; they produce headlines. But adolescence is not, primarily, a behavioural phenomenon. It is the developmental period during which a person assembles, from a vast and chaotic input, a working sense of who they are — what they care about, who they trust, what they will and will not do, where the boundaries of their inner life sit. This assembly is normally slow, recursive, and depends on long stretches of uninterrupted internal time. It depends, in particular, on the experience of the inner self being separable from the social self that has to be performed. The conditions of algorithmic life rearrange that assembly in ways that the behavioural frame does not capture. What changes The young people I interview, read, and have worked with consistently describe a distinct phenomenology. It is not that they have lost selves. It is that their selves have become more responsive — more shaped, in fine grain and in real time, by what the surrounding systems reward. A teenager who posts and watches the response, modifies the next post, watches that response, modifies the next, is doing something that previous generations also did with peers — calibrating self-presentation against social feedback. The difference is that the feedback is now produced by a hybrid of algorithmic ranking, partial human signal, and a dataset of millions of similar performances. It is faster, more granular, and harder to ignore. The feedback loop is tighter than the slower internal loop that used to do the integrating work. Identity formation under continuous prediction tends toward responsiveness rather than coherence. The result, in many of the cases I have looked at, is a self organised more around legibility — the property of being readable to people, systems, audiences — than around inward continuity. These young people are often acutely self-aware: they know they are performing; they know the algorithm is shaping what they see; they have a sophisticated metacognition about being watched. What they often lack is the slow internal time in which to do anything with that awareness. The cost is not what people think The popular framing of this is that young people are losing themselves in social media, becoming addicted, being deformed by their devices. That framing is not quite right and, more importantly, it is not useful — because it points the conversation at restriction, when restriction alone (as I've argued elsewhere) does not produce the developmental conditions a young person needs. The actual cost, in the developmental sense, is subtler. It is not loss of self. It is a self whose centre of gravity has moved outward. The person continues to function, often impressively. They report feeling fine, often genuinely. But the integration that adolescence is supposed to do — pulling disparate experiences into a working interior coherence — gets done less, or done in different terms, or left for later. What "later" looks like is something we will be finding out for the next two decades, as this cohort moves through adulthood. There are already early signs in the clinical and educational data that some of the integration work is being deferred to the early twenties, where it then has to be done in conditions much less suited to it than adolescence was. What designers, parents, and clinicians need The implication, I think, is that the people most responsible for young people's environments — those who design the systems, those who parent them, those who treat them clinically — need to operate with a developmental frame, not a behavioural one. The questions are not how much screen time? They are: what is this system asking of this young person's interior? What is it rewarding? What slow internal work is it making harder? What conditions outside it are doing the integration that it is making harder? These questions do not yield clean policy. They yield design decisions, parenting decisions, and clinical decisions, taken case by case, with developmental expertise as their grammar. It is unromantic work. It does not generate viral coverage. It is, however, the work that the next generation of young people is going to need us to do. Further reading - Turkle, S. (2011). Alone Together. New York: Basic Books. - Twenge, J. M. (2023). Generations. New York: Atria. - Haidt, J. (2024). The Anxious Generation. New York: Penguin. - Schüll, N. D. (2012). Addiction by Design. Princeton: Princeton University Press. ## Ensaios em português Ainda não há ensaios publicados em português. Índice: https://drjorgecampos.com/pt/journal