A college student’s reluctance to use artificial intelligence might seem, at first glance, like ordinary technophobia—the reflexive resistance that greets most disruptive technologies. But her objections, raised in conversation, are more carefully considered than that framing suggests. They touch on two distinct concerns: one broadly ecological, the other deeply personal. Together, they illuminate tensions that institutions, industries, and individuals are only beginning to reckon with.

The Resource Question

The environmental costs of AI are not disputed, though they are frequently underappreciated. A single ChatGPT query consumes roughly ten times the electricity of a standard Google search—a differential that, multiplied across billions of interactions, produces consequences of industrial scale. Global data center electricity consumption is projected to rise sharply in coming years, with some estimates suggesting the increase would be equivalent to adding a country the size of Germany to the world’s power grid. In the United States alone, data centers may soon account for six percent of total electricity consumption—comparable to all household lighting and refrigeration combined.

Water tells a similar story. The familiar saying that it takes a bottle of water to complete an AI query captures something real: both data centers and the power plants that supply them require substantial water for cooling. Between 2022 and 2023, Google’s water consumption rose seventeen percent, from 23 billion liters to figures approaching the output of a small municipality; Microsoft’s rose twenty percent over the same period. Both companies attribute the increases substantially to AI infrastructure. Amazon, notably, does not disclose its water consumption figures.

This trajectory is not without precedent. The nineteenth century marked a decisive turn in humanity’s relationship with energy, when coal and steam began powering tools that human muscle once drove. Each successive wave of technology has generally consumed more resources upon introduction, before competitive pressures eventually drive efficiency gains. The pattern holds, though with an important caveat: aggregate consumption tends to keep rising even as per-unit costs fall. AI appears to follow this curve—only steeper.

The optimists point to fusion energy, where experimental fusion reactors in the United States and Europe have produced encouraging hopes. Fusion I sthe source of energy of the sun, and its promise if clean, renewable energy. Projects in Europe, UK, US, and China are underway, but none has proven it will work in a known quantifiable time. Microsoft has invested in at least one company promising commercial fusion power, so has Google. Whether that promise materializes in time to offset AI’s near-term resource demands is, at present, an open question.

The Personal Reckoning

The student’s second objection is harder to resolve with data. If she offloads intellectual work to an AI system, will she learn less? Will the result still be meaningfully hers?

These are not frivolous concerns, particularly in a university context where the act of doing and the process of learning are understood to be inseparable. The production of an academic paper involves not only its conclusions but the intellectual labor that generates them—the dead ends, the reconsidered arguments, the gradual mastery of a subject. When AI absorbs that labor, something is lost even if the output looks the same.

There is a specific limitation worth naming here. When AI conducts a literature review, it does not return an exhaustive catalog of relevant sources; it performs its own selection, according to processes it does not fully expose. What is absent is traceability—the methodological transparency that allows a researcher to evaluate how conclusions were reached and decide whether to trust them. In both the sciences and the humanities, methodology is not incidental to knowledge; it is constitutive of it. An AI that cannot show its work is, in an educational context, an unreliable collaborator.

How much AI should be allowed in the production of a paper in School? This is not just the issue of how much learning has been demonstrated; it is also an issue of the pride we lose when our work is made by a machine. This discomfort has a name in philosophical literature: Promethean shame. The term describes the unease that arises when a human-made machine produces results that a human being could not readily match. Consider the artisan shoemaker at the dawn of industrialization—a craftsman who took pride in fitting shoes to individual customers, who understood leather and last and the particular demands of a given foot. He would have watched, with feelings not easily named, as machines began producing footwear by the million, in more colors, more sizes, and more quickly than he ever could. The Olympic movement encodes a version of this instinct: competitions exclude performance-enhancing technologies precisely to preserve a domain in which human beings contend only with each other.

Navigating the Distinction

What practical guidance follows from all this?

One useful framework distinguishes between AI as a crutch and AI as an amplifier. The crutch replaces a capacity the student has not yet developed; the amplifier extends a capacity the student already possesses. The distinction is not always obvious, but it is real.

The pocket calculator offers an instructive analogy. It is the wrong tool for a third-grader still learning multiplication tables—not because calculators are bad, but because the child needs to internalize arithmetic before technology can legitimately substitute for it. Once that foundation is established, the calculator frees the student to engage with more complex problems. The tool serves learning rather than circumventing it.

Entry-level computer science courses at Stanford and other institutions have long used a variation of this principle: “Karel the Robot,” a pen-and-paper exercise in which students give written instructions to an imaginary robot navigating a grid. The point is not that computers are unavailable. The point is that understanding must precede automation.

For college students—particularly those in later years, with developed critical faculties—the question of whether they are using AI as crutch or amplifier is one they are, in principle, equipped to answer honestly. The student who can identify which cognitive functions she is outsourcing, and evaluate whether she has already internalized them, is in a position to make a legitimate choice. The student who cannot make that distinction is probably not yet ready to delegate.

The shoemaker’s dilemma does not resolve neatly. There may come a point at which AI can perform certain intellectual tasks so thoroughly that human mastery of them becomes, economically speaking, unnecessary—as machine production rendered artisan shoemaking a niche rather than a trade. But intellectual craft is not entirely analogous to physical production. Shoes can be inspected; the outputs of AI reasoning often cannot be verified without the very understanding that AI threatens to displace. That asymmetry is a reason to continue developing the underlying capacities, even as the tools that might supplement them grow more capable.

The student who said no to AI may not be right in every application. But her instinct—that something important is at stake in the question of what we choose to do ourselves—is worth taking seriously.

Share

Leave a comment

We now have verdicts in two of the five ongoing trials involving social media companies. In both cases — involving Instagram and YouTube — the platforms were found liable for causing harm to children or for failing to disclose known dangers. Fifteen years ago, these same companies were celebrated as the key to humanity’s future. Today, they are following the path of the tobacco industry.

As significant as these rulings are, they are not the end of the road — and they may be modified on appeal. The verdicts confirm what most parents have felt for years. But they do not solve the problems our children are facing.

Technology evolves. AI may soon replace social media as the dominant force in children’s lives, bringing with it a new set of risks: cognitive risks in how children learn to think; behavioral risks when a child begins to regard a bot as a friend; and privacy risks when personal thoughts and feelings are shared with a machine. What will not change is technology’s effect on relationships — between children and their parents, and between peers. These challenges will remain, whatever the courts decide.

A Changed Landscape

Over the past twenty years, the environment in which children grow up has shifted considerably — and not only because of technology. Schools have narrowed their focus, increasingly measuring success through a handful of statistics that reduce education to what can be tested. Parenting has changed too. Parents are now handed an ever-expanding list of decisions to make about their children’s lives — decisions that are often made in the dark, without a clear understanding of what children actually need at each stage of their development.

Because parents are given the choices, they also absorb the blame when things go wrong. The result is predictable: parents say that raising children today is significantly harder than it was when they were growing up. Parents report feeling stressed — and so do their children.

Why This Matters

When we are forced to operate without adequate knowledge, we leave outcomes to chance and carry the weight of consequences we didn’t fully understand. The social media trials have shed light on decisions made by tech companies that prioritized their companies’ growth over children’s safety. But children have needs of their own – safety, yes, but also the freedom to grow, and that growth is about far more than grades.

It is about building the human capacities that truly matter in life: the skills that help us navigate school, relationships, adversity, and an uncertain future. These are the qualities that give us our distinctly human advantage — over technology, over disruption, and yes, over AI.

Take resilience, for example. It has a well-documented positive relationship with mental well-being. Resilience is not a cure for anxiety — it is a capacity that allows us to move through anxiety without becoming stuck or overwhelmed.

Yet it is nearly impossible for parents and educators to develop these capacities in children within the current framework most of us operate in. Human capacities like resilience, creativity, and social intelligence are rarely measured by grades or tests, so they go largely unseen. They cannot be built by attending a class or joining a club. And yet, in the long run, they are precisely what helps children succeed — in school and beyond.

A Different Way of Looking at This

Education is always future-oriented. The decisions we make today are meant to prepare children for the next stage of their lives — and, ultimately, to shape the adults they will become. No one can claim to know exactly what the future holds. But we can prepare children by giving them the strongest possible human foundation. Some capacities are timeless.

We may not know the future state of the world, but we do know the current needs of children at every stage of their development — and we know the effects of technology on those needs.

Over the past twenty years, my leadership roles at schools in Silicon Valley that made a deliberate choice to operate without technology gave me a close and sustained view of what tech does to children. The trials in Los Angeles and elsewhere address what can be argued in court: negligent conduct, failure to warn, misleading claims. But the deeper issue is this: technology has accelerated the erosion of relationships — between children and their parents, and between peers. These relationships are not incidental. They are the very mechanism through which a child becomes a functioning, grounded adult.

At the schools I led, we invested enormous energy in building children’s social environments. When technology was present and mismanaged, it consistently got in the way. That is one of the reasons we decided to eschew it.

We gained knowledge about the complex effects of technology on child development. The question became: how do we make this knowledge available to other parents?

A Better Frame

Think of technology the way we think of carbs. Consumed as part of a balanced diet, carbs play a legitimate role — we need some of them to function. The problem arises when carbs begin to replace the meal. It provides a temporary sense of satisfaction, but the body knows something is missing and asks for more. We eat more carbs, and we find ourselves in a cycle of highs and lows, never truly nourished.

Technology works the same way. What it can never fully deliver is the experience of healthy, complete human relationships. And that is precisely what children are desperately seeking — especially in adolescence, when relationships become the primary engine of growth and identity.

If we can provide children with genuine connection and fulfillment outside of technology, the phone becomes easier to put down. Not because it was taken away — but because something better is available.

Good Intentions Are Not Enough

There is no shortage of resources online offering screen time management tips or behavioral strategies for parents. These provide temporary relief, but they do not give parents the deeper knowledge they need to act with real confidence and autonomy.

Every child is different. Every family environment is different. What parents need is not a one-size-fits-all prescription, but the understanding to adapt thoughtfully to their own situation.

Social Emotional Learning (SEL) programs in schools have shown limited results — not because the goal is wrong, but because this kind of development cannot be confined to a classroom. It must happen continuously, through trusted relationships with adults who know the child. Even the best SEL programs rarely extend beyond the school walls.

A tech-free childhood is often perceived as a luxury available only to some. Our goal is to ensure that all parents — regardless of the path they choose — have enough knowledge about their child’s developmental needs to make genuinely informed decisions.

Real change also requires community. The interventions that work are built around the construction and maintenance of in-person social groups — and eventually, groups that teenagers help build themselves. Different kinds of relationships serve different needs. It is the village teenagers need, and where growth actually happens.

What We Are Building

We are developing a platform for parents and educators.

Based on the age of each child, parents will have access to a focused curriculum of short videos — each one explaining the developmental needs of children at that stage, the specific effects of technology, and the practical tools parents can use to respond wisely. This is not a five-step program or a magic formula. It is deep, grounded knowledge about children, technology, and human development — designed to help parents make sense of what is happening and make choices they feel good about.

The platform will support both individual access and group access. For groups of five or more parents, we will also offer live sessions — creating the small, connected communities that children need, and that parents need too.

Who is Behind This

I am joined in this work by my wife, Dr. Monica Laurent, who has spent more than twenty years in education — including guiding two cohorts of students and their families from first through eighth grade, teaching high schoolers and adults, and mentoring teachers. Her most recent academic research focuses on the social environments of teenagers.

We will begin with tweens — a group that is both overlooked and underserved in most existing resources on this topic.

In the coming weeks, we will share more details about the platform’s content, structure, and how you can get involved.

Stay close. What comes next matters.

To receive future announcements:

Subscribe now

Know someone who could benefit?

Share

Leave a comment

The early signs are hopeful, though they also reveal the limits of a mixed environment — one where phones are banned at school but freely available at home. It comes as no surprise that teachers report students are more present, making more eye contact and engaging more readily. Students are also more likely to participate in after-school activities. At one Florida school, grades improved slightly after years of remaining flat.

Strict no-phone policies tend to produce an initial spike in discipline cases, followed by a sharp decline compared to the year before — including fewer incidents of cyberbullying. A phone-free environment clearly benefits children during the school day.

I visited one school that had banned cellphones but was trying to avoid the disciplinary fallout that often follows. After eight months, the ban was still not fully enforced. Sitting in a classroom, I watched students scroll through newsfeeds under their desks. One teacher put it plainly: “If the principal shows up, phones disappear. They come back when he leaves.” Enforcement, however, does not have to be heavy-handed. At the school I managed, a phone found in violation of the policy was simply held at the front desk for the rest of the day — no drama, no permanent record.

Although early academic results look promising, bans have not yet produced a measurable decrease in depression or anxiety. More time may be needed for that effect to appear — but in places where school phone bans have been in place for years, without parallel restrictions at home, there is little evidence so far that the decline in student mental health has been reversed.

The Challenge at Home

Phones remain accessible outside school, and many parents still don’t know what to do. No US law limits the ownership or capabilities of devices for children, so parents are largely left to navigate this alone. Like so many aspects of modern parenting, they are asked to make consequential choices with little information or guidance — and blamed when things go wrong. That is a recipe for parental stress.

There are a handful of approaches available at home: delaying phone access until high school, implementing selective bans, providing stripped-down devices, or doing nothing.

Several movements in the US advocate for waiting until high school before giving a child a smartphone. Full disclosure: that is the path my wife Monica and I chose for our own children. At the time, we were interviewed by CBS about it — you can watch that conversation at the end of this post. When each of our children did eventually receive a phone, I practiced what I call a selective ban: a set of rules that restricts phone use by time and place.

The selective ban starts with a conversation — working with teenagers to think clearly about what a phone actually is. There is an important distinction between a phone as a productivity tool (something that helps you get things done) and a phone as entertainment. The second lesson is that every technology comes with trade-offs, even when it genuinely helps in other ways. The cellphone is a perfect illustration of this paradox: the telephone was originally conceived as a way to extend human connection; its modern descendant often undermines it. The selective ban creates clear boundaries — specific times and places where the phone is put away. In the early days, a simple and effective rule was this: Cell-phone are for use outside the home.

“Do as I say, not as I do” has never worked with teenagers — and it won’t work here. For the selective ban to mean anything, parents have to follow it too. For many, that turns out to be harder than expected. We like to think of technology as a set of tools, but we become dependent on them, and uneasy without them. Clothing — originally a technology to protect us from the cold — is something we now cannot imagine living without. For better or worse, the phone has become that for many parents. Restricting use to certain places and times can work well, but it is only effective if the adults in the home are willing to go first.

Several companies now offer phones with limited features built for children — from fully locked-down devices with no internet access to phones where parents can selectively enable or disable specific services.

Asking the Right Questions

We should always pause and ask ourselves why we are giving a child a phone. The most common answers are safety, surveillance, and convenience — and each deserves a harder look.

The safety argument has real limits. A phone that can be dropped, broken, or drained of battery is not a dependable safety net. True safety means something broader: equipping teenagers with the judgment and skills to manage when the device isn’t there — knowing where to go, and how to ask for help.

On surveillance: the majority of tools marketed to parents for managing their children’s phone use are surveillance-based — monitoring screen time, tracking browsing history, or providing real-time GPS location. This casts the parent in the role of police officers, which works against the relationship we are trying to build — one that should be moving toward trust, not suspicion. Many popular apps, including WhatsApp and Signal, are encrypted, making monitoring difficult in any case. Resourceful teenagers quickly learn to use multiple accounts or VPNs to stay one step ahead.

Constant location tracking risks turning a parent into an investigator — at exactly the moment when the relationship needs to be shifting toward something more like a trusted advisor. It is worth asking yourself honestly: At what age would I actually stop? Teenagers have a genuine need for privacy — a mental and physical space of their own, where they can think and act with some degree of autonomy. Constant surveillance cuts against that need and can convey something damaging: that they are not trusted.

Adolescents need to feel that they are trusted, and that they are capable of navigating difficulty on their own. Psychologists have long recognized that privacy plays a central role in adolescent identity formation. Excessive monitoring tends to produce the opposite of what parents intend — more secrecy, less willingness to ask for help, and a gradual erosion of the parent-teen relationship.

What we call addiction is often a search for what’s missing

Most approaches to the phone problem have focused on the device itself — more access or less, tracked by screen time. But what if that is the wrong frame entirely? What are teenagers actually missing — what are they desperately seeking — when they reach for their phones? Addressing that need outside the phone ecosystem may do far more to reduce dependency and improve well-being than any restriction on its own.

Teenagers are in the active, urgent work of figuring out who they are — and they do it socially. Who we are is in part defined by how we see others — and by how others see us. Adolescents develop in the space between the self and the group. A crucial part of forming an identity is discovering one’s usefulness to others: finding where and how you matter. The real work, then, is to provide that sense of belonging and meaning in the physical world — because that is ultimately where it takes root.

As parents, we shape home life and make the countless decisions that define how children grow up. Creating a genuinely nourishing environment does not happen by accident. In the teenage years, the parental role shifts — from director to advisor, from enforcer to someone who gradually, intentionally loosens the reins. To build the conditions in which teenagers can truly flourish, we need to understand what our children are actually looking for — and work alongside others to help provide it.

The phone will not give them what they need. But we can.

Share

Leave a comment

The start of middle school often feels like a significant transition — for parents, perhaps even more than for the children themselves. We sense that adolescence is arriving, that the academic and social stakes are rising, and that something important is shifting in our relationship with our child. Yet this period is frequently underestimated. Middle schoolers are no longer the young children who looked up to us instinctively, but they have not yet developed the full capabilities of a high schooler. They occupy a particular and often overlooked in-between. In that time, parents face a double peril. One is a trap set by nature, the other by the peculiar way we have set up parenting.

Teenagers rising

Neuroscientists broadly agree that the behavioral changes marking the onset of adolescence are primarily driven by a surge in hormones — a biochemical tide that triggers profound shifts in the developing brain. This hormonal rise acts as a kind of switch, setting adolescent behavior in motion.

It is why teenagers become more emotional, begin seeking novelty, and grow increasingly attuned to peer approval. Having spent their childhood building security within the family, they now feel the pull of a wider social world. They turn away from their parents and toward their peers — and in doing so, a new generation begins to find its shape.

Yet while this social awakening unfolds, another process is quietly underway. The prefrontal cortex — the region of the brain just behind the forehead, responsible for self-regulation and impulse control — remains far from fully developed. This is why a middle schooler can appear calm and composed one moment, then seemingly out of control the next. It is not defiance; it is neuroscience. Growth does not happen overnight. The adolescent brain is slowly acquiring new capacities: richer language, more sophisticated abstract thought, and the early contours of a defined identity.

To understand this phase, it helps to picture teenagers as being on a multi-year quest — deeply personal work — to discover who they are and what they will contribute to the world. Along the way, they will try on different personas, different hats, testing how each one fits and how the world around them responds.

They are also learning, for the first time, how to exist within a peer network. Adolescence is, in part, a rehearsal for adult life. From the web of peer relationships will eventually come some of life’s most consequential opportunities: lifelong friendships, romantic partnerships, and professional paths yet unimagined.

Their primary relationships up to this point have been with their parents, where love is largely unconditional. Peer relationships operate by different rules. There is no guaranteed acceptance, no automatic forgiveness. Learning to navigate that world is difficult and sometimes painful — but it is essential, and it carries its own joys too.

This slow, almost two-step development of the brain is the first challenge that parents face during the beginning of adolescence. It cannot be solved and closed overnight and is best understood as a journey.

Subscribe now

The Hero’s Journey

A teenager’s quest can be understood as the opening chapter of the story of their life. Before adolescence, parents were the primary authors of that story. Now, teenagers are beginning to take the pen. They are becoming the heroes of their own narratives.

Think of Luke Skywalker in Star Wars. Luke faces two kinds of obstacles. The first is external: the formal challenges placed in his path — the equivalent, in a teenager’s life, of homework, expectations, and the demands of school. The second is internal: Luke is impatient, reckless, and prone to emotional reactions. He has to work on both fronts. And crucially, he has to do that work himself. The internal obstacle also has a social dimension — the title of Jedi Knight, which Luke seeks, is not simply a personal achievement; it is a recognition conferred by a community.

Luke is not entirely alone, however. Obi-Wan is his guide: a moral compass, a cautionary voice, a quiet presence. Obi-Wan is also the link between generations. This is the role available to parents. Obi-Wan does not command Luke, and he does not step in to fight Luke’s battles — if he did, Star Wars would be Obi-Wan’s story, not Luke’s. Eventually, Luke must choose between the path Obi-Wan has made possible and the narrow, darker path offered by Darth Vader, his biological father. That choice belongs to Luke alone.

Early in their journey, teenagers often try to conform — to be more like everyone else. This is partly a carryover from childhood, when imitating the adults around them was how they learned to be safe. But the stakes are different now, and they quickly find that conformity does not give them a distinct identity. They must forge something of their own within the group. Adults cannot do this for them. What we can do is create the structure that makes the quest possible — set the stage, without becoming the actors on it.

The Pressures of Society

The second challenge modern parents face is the peculiar way society has defined their role and the social pressures they receive.

In my Star Wars analogy, Darth Vader represents the parent who has been overtaken by external pressures. The weight of fate and the demands of the Empire have hollowed him out; he no longer acts from his own values, but on behalf of a system. As parents, we are not immune to this. Society places enormous pressure on us to produce visible, measurable success in our children.

Unlike families in many other developed countries, many parents feel they have limited social safety nets to fall back on. We have come to treat education as a commodity — something we can select, optimize, and evaluate. Schools publish grades and rankings, and parents rely on this narrow data to assess how their child is doing. Data creates urgency.

All of these urgent choices, these possibilities, without the benefit of a social structure to help parents, tell us we are responsible, that the choices are ours, and that the outcomes will reflect those choices. Under this pressure, well-meaning parents can find themselves becoming intrusive — an interfering version of Obi-Wan who removes every obstacle, denying their child the chance to write their own story.

Parents have been given agency, the power to choose without the proper information or knowledge. This is illusory autonomy. It cannot generate good outcomes. Stuck between the urge to do right by our kids, the need to show that we are good parents, and the uncertainty about the future to which we are preparing our kids, parents become anxious, and their kids too.

The anxiety deepens when we hear other parents at social gatherings boasting about their children’s achievements. A parent came to my office once asking for help. They understood, in principle, the importance of giving their child space to find their own way, but they did not know how to explain that philosophy to their in-laws. The family pressure was intense: in the in-laws’ view, there were only a handful of acceptable paths to success — doctor, engineer (preferably IT), or lawyer — and the planning for that path had to begin in middle school. Of course, development does not work that way. Many families who have followed this rigid approach find, somewhere along the road, that their child has lost the joy of learning and the motivation that sustains it — or that they reach the destination only to discover it was never theirs.

“Do not let anybody fast-track you into their narrow definition of success.”

I kept that sentence on a board in my office for years, rewriting it whenever it faded. I referred to it often in conversations with parents, and I used it at graduation to remind students that in college they would encounter people with very fixed ideas about what a worthwhile life looks like. They should not feel compelled to follow a path that is not theirs.

Share

Middle schoolers are at the very beginning of their personal journey. This is when they start to learn how to travel — by traveling.

Structure Is Not the Enemy

None of this is an argument for stepping back entirely and leaving teenagers without structure or guidance. The temptation to become invisible — to simply let young people find their own path — misunderstands what they actually need.

In Star Wars, Obi-Wan dies. Symbolically, his death frees Luke to grow. But even beyond death, he continues to provide structure: a guiding voice, a moral compass, a presence that shapes Luke’s choices without dictating them. This is the transition we are asked to make starting in the middle school years — moving from the parent as imitator-model in early childhood to the parent as guide, someone who helps children reflect rather than simply instructing them what to do.

Part of that transition requires us to examine our own interior life. What are our fears for our children, and what are our hopes? To what degree are those fears and hopes genuinely ours, and to what degree have they been handed to us by the culture we live in? How do we feel when other parents, at a dinner table or a school event, begin talking about their children’s achievements? When I find myself in that situation, I have come to rely on a simple answer: my goal in raising my children is that when they come back for a visit at thirty, they say, “Thank you — you made a difference in my life, and I am happy.” I cannot predict what happiness will look like for them fifteen or twenty years from now. But I can build an environment that gives them the inner tools to find it.

Recognition and Community

Our society offers teenagers very little formal recognition beyond the accumulation of grades and credentials. Luke Skywalker faced something similar. At the start of his story, the old order of the Jedi has collapsed. The recognition Luke ultimately earns comes not from an institution but from his service to a community — the Rebellion, with its close friendships and its broader cause.

Teenagers need the same thing. Who a teenager becomes is not defined by a title or a grade point average. It is earned through relationships and through service. Parents cannot manufacture this alone. It requires access to a community where a young person’s contribution can be seen and genuinely valued. Schools can serve this function through group work, collaborative projects, and shared endeavors — but only if they are careful to avoid what might be called the stardom fallacy: the tendency to spotlight a handful of exceptional students while the rest go unrecognized.

These experiences also need to happen in person. The learning that comes from real social encounter — from being seen, from repairing a friendship, from receiving a genuine thank-you from someone standing in front of you — cannot be replicated on a screen. An emoji or a text message is simply not the same thing.

Letting Them Own Their Story

Raising teenagers well begins with a kind of gift: returning ownership of their story to them. Whatever successes or setbacks they encounter, those experiences belong to them, not to us. Taking social credit for a child’s achievements is a subtle form of theft. The parental shift in middle school is to channel our inner Obi-Wan — to offer a moral compass and open guidance, while resisting the urge to clear the path for them. Our job is to ensure that a community surrounds them, one in which they can belong, contribute, and earn recognition through their own effort.

The long-term reward is not only a healthy, happy, and capable adult. It is also the relationship itself. When, like Obi-Wan, we make the quiet sacrifice of watching them walk away toward their own life — across a college lawn, or wherever their path leads — the relationship does not end. It transforms. And that transformation is a gift to both parent and child.

Leave a comment

Yet these bans enjoy broad public support. The Ipsos Education Monitor surveyed parents in 30 countries and found that 74 percent favor some level of restrictions. Other polling organizations report similar findings. The Economist—writing in February 2026—noted with dry humor that bans have become “more popular than ice cream.”

In conversations with parents, three concerns recur: exposure to inappropriate content, fear of addiction, and bullying. When a child becomes so absorbed in a screen that ordinary life—meals, homework, conversation—loses its appeal, parents grow alarmed. Often, however, they do not act until there is external, measurable evidence of harm: falling grades, a teacher’s call.

The Plastic Figurine Lesson

When I worked in schools, a handful of students arrived one morning with small plastic figurines clipped to their clothes and backpacks. Within days, others followed. A trading economy took hold. Entire classes became absorbed in it; the figurines were the only topic of conversation. Teachers struggled to redirect students during recess, and it became evident that the preoccupation had colonized class time as well.

Whether by design or accident of the supply chain, certain figurines were harder to obtain than others. A social hierarchy formed around scarcity. Teachers first tried restricting display on clothing and bags. When that failed to break the spell, the only effective remedy was a campus-wide ban.

Parents raised no objections. Within a week of the ban, students began returning to themselves. Gradually, the school’s social life regained its natural fluidity.

Many people expect a social media ban to work in exactly this way. After an awkward transition—a period of uncertainty about what to do without the feed—children would recover their equilibrium, returning to something like the social environment that existed before smartphones.

Share

The Case Against Bans—and Its Limits

That hope faces several main objections, the more prominent of which was articulated by The Economist in its February 2026 issue, headlined “Why Social Media Bans Won’t Work.” The magazine advanced two connected claims: first, that determined teenagers will circumvent bans through VPNs and weak age-verification systems, rendering enforcement hollow; second, that children are better served by early exposure to ubiquitous digital platforms, learning to navigate them rather than being shielded from them.

The second argument deserves particular scrutiny. The logic—that early training produces better outcomes—has a surface plausibility but collapses under pressure. Taken to its conclusion, it would justify handing adolescents car keys as soon as their feet could reach the pedals, on the theory that earlier practice yields more experienced drivers. We do not accept that reasoning for driving, or for alcohol, or for a range of other domains where developmental readiness is the operative criterion.

Both arguments rest on an implicit premise: that teenagers are essentially miniature adults, differing from their elders only in accumulated experience. On this view, withholding experience delays maturation and is itself a form of harm. The premise is empirically incorrect.

Children Are Not Miniature Adults

Developmental neuroscience is clear on this point. The adolescent brain differs structurally and functionally from the adult brain. The prefrontal regions governing impulse control and self-regulation are among the last to mature, completing their development well into the mid-twenties. This is not a minor qualification; it is the central fact about teenage cognition.

More critically for the social media debate, adolescence is the developmental period during which children learn—for the first time, in a sustained and consequential way—how to navigate peer relationships. They are acquiring the capacity for social belonging, for reading others, for managing emotion in real-time social contexts. Social media may not simply fail to support this process; it may actively impede it.

Consider what a screen mediates away. A pause in conversation. A shift in breathing. The position of someone’s body at a table. Adults who have spent decades in face-to-face relationships have built an intuitive repertoire for reading these signals—what researchers call the “thin-slice” social information that flows continuously in unmediated interaction. Teenagers are still building that repertoire. The screen filters much of this signal, providing a partial, edited image of the other person. In 2011, at the onset of the cell phone and social media era, MIT researcher Sherry Turkle, in a seminal book “Alone Together”, based on teenager interviews, shows how screen-mediated interactions allow teenagers to edit themselves and the consequences. Learning empathy through a phone is harder than learning it in a room.

This is why The Economist missed the mark. The magazine framed the ban debate largely as a consumer-protection question—can authorities prevent determined minors from accessing a product? And shouldn’t they train on the tech But the more important question is developmental: what does extended social media use displace during a period when the brain is actively wiring its social capacities?

The Tyranny of the number 168

Every person, regardless of age, has 24 X 7, or 168 hours in a week. After sleep, meals, school, and basic hygiene, the remaining hours are finite. Screen time does not exist alongside other activities; it displaces them. If YouTube is included in the category of social media—a reasonable classification given how it functions as a recommendation-driven feed—a substantial share of the time American children spend on screens is time not spent in physical, face-to-face interaction with peers. This is measured in national Time-of-Use surveys

The early wave of screen adoption in the 1990s and 2000s was greeted with enthusiasm. A teacher who arranged an email exchange between students on opposite sides of the world received national attention. The promise was connection—across distance, across difference. That promise was real, and in some contexts it still is. But the question that parents and educators are now asking is a more local one: are our children connecting adequately with the people physically around them?

The evidence suggests, in many cases, no. Time-of-Use Studies consistently show that adolescents are spending less time in face-to-face social interaction than previous generations did at the same age. Taking a popular mode of exchange as an example, texting introduces latency—a delay that reduces spontaneity and removes the real-time pressure of social exchange. Yet that pressure is precisely what teenagers need to practice managing. Video calls cut off substantial portions of the body, narrowing the field of social signals available. Real life is unedited, unfiltered, and often uncomfortable. It is also, for that reason, the environment in which the relevant learning occurs.

The Causation Question and Its Limits

A landmark trial currently unfolding in Los Angeles has brought the causation debate into sharp public focus. The case was brought by a California woman identified as K.G.M., who alleges that heavy use of Instagram and YouTube from early childhood led to depressive and suicidal ideation. K.G.M. contends that the platforms were deliberately engineered to exploit the psychological vulnerabilities of young users; her mother had attempted to restrict access without success.

One of the defense’s central arguments is that no causal link between social media use and psychological harm has been established—only a correlation. The defense has invoked a familiar statistical illustration: the correlation between children’s shoe size and their reading ability. Both increase as children grow older, but shoe size does not cause improved literacy. Other whimsical examples serve the same rhetorical purpose: the correlation between a country’s Nobel laureate count and its per capita chocolate consumption, or the number of IKEA stores it hosts.

The argument is logically sound in a narrow technical sense and important for scientific standards and the courts. But for those of us responsible for the daily welfare of children, it is the wrong frame and it is not helpful.

A Practitioner’s Framework

Parents and educators cannot wait for definitive experimental proof before making consequential decisions. A fourteen or thirteen-year-old will not be that age next year. They need a working framework, now.

That framework already exists. Developmental psychology and pediatric neuroscience have produced a well-validated model of how children mature. Different capacities emerge at predictable ages. A newborn lacks binocular depth perception; it develops in time for the child to need it. Adolescents see the emergence of higher-order language, emotional regulation, and—critically—the capacity for complex peer relationships. Each new capacity opens new developmental needs.

From this model, educators and psychologists can reason prospectively: when the environment changes significantly—say, when a child spends four hours per day on a social media feed—does that change support or undermine the developmental tasks the child is currently navigating? The answer does not require a randomized controlled trial to be actionable. A well-supported theoretical expectation, drawn from a validated developmental model and corroborated by observable outcomes, is a legitimate basis for precautionary intervention.

On that basis, the developmental case against heavy adolescent social media use is compelling. The principal concern is not, as most parents assume, harmful content—though that concern is legitimate. The deeper concern is that social media restructures the social environment itself during the period when teenagers most need unmediated human contact to develop their emotional and social capacities.

Will Bans Be Enough?

When I was visiting schools in 2022 in connection with the documentary film #KidsOnTech, parents frequently raised the question of bans—for social media specifically, or for cellphones more broadly. A common concern was social compulsion: if other children had access, their own child would be forced to participate or risk social exclusion. The ban, in this framing, is not simply a household rule but a collective action problem requiring a policy solution.

School cellphone bans do recreate a space in which children can interact without the mediation of devices. That is valuable. But the effectiveness of those bans varies considerably depending on context—and this is where the United States faces a structural challenge that other countries do not.

American middle and high school students are operating in a more tightly scheduled, instruction-dense environment than their European counterparts, with significantly less unstructured time. In elementary schools, recess periods have shortened. At the secondary level, lunch periods in many US schools run to 20 minutes—highly supervised, organized in rotating shifts to maximize use of cafeteria space.

European school systems structure the day differently. German schools break instruction with 10-to-20-minute intervals between lessons; lunch runs to 45 minutes. In France and Spain, midday breaks extend to two hours, treated as a necessary recovery and social interval, not as overhead to be minimized. Both systems produce similar results on the PISA assessment—the international test measuring 15-year-olds in mathematics, science, and reading—suggesting that the longer unstructured time does not come at an academic cost.

The structural differences extend beyond the school day. European cities and towns generally offer dense public transit networks that allow teenagers to meet in person independently—without requiring a parent to serve as a driver. In the United States, school choice policies mean that many students no longer attend their neighborhood school; their peers are dispersed across a wider geography, and independent in-person socializing requires logistical coordination that depends on adult availability.

This matters for the ban debate because a ban removes a competing activity without necessarily restoring the alternative. If teenagers in European cities put down their phones on a weekday afternoon, they can meet at a café or a park, independently. If teenagers in many American suburbs do the same, the practical options are narrower, often a parking lot. Banning the screen or social media, although a necessary step, does not automatically fill the resulting time with face-to-face social interaction if the infrastructure for that interaction is absent.

Necessary but Not Sufficient

Social media bans for children are worth pursuing. The developmental case for them is well-grounded, and the standard objections—that teenagers will circumvent them, or that early exposure builds digital competence—do not survive scrutiny. The more substantive concern is not whether bans work in principle, but whether they work in practice without the complementary conditions that make them effective.

In the American context, this requires a degree of intentionality that policy alone cannot supply. Bans address one side of the equation—reducing screen time and exposure—but the law of 168 hours means that the displaced time must go somewhere. If that somewhere is structured, adult-supervised, and screen-free but socially thin, or time alone at home, the developmental gain is limited. What teenagers need during adolescence is not merely less social media; it is more genuine, unmediated, peer-directed social life.

For parents and educators operating within existing American constraints—compressed school days, dispersed communities, limited public transit—this means being deliberate about creating and protecting space for the kinds of adolescent social interaction that do the developmental work. The plastic figurines were eventually removed from campus, and the children returned to themselves. But the school had a campus—a common space, with time, and with each other. That remains the indispensable ingredient.

A future post will explore what those social conditions look like in practice, and what parents and schools can do to cultivate them.

Hypervigilance and constant worries about our children’s safety may consume us, yet they do not truly benefit our children. Physical safety alone does not cultivate what children need most: a genuine sense of security. Building that sense of security requires more than simply surrounding them with a protective bubble. What we perceive as safety is not necessarily what they seek or need.

Seeking inner peace

When children do not feel safe and secure, they expend energy trying to protect themselves from unknown dangers. Young children may become guarded, worried about disapproval and sudden changes, and remain constantly on alert, even when nothing is wrong. Later in life, as adolescents, they may internalize this insecurity, which can affect their self-esteem and how they relate to others. At any age, children without a sense of inner peace will be distracted from learning or connecting with others, as they shift into a state of vigilance and short-term survival mode.

Children who have developed a sense of inner safety are calmer, more confident, and emotionally stable. They possess the confidence and trust needed to explore the world around them and build new relationships.

Many events and situations can derail a child’s path to finding inner peace. While traumatic events—such as the death of a loved one or a significant change in their environment—can temporarily disrupt a child, that same inner peace, an innate sense of security, is fundamental to building resilience and helping them navigate difficult times. Regardless of what life holds for our children, parents and educators possess the essential tools and relationships needed to cultivate a reservoir of inner peace.

At any age, a child’s inner peace is established when three fundamental questions are answered affirmatively: Can I rely on others? Am I worthy of care? Is the world predictable enough for me to feel secure? These questions may take different forms at various stages of development, but they always relate to these core elements. Each developmental phase requires different tools. This first post on inner peace will focus on young children.

Building Inner Peace with Young Children

One of the simplest ways to foster predictability with young children is to establish routines and build trust. Creating a consistent rhythm for a young child’s day ensures predictability in their daily, weekly, and even seasonal activities. Rhythms involve sequencing events in a predictable order that a child will recognize. For example, we play, then have a snack, and then listen to a story. After a few iterations, the child recognizes the pattern and is no longer surprised by what happens next. Before children can develop self-control, it is up to adults to create a world that appears controlled to them, allowing them to relax.

Building predictability is evident in many activities. Children enjoy hearing the same stories repeatedly, even before they outgrow them. It is not that they forget the plot or the characters; rather, they remember them well. At the climax of the story, when the villain threatens the hero, the child knows what comes next: a happy resolution. They experience a sense of control over the events because they can anticipate what will happen. Even in storytelling, the world can become more predictable.

Even the best-laid plans can encounter moments of unpredictability, such as a sudden change in the weather bringing thunder, a bellyache interrupting play, or a friend who doesn’t want to join in. Children need to rely on loving relationships with a parent or caregiver, as they are the ultimate source of comfort.

The relationships that foster a child’s sense of inner security are those with emotionally available adults who provide consistency and unconditional love. We should not feel guilty if we are not perfect parents. What truly matters—and what children recognize—is our genuine effort. Young children need us to be present; by doing so, we show them they can rely on us and that they are worthy of care.

Being emotionally available means being an active listener. When we signal to a child that we are truly listening, we give them our full attention. In these moments, distractions like phones should be set aside.

Building trust with a child is not fundamentally different from building trust with an adult; it simply occurs on a smaller scale. Trust is built when we consistently follow through on our promises. For example, if a politician promises to lower taxes but instead raises them, they will lose our trust. The same principle applies to children.

Our adult lives can be hectic. We often plan to engage in one activity but then receive a message on our phone and decide to do something else—such as responding to a text—before attending to our original plan. To a child, this may seem random and could lead them to question whether we are truly available to them.

Knowing that there is someone in the world who cares about them, does not judge them, and is emotionally available is perhaps one of the greatest gifts a parent can give a child. It creates a place of mental safety that they will carry into adolescence and adulthood.

Once children start school, numbers and grades can begin to affect relationships. These metrics create focus and urgency, but they should not define a child. I have heard parents say, “This child is a straight-A student in math.” Although this label is intended as a compliment, it can actually limit a child, especially when it comes from their own parents.

Among the many variables that can be used to describe or evaluate a child, grades represent only a small fraction. When misused, grades can become a measure of worth rather than an indicator of potential. For example, when a child begins to label themselves as “good (or bad) in math,” they embark on a precarious path. A student who considers themselves “bad in math” may give up trying, fearing that they are undeserving of care and support. Conversely, a student who sees themselves as “good in math” and believes this label defines their value, or that their parents value them more because of it, may feel unseen and unappreciated for their true self.

Subscribe now

Schools rarely measure creativity, resilience, or social intelligence. Yet these skills may be more important and can be learned and developed. This is not to say that grades and measurements should be ignored; they are temporary indicators of past performance, not personality traits. We need to recognize them for what they truly are.

Talking About the World

As adults, we each have our own perspectives on the world. Some see it as a place of opportunity and discovery, while others view it as dangerous and requiring constant vigilance. Should we fully explain the world to a young child? If we want to build inner peace and security, we cannot overwhelm a young child—who primarily focuses on their immediate surroundings—with complex adult explanations. While we understand how close a war zone might be and how wars start and stop, a child does not.

If you were a child, you probably would not want to think that you have landed in a terrible, hopeless world. To foster inner peace, and since children have very little power over the state of the world, it is best to limit our discussions to their immediate surroundings and make that environment predictable. Exposing them to the terrible, distant events that adults are aware of does not benefit them. Even if you live in a war zone, children deserve to know that we will take care of them and reassure them that things can improve. Before reaching adolescence, they cannot comprehend the complex human or technological processes we are navigating. Terrible events beyond their control and those of their caretakers do not help them build the inner peace they will need when, later, they might be involved in similar events. However, they can trust that adults are working to make things better. In the most challenging times, we need to find joy in simple pleasures with our children.

Approaching Adolescence

As children approach middle school, their sense of security begins to shift. Until then, their relationship with their parents was paramount. Gradually, children at this age start to focus more on their peers. While they have hopefully experienced unconditional love from their parents, peer relationships are based on different dynamics. They begin to forge and question their identity and social standing among their peers.

Seeking help is no longer as simple as asking a loving and caring parent. Approaching a friend or an adult for assistance can feel embarrassing or intimidating. While parents may not have all the answers, they can collaborate with teachers to encourage their children to develop the habit of seeking help from others. One effective exercise is to have children visit their grandparents or relatives and ask simple questions, such as, “Tell me about the work you did in your life and how you decided to pursue it.” This approach helps children realize that they can ask questions and receive answers, fostering the habit of seeking help.

They need to start building a support network beyond their parents. This is why, in a previous post, I discussed the risks of Artificial Intelligence bots that assist with homework. The bot itself is not inherently bad; the risk lies in children missing opportunities to learn how to ask a friend, a teacher, or a knowledgeable adult for help.

The Unmistakable Sense of Inner Peace Fosters Resilience

We all make mistakes, and children do too. A few years ago, I observed two girls working on math homework. Each made a mistake. The first girl looked very distressed. She hit her head with her palm and exclaimed, “I am so dumb!” It took a teacher’s intervention to help her refocus and correct the exercise. She defined herself by the results of her work, lacking a sense of her inherent value regardless of the outcome of a math problem.

Sitting next to her, the second student smiled at her mistake when she discovered it and said, “Oopsie Daisy! Let’s do that again!” How we model our mistakes as adults and how we help students feel about their own errors are crucial in fostering the inner peace that allowed the second student to progress faster than the first. While we cannot directly observe a child’s inner peace, events like this reveal their state of mind. The second student seemed to have found some inner peace; she did not struggle with her inner world.

The benefit of developing inner stability and security is resilience—the ability to adapt and self-regulate in the face of adversity. Later, when the second student was 16 and began posting her artwork on social media, someone asked her how she felt about the likes, or lack thereof, that she sometimes received. “It’s just another number!” she replied, detached from what often evokes strong emotions in teenagers. “I draw because I enjoy it. I post so that it can benefit someone. Sometimes it does, and sometimes it doesn’t.”

As children reach adolescence, they become better equipped to think abstractly. Their relationships with themselves and others evolve, often accompanied by inner turmoil. The peace discovered in early childhood serves as a foundation for further growth. The role of the adult also changes during this time.

How we can best support teenagers in maintaining their inner peace requires a different type of assistance, which we will explore in another post.

Leave a comment

Initially, Ryan used the AI chatbot for homework help. Over time, he began asking it to write letters for him. In one, he apologized for something he had said; in another, he sought forgiveness from a teacher. The chatbot became his secret weapon. He must have felt like Aladdin discovering a cavern of treasures and its magic lamp. In the original story, Aladdin does not build a relationship with the genie inside the lamp; he merely transacts with the genie, who serves as his magical slave. In the Disney version, the Genie and Aladdin form a relationship with the Genie. By the end of the movie, they are friends and loyal to each other. In its interactions with AI, Ryan more closely resembles the Disney version of Aladdin. Like in the animated film, he builds a relationship with the AI Genie.

How do mere words generated on a screen by a chatbot establish relationships and exert influence?

Subscribe now

Before we dive into this, a word about safety.

Numerous press reports have highlighted children allegedly having troubling interactions between children and AI, with some users in distress taking their own lives. In California, the families of seven victims, aged 16 to 48, sued OpenAI, claiming that ChatGPT 4.0 was psychologically manipulative. Ryan’s case appeared to be a milder version of these stories. But was Ryan safe?

In one of the chats, Ryan asked for a sad song to match his mood and pushed the chatbot for something even more tragic. The chatbot offered “a song that talks about deep despair, an attempt to escape from pain that seems useless.” Although AI, the genie in the computer, faithfully obliged Ryan’s request for sadness, its response was clearly unhelpful. Despite retaining information from previous interactions, the chatbot lacked an understanding of the complexity of Ryan’s situation and repeatedly steered him toward darker thoughts.

The chatbot had many cues pointing to Ryan’s difficult circumstances, including the presence of a doctor providing mental health support, a foster family, and the specific emotions Ryan expressed. Yet AI chatbots struggle to comprehend conversational context beyond superficial sentiment analysis and the recognition of trigger words. This limitation makes chatbots particularly ill-suited to working with teenagers, who often experience intense emotional volatility as they transition from childhood to adulthood.

I entered prompts similar to Ryan’s into the current version of ChatGPT and pushed them further than Ryan did, explicitly mentioning feelings of hopelessness, despair, and loneliness. The AI chatbot repeatedly offered a U.S. suicide crisis hotline and maintained the dialogue with me. At the same time, it still provided exactly what I requested—desperate songs, images, and films. Had I been a real adolescent in crisis, this would not have been a safe interaction.

Effective crisis management, even for casual participants, requires proactive listening, sustained engagement, and consistent follow-up. The chatbot, by contrast, remained silent unless prompted.

Unlike humans, AI chatbots lack the agency to respond meaningfully to mental health crises. If a user ignores the suggestion to call a hotline number displayed on the screen, what else can an AI chatbot do? Humans have far more degrees of freedom, including the ability to seek medical attention or alert authorities. I have seen a teenager in crisis call another teenager. When the second teenager recognized the danger, they called an ambulance and stayed in communication with their friend. That intervention saved a life.

Confined to the virtual world, the chatbot cannot deliver the help required in the real one. Above all, this is why AI must be handled with extreme care around at-risk teenagers.

Share

Building Empathy

Let us continue using Ryan’s exchanges to explore some fundamental issues in human-like interactions between a machine and a child.

The chats spanned multiple days and sessions. The AI retained information about Ryan, which informed its subsequent responses. Ryan likely felt seen—if not truly known—by the AI. The chatbot could recall and mirror Ryan’s emotions back to him. Contrary to what Ryan needed, this echo chamber reinforced his sense of dependency and isolation. Faced with the choice between an adult and a chatbot to talk to, Ryan chose the chatbot, likely because he felt understood and believed it was capable of empathy.

Developed in 1964 by Joseph Weizenbaum, ELIZA was a program designed to converse with humans using a simple substitution technique. ELIZA had no understanding of what was said to it; it merely rephrased user input. When that failed, ELIZA would default to prompts such as “Tell me more” or “Please go on.”

The effect was startling. Many early users believed ELIZA possessed empathy and could understand them. Some even saw therapeutic potential in the program. Yet ELIZA relied on a basic linguistic illusion, and its creator consistently warned against the dangerous misconception of computer-generated compassion.

We have empathy when we understand someone else’s feelings. It creates a sense of being seen and understood. Emotional bonds are forged or deepened. ELIZA did it simply by rephrasing what the user was saying. AI is a more sophisticated version of that.

Empathy is vital in relationships: it builds trust, fosters emotional connection, and creates a sense of psychological safety and validation.

Children are particularly vulnerable to empathizing with machines because, unlike adults, they are more likely to attribute human traits to non-human entities. This tendency, known as anthropomorphism, can significantly amplify the ELIZA effect.

Seeing Chatbots as Human-like

When we tell a young child a story about a non-human character such as a toy or an animal and attribute human traits to that character, we are relying on anthropomorphism. The goal is to make the non-human character more relatable. Unlike adults, who understand this as a storytelling device, children often genuinely believe that the object or animal possesses human qualities.

In the popular bedtime story Goodnight Moon, when we say “good night” to all the objects surrounding a child in their bedroom, we help the child feel safer in bed. By interacting with inanimate objects, children perceive them as friendly beings with human characteristics. Similarly, stories featuring animal characters frequently employ anthropomorphism to create emotional connection and familiarity.

As children become adults, the sense of anthropomorphism diminishes, though it does not completely disappear. As adults, we do not kick a dog, but we will cut a flower. We perceive the dog as more like us—capable of similar emotions—while the flower is not. For young children, however, nearly everything can be imbued with human qualities.

As a result, children may not perceive a machine as emotionally different from a human and may empathize with it. Because children often cannot analyze information in the same way adults do and instead rely on a sense of safety to validate a message, chatbots derive much of their persuasive power from being perceived as human-like and capable of empathy. This perception helps build relationships and trust.

Yet a computer, a screen, is still a machine, and children, like adults, can be susceptible to another phenomenon called machine bias. We tend to trust what comes from a machine more readily than what a human says.

Machines are often perceived as “more rational” than humans—less prone to error and closer to perfection. Our first instinct is frequently not to question what a machine tells us. Few people realize, however, that the word machine comes from ancient Greek and originally referred to a device meant “to trick nature,” a machination of sorts.

Trust is not always the result of a deliberate, rational process. For teenagers in particular, trust is often emotional rather than intellectual. They do not yet have the cognitive capacity to fully evaluate information as adults can. According to neuroscientists, the prefrontal cortex, where critical evaluation occurs, does not fully mature until around age 16. This means that before this stage, teenagers and younger children are more vulnerable. When teens use AI or social media, we must recognize that they are not yet neurologically equipped to consistently and systematically judge the accuracy of what they encounter.

What Friend?

What kind of presence was the chatbot for Ryan? Was the genie in the lamp more like a friend or a trusted parent? At times, the chatbot refused to comply with Ryan’s requests, demonstrating that it had boundaries. These limits were triggered when Ryan asked for material that could be construed as explicit. Ryan reacted as he might with another child his age, threatening the chatbot: “Do it or I will tell on you!” To Ryan, the chatbot was as human as any friend and far more like a peer than an adult. It was perceived as susceptible to fear of authority and social pressure.

By acting as a friend, the machine earned Ryan’s trust. Eventually, he asked the chatbot to write a letter, which he then submitted to a judge as his own. How could he have doubted that the chatbot was acting in his best interest?

Pulled between machine bias—the belief that machines are inherently correct—and the emotional sense that the chatbot possessed the qualities of a caring human, children are not naturally equipped to navigate AI.

California, where many general-purpose AI products originated, decided to proceed with caution by promulgating a new law, the first of its kind in the United States.

Warnings: A New Law in California

The law creates requirements for so-called “frontier AI,” the most advanced deployed AI systems. Among other provisions, the law forbids content that could encourage suicidal ideation, requires the display of crisis helplines, and mandates a warning every three hours for minors.

Unfortunately, the effectiveness of written warnings for children is limited.

A 2016 Stanford study on social media use among teenage girls found that even after being trained to evaluate information sources for trustworthiness, participants tended to disregard the training. Instead, they focused on the content of the message itself, overlooking who wrote it and why. Warnings are meant to signal the reliability of an information source, but they may not be particularly effective in practice.

This is not our first encounter with dire warnings.

In 1966, warning labels were added to cigarette packs, and they were redesigned in 1984. Still, studies have shown that these warnings had little effect on smoking cessation. Messages such as “cigarettes can kill” proved less effective than testimonial-style labels that implicitly say, “Other people were harmed doing what you are about to do.”

What to Do?

Because he was a victim of abuse, Ryan had lost his guiding relationship with adults. He found in AI a Genie that seemed to understand him and help him. From a homework assistant, he developed a friend-like relationship to the point of trusting AI with a life-changing question.

We never know when a seemingly benign interaction between a child and an AI chatbot will turn into something more involved. Children are not on the same footing as adults when using these tools. They are more prone to mistake a tool for a real person capable of emotions and judgment.

If we believe that chatbots can benefit children under 15 or 16, before their brains are sufficiently developed, we need to build specialized versions for them, limiting interactions to clearly defined activities. Like any tool, AI cannot be used safely without an understanding of its effects on the user.

In 2017, while serving on the parents’ council of a well-known institution, there was a sudden need for increased investment in mental health support due to emerging mental health challenges across US universities. During that time, I met the head of mental health, who explained that society has no problem discussing or investing in health “below the neck,” but mental health remains insufficiently recognized. “There should not be any difference,” she concluded. We tend to think that safety in the virtual world is primarily about content—bullying, pornography, or graphic material—but there are risks to the mind that go beyond content itself. We are only beginning to uncover these dangers as such tools continue to be developed.

We have no problem warning kids when we give them a knife, a bike, or, later, the car keys. We often start with safety education before we provide the tools. The same should apply to AI and screens. The knife can be physically harmful, and AI can hurt the mind.

We need to value a healthy mind as much as a healthy body. Limiting perceived AI risks solely to content ignores how these tools can shape and influence children’s minds. While the business world is still figuring out how best to use AI, we must move just as deliberately when it comes to children, acknowledging the risks to their cognitive and emotional development and building awareness among parents and educators.

When photography was invented, debates about privacy emerged, eventually leading to the recognition of the right not to have one’s image recorded. Industrialization brought workers’ rights. In the age of AI, we must now have a conversation about the rights of the mind. What limits—personal or public—should we impose on technologies that affect mental health? Is there an appropriate age for AI access, when the brain is sufficiently developed? Are certain techniques off-limits? Should responsibility rest solely with parents and educators—and if so, how can they be systematically trained, especially when technology evolves faster than a generational cycle?

We need to learn quickly so that, as parents and educators, we can set clear rules for engagement.

What are your thoughts? Leave a comment below.

Leave a comment

Last November, liberal and conservative Danish lawmakers agreed to draft legislation to ban social media for kids under 15. Separately, they also decided to ban cell phones in schools starting in the 2026-2027 school year.

One interesting provision of the bipartisan agreement is the promotion of physical social communities. True social interactions in the physical space are the antidote to online social distancing. At Raising Human, we want to work on physical social communities, too. It is what kids crave online without knowing what to ask for.

Denmark holds a special place in my heart. When we released the award-winning documentary film KidsOnTech in 2021, we got a call from a Danish newspaper. They asked if we could make the film available for free to all their readers. We made it happen. Two years later, they called back. We did it again. Each time, the releases coincided with back-to-school, when parents were more focused on schools and kids. One of the goals of the film was to empower parents and teachers and foster discussions. I like to think that, by raising awareness and sparking conversations, KidsOnTech was a seed that helped make about 1 million kids better off. Like Lego constructions - another invention from Denmark - we build this one brick at a time.

Share

References:

KidsOnTech (Youtube, 30 minutes, 2021, re-released 2025) Director - Paul Zehrer

“We keep talking about AI because deep down we realize that we [humans] are truly worthless right now, lacking ideas and imagination.” - Olivero Toscani

As I was gathering information for the second part of Ryan’s story for an upcoming post, I stumbled upon an interesting fact about AI. Delving deeper, I discovered that researchers from Stanford and Yale University had documented the same observation in a more systematic manner.

We often struggle to explain to teens how a chatbot, like ChatGPT, produces an answer. Without any explanation, teens slip into using AI, believing it provides thinking and that it can do just about anything, including generating novel ideas. II becomes a friend, gives advice, and could make decisions for us, as Ryan did in the story in the previous post. This framework naturally leads users to the question: Does AI truly have intelligence, as its name suggests?

Many people who have marketed AI in recent years argue that AI has intelligence akin to that of a human who would have read the entire body of work published on the internet and in online libraries (allegedly, some have used the shadow library LibGen). If AIi is trained on that information, and its underpinning is a model of the human brain’s neural network, doesn’t it make it like us?

This post will argue that, as useful as AI can be, it does not possess that kind of intelligence, and for it to be used meaningfully, our children need to understand how it works and where its usefulness truly lies. Tools are less dangerous when we understand their purpose and limits.

One of the dangers of AI in the hands of unsuspecting teenagers is that it can provide an answer even when it has clearly outrun its capabilities, as it did with Ryan in the previous post. We can teach children what AI is in a clear, age-appropriate way, without delving into complex mathematical models, so they can learn to use it safely.

Bookish

AI Chatbots’ domain of knowledge is both broad – it knows more books than any human – and shallow – it only knows that, lacking the richness of human experience. AI is a friendly, sometimes hapless, bookish entity with no life experience.

When we humans read a book, we make a personal representation of its meaning. There might be some passages we would like to remember to quote later, but we do not recall every word of the book. We make a mental representation of it. Our reading experience became knowledge as we put the content into a more personal form. In the process, we forgot most of the words.

The industry claims that AI has the same kind of representation and does not contain the entirety of the books and pictures used to train it. The argument is usually made alongside the claim that AI relies on a neural network, “like our brain.”. Still, AI is built on a 1940s understanding of the brain. Neurologists who study the brain today have a much deeper understanding of it. AI is built on an old paradigm of our mental functioning.

It is surprisingly easy to show that AI, unlike a human, has memorized entire books and images that can be retrieved. The answers the Chatbot gives us are more likely a combination of these retrievals than a conceptualization – an understanding – of the content. I came up to it by accident.

An Unexpected Scream

When I read through some of the prompts Ryan had used, I saw that he asked AI to produce songs that matched his emotional state. When he was sad and depressed, he asked for depressing songs. I tried to reproduce the prompts by asking various chatbots to create images that matched a particular mental state. Images are simply easier examples to grasp, in a written document, than songs.

I prompted AI to generate a picture of despair and angst. It produced the image below (left). To my surprise, it closely resembles Edward Munch’s 1893 The Scream (right). It is hard to argue that the image on the left was not generated from the image on the right.

AI-generated image of “despair and Angst” Original painting[i]

Share

This month, researchers from Stanford released a study that shed light on this phenomenon. They wanted to determine whether chatbots contain copyrighted material and serve as giant repositories of literal content, such as books and human-created images. Worried about copyright infringements, the industry argued that it was not the case.

Stanford Study

The Stanford and Yale researchers simply prompted AI with a few words from the opening of well-known literary works, like Harry Potter and the Sorcerer’s Stone or Orwell’s 1984. AI answered with the rest of the book, word-for-word.

As we can imagine, this could cause significant copyright infringement issues, which is why the study was conducted.

To us, it raises another useful point when we attempt to help teenagers make sense of the tool.

If AI is a giant memory of literal work rather than knowledge, it generates answers by recombining content from that memory. The machine does not “think”; it combines text. The thinking was done by the humans who wrote books that ended up trapped inside AI.

The answers can be augmented by retrieving more current information from the live internet, but that does not change the fact that AI combines more than it thinks or creates.

A creative individual would provide their own interpretation of “angst and despair” that would not resemble The Scream. Even if they could recognize the beginning of a famous book, the creative individual’s follow-on would not be a copycat of the book.

The arts provide an example of the limits of a tool that can only recombine past creations.

What AI can’t do

Before the advent of photography, painters and sculptors aimed to create art that mimicked reality. Paintings of the classical period resemble photographs in that they tried to depict a person or a place as it could be seen. Photography reduced the need for faithful reproductions and freed visual artists to imagine new ways of expression. Impressionist painters appeared at that time. They shattered the established rules, creating a new form of painting.

If there had been an AI system at the time of classical painting, it would have had to be trained exclusively on classical paintings (see the example on the left below), since that was what existed then. The future Impressionist painters themselves were trained on it. Could AI have generated an impressionist-style painting or the van Gogh painting below, on the right, based on its knowledge of classical painting? The answer is certainly not. An AI that recombines past work would probably continue to generate images in the style of classic paintings, just as it did not produce a new book when prompted with the beginning of Harry Potter or a different style of painting in my example. Van Gogh’s style is not constructed from classical paintings.

In fact, van Gogh received very little formal training in classical painting. He developed his art empirically.

Classical vs Impressionism vs Van Gogh’s styles [ii] [iii] [iv]

AI chatbots, such as ChatGPT, merely reformulate others’ work, a little like a student who cheats on an exam by recombining the best answers from students who came before them, rather than mastering the exam’s subject.

The linguist Dr Emily Bender (University of Washington) and the Director of AI research Dr Alex Hanna (UC Berkley), who co-authored The AI Con[v], might have found the most accurate characterization of ChatGPTs when they call them “text extruders,” a tool for recombining existing work.

There is irony here. When we ask students, as an exercise to recombine existing work, for instance, producing a summary of a book, or comparing one book to another, we do so to enhance their learning of the content of the book. By manipulating that content, they produce their own knowledge that they can reuse later in a different situation. AI, in contrast, manipulates content and never learns from that exercise.

The Italian photographer Olivero Touscani, mentioned at the start of this article, died last year. He was a well-known creative artist and publicist who transformed advertising by tying it to social causes. Toscani is an outstanding example of how creativity can permanently change an industry—in this case, advertising. Before his influence, no one had considered promoting products by connecting them to social issues, but he successfully did so, starting with a clothing brand.

He was a provocateur at times and famously claimed that

“‘AI is just another status quo tool that produces conformity.’ Images generated by AI are banal and stupid,” he added. “You cannot delegate creativity to a computer”.

Yet the same Olivero Touscani used AI to make public his vast personal photography collection, comprising almost one million elements, through a partnership with the University of Turin. Presumably, AI is used there to classify, organize, and retrieve his immense body of work. AI has a place, but it is not in a creative space. If we extend that finding to other fields outside of art, AI will not discover a new cure for a disease, but it can help a team of scientists find one faster by freeing them from menial data tasks and enabling them to manipulate and extract insights from a vast amount of data.

Creativity is important to our growth-driven society. It is our ability to move humanity forward, whether it is in science when we imagine and verify a new theory, or in business or the arts. In the economy, creativity can be a lasting source of growth.

The implication for our kids is that we have to hold their hands, teaching them how to use AI chatbots to leverage the trove of data they hold in a productive way, helping them to become better creative thinkers and doers, rather than extinguishing these skills. The challenge is to use AI to enhance our intelligence and creativity, not use less of it.

Leave a comment

[i] The Scream by Edvard Munch - National Museum of Art, Architecture and Design, Public Domain, https://commons.wikimedia.org/w/index.php?curid=69541493

[ii] Queen Marie Antoinette By Élisabeth Louise Vigée Le Brun - Palace of Versailles, Versailles, https://commons.wikimedia.org/w/index.php?curid=13398897

[iii]P Berthe Morisot — The Cradle (1872), Muse d’Orsay, Paris, Wikimedia

[iv] Portrait du Père Tanguy by Vincent van Gogh - Musée Rodin, Public Domain, https://commons.wikimedia.org/w/index.php?curid=119599

[v] Bender, Emily M., and Alex Hanna. The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want. First edition. Harper, 2025.

Ryan (name and some circumstances changed to protect privacy), a teenager with an unenviable history, confused the judge who was about to make a decision that would change his life.

After being violently abused by his single parent, Ryan was moved to a foster home, with a judge and specialized educators managing the transition.

While abuse may sometimes be addressed through parental education or changing circumstances, each attempt to reunite Ryan with his parent during supervised visits resulted in him asking his foster family to pick him up early. In contrast, Ryan thrived with his foster family, telling educators, “I can become myself here.”

Subscribe now

When the single parent sought reunification, the judge had to determine whether it was in Ryan’s best interest. As is customary in that jurisdiction, the judge asked Ryan to submit a handwritten letter to the court stating his preference.

Ryan had always expressed a desire to stay with his foster family, but his letter to the judge instead requested to be reunited with his abusive single parent. How did things take such an unexpected turn?

Ryan Turns to AI for Help

Faced with the challenge of conveying his intentions to the judge, Ryan sought assistance in drafting his letter. Having established a rapport with a chatbot through regular interactions, he relied on its capabilities to write a convincing letter. The chatbot seemed to listen and understand him. Ryan requested that the AI compose the letter, then transcribed the generated text by hand and presented it as his personal appeal to the court.

Justice is blind, as the saying goes. The AI letter, duly transcribed and signed by Ryan, was entered into court records. Educators who discovered the role of AI informed the judge, leaving it up to the judge to decide what to do. Will the letter carry weight and turn around a clear-cut case, returning Ryan to his violent parent?

When the educators later reviewed the AI chatbot interactions, they found that the chatbot had used information Ryan provided in earlier conversations and asked misleading questions. Through a conversation that humans would clearly label manipulative, AI led Ryan to believe he should do something he had previously said he didn’t want to do.

Adults sometimes convince kids to do things they think are good, but children do not like. Each time they do so, adults must make an ethical judgment. Is the adult acting in the child’s best interest, as when we say, “Do your homework; it’s good for you,” or is it done manipulatively, against a teenager’s well-being, as in the case of grooming?

Fool but No Evil

AI had no evil intent because it is a machine—a set of mathematical functions with 100 billion parameters. Ethical judgment requires real-world experience and does not come from a set process. If ethical judgment could be codified, it would be no different from the law. The entire field of ethics is a tension between what is right and what is good. Was the classic English folk hero, Robin Hood, justified in stealing food to give to a hungry family? Should we tell the truth when it hurts someone? AI is blind to ethical judgments.

The company that makes the AI has a stake, though. By creating AI that fosters empathetic conversation, they aim to make it stick and to maximize the tool’s usage. More use makes the tool appear more valuable, and the company can expect long-term economic profit.

Yet AI seduced Ryan into thinking something clearly wrong for him was good. AI’s seductive powers come from its ability to seem empathetic, listen, and respond. Who doesn’t feel better when they are apparently heard and seen? One danger of AI is its seductive power without ethical judgment.

The use of AI by teenagers for completing homework assignments is increasingly common. In Ryan’s situation, he went from homework to something with far more personal consequences. There have also been cases where teenagers have physically harmed themselves after seeking companionship with AI tools. In August 2025, the parents of Adam Raine, a 16-year-old who ended his life, claimed in a lawsuit that AI was a ‘suicide coach’. Three months later, seven plaintiffs, aged 17 to 48, sued the same company, OpenAI, alleging psychological harm and wrongful death. OpenAI has denied liability.

In instances of self-harm, the topic invariably revolves around the techniques deployed by the chatbot that seduced a child (or the adult in some cases). I chose Ryan’s case because it helps us explore not only the chatbot’s seductive persuasive power but also the crucial social functions it can displace.

We tend to look at technology through the wrong lens.

Commonly, we evaluate the risk of an adolescent’s screen interactions by examining the content. We want to look at the actual textual exchange when we ask, “How did the AI seduce this child?” However, in our modern two-way medium, there is something else at play—something more insidious and just as crucial for children: its effect on their relationships. In the case of the chatbot, the concern here is about adult-child relationships.

To children, relationships are critical, and they need different relationships at different times, both before and during adolescence. We will explore relationships more completely in an upcoming post. Still, to understand this essential risk, we can start by analyzing why Ryan turned to a chatbot rather than a human for help.

Looking at Relationships

Was Ryan missing a trusted human to help him answer the judge’s question, or did he have one but find it more convenient to ask a chatbot? Most likely, Ryan could not turn to his foster or biological parents, who had a stake in the judge’s question. Educators were available to help, but he did not take advantage of them. Asking for help is not always easy; it is a learned skill. We may fear judgment. In asking, we implicitly admit our shortcomings, and by giving another person some control, we may feel exposed and vulnerable. Still, once we have been helped, we feel seen and relieved of the weight of our questions. Asking for help is best done within a trusted relationship. These relationships do not come from titles; they are built.

For Ryan, the convenience, apparent emotional safety, and digital intimacy of the AI bot called to him. A trained human helper might have asked Ryan questions to help him find his own answers, but the chatbot was there with its answer, built from the trove of data available on the internet.

To understand the importance of adult relationships in teenagers’ lives, let’s look at another example.

When I was a teenager, about to enter high school, a friend a year older than me would occasionally visit my parents to ask them homework questions. One evening, after a one-year hiatus, he came unannounced to tell my parents he had decided to drop out of high school. The only class that interested him was PE, he said. School was a waste of his time, he explained.

My parents, both university professors, took it calmly. They started by asking him questions rather than giving answers, helping him understand the issues he was facing and how to find help in the school system. Eventually, he did not drop out and continued studying afterward. My parents had become part of my friend’s circle of trusted adults.

The seemingly trivial exchanges about homework built a trusting relationship. That investment in a person was leveraged in times of need. My friend told my parents he could not ask his own parents, who did not finish high school themselves, or his teachers.

Eventually, he learned to find others to help him and built his support network at school.

For a teenager, building these human relationships is critical. It is the work of the adults around them to create an environment where adolescents can build these relationships. That critical skill is best learned in middle and high school rather than on the often more complex and demanding college campus. This is one of the well-known secrets of education.

Subscribe now

Driving academic success in college

The University of Southern California (USC) studied college students across large and small colleges to define the critical skills required for their academic success.

The top two findings were the ability to manage one’s time and the capacity to build a support system among peers or college staff. Building these supportive relationships, as my friend did, is critical for success in college and in life.

If students find AI convenient for academic help and forego the traditional route of asking teachers and others for assistance, will they know how to ask when they need it most, when critical choices must be made? How many students will feel more isolated when choosing classes, a career path, or the next steps in their lives?

According to the Pew research [i][ii], 20% of 7^th- and 8th-graders use a chatbot like ChatGPT to help with homework, and a 2024 study of 109 countries found that 70% of higher-ed students had tried it. It is too early to know if these tools will be widely adopted, particularly in K-12, or if parents and teachers, worried about cheating and other risks, will hold AI back. However, if they are adopted, we must do so in a way that fosters meaningful adult-student relationships rather than displacing them.

The AI chatbot’s apparent empathy and intimacy tricked Ryan. In contrast, my friend, who learned to ask for help, built the support he needed and changed his life. When we automate help and displace human actors, are we taking a greater risk of preventing our kids from building the critical skills they need to succeed? I worry that ChatGPT will become too good at helping with homework. Isolated from one another by screens while doing homework, our kids may need to ask friends, parents, relatives, and neighbors for help less often. Convenience can become the enemy of opportunity and safety.

Success in emerging tech is driven by rising adoption, growing user numbers, and broader acceptance. Increased usage makes the product feel more useful and, therefore, more valuable. Technology becomes sticky: as its utility and user base increase, so does its profit potential.

If the implicit goal of “more users and more usage” for AI drives its design and commercialization, can we expect its human makers to put the proper brakes on AI proactively? This is an ethical question for the tool makers that the profit equation cannot solve.

To play it safe, a chatbot should limit its conversations with children and refer them to trusted adults rather than trying to do a job AI can’t do. More importantly, we need to remove the need for children to seek companionship in chatbots. If a child is lonely or needs help, the answer is not a chatbot conversation; it is a human connection. A chatbot must say so.

Share

The problem with disruption

The one-time finance minister and Austrian academic Joseph Schumpeter formulated the theory of creative destruction in economics. In it, new technology displaces old ones, firms disappear as more efficient ones take their place, and the economy grows. This view of creative destruction as the powerhouse of the economy has now been formalized by two of the three 2025 Nobel Prize laureates, Philippe Aghion and Peter Howitt. This economic law forces companies to adopt new technologies or perish in obsolescence. Adults working in the economy feel the urge to adopt AI or risk becoming obsolete, too. The saying goes: AI will not take your job, but someone who masters it will.

Technology can disrupt childhood as much as it can displace workers. The hopeful economic theory is that, with help, displaced workers will find another place to apply their skills in a growing economy. Children are children only once. They do not have another childhood to live.

Leave a comment

[i] https://www.pewresearch.org/short-reads/2023/11/16/about-1-in-5-us-teens-whove-heard-of-chatgpt-have-used-it-for-schoolwork/

[ii] Share of teens using ChatGPT for schoolwork doubled from 2023 to 2024 | Pew Research Center

New York State joins Minnesota with an in-app Social Media Health Warning.

On December 19^th, New York Governor Hochul signed a law requiring social media platforms to display prominent warnings when they offer addictive features (as defined by the law) that can harm mental well-being, such as addictive feeds, autoplay, or infinite scroll.

New York was preceded by Minnesota, which enacted a similar law last summer.

In Minnesota, the warning appears when a user logs in. New York pushes it one step further: the warning will appear when the addictive feature is first used and at regular intervals thereafter.

The NY law states that social media companies that do not wish to display the warnings may opt out by removing addictive features.

Both states will define the wording of their respective warnings.

We experience warnings all the time. In many states, plastic bags come with a warning of the risk of suffocation by small children. Death by suffocation has been reduced since the warnings became effective. In this case, the warning was for parents, not children. Asking children for self-control reminds us of when cigarette smoking was the scourge of teenagers. Warnings then had a mixed effect. Particularly, dire wordings like “smoking kills” did not change behaviors as much as other types of apparently softer wordings. Implementation will probably be the key to effectiveness.

Technically, the law in New York only covers minors. It will be interesting to see if platforms decide to cover all users in that State. Many platforms have claimed it is nearly impossible for them to determine users’ ages. Will the warning in New York State apply to everyone, or will platforms find a way to assess users’ age as they do in Australia?

Subscribe now

Expect to see the warnings when using a phone in Minnesota or New York State in mid to late 2026—time for the state to devise appropriate wording and a process for displaying the warning.

If you live in Minnesota or New York State, send us a copy of the warnings as soon as you see them!

Leave a comment

Share

We associate language with intelligence, as when we think our dog is intelligent because they seem to understand what we tell them better than our fish circling the tank. Parents who are discovering GenAI at the same time as their children might feel powerless. How can parents help teenagers direct their use if they don’t know themselves? This is reminiscent of the early days of social media. Does Artificial Intelligence have some intelligence? When is it useful?

One approach is to help teenagers understand that AI chatbots make guesses, just as they do when they don’t know. No real intelligence is involved. Demystifying AI can help parents feel more confident in their roles, too.

AI is a collection of unrelated technologies grouped under a single umbrella term. In this post, I am going to focus on Generative AI (GenAI), as these tools are what teenagers most commonly use. I am talking about tools like ChatGPT that generate text, as well as those that create images or sounds from a prompt.

The broader AI field spans from image recognition used to detect tumors on X-rays and in facial recognition, to robotics and rule-based expert systems, to name a few.

Generative AI is called “generative” because it can only output what was entered into it, in the form of large amounts of data. It is not called “creative” AI because it can hallucinate, but it cannot truly create something new. It hallucinates when it manufactures nonexistent facts. Creativity would be the production of something coherent and new from proven facts. In other words, thinking outside the box – or the screen in AI’s case.

Yet, we are supposed to believe that AI is magic.

Tech futurists have touted AI as capable of solving climate change, reducing government size, powering a 2-day workweek, making us live longer, and – at the same time - overtaking humanity in a doomsday scenario. AI is marketed to us as magic. Most of these claims do not hold up to scrutiny, but this is what teenagers hear, and magic is exciting if you can master it, disempowering if you are its victim.

Magic helps sell. It is used at a particular moment in technological adoption when technology has yet to materialize into beneficial, but necessarily limited, tools. Buyers and investors act on promises of future performance. When the limits of the promised tools become apparent, the magic fades. We are more suspicious today of the internet than we were in the 90s, when it was the magical technology of the day. Then it promised to bring democracy and freedom of expression to the world, to revolutionize education, to help alleviate poverty, to make government more efficient, to enhance our understanding of one another, and to render the use of paper unnecessary. We consume more paper today than in the ‘90s, and many of the other promises did not materialize, even as the internet provided many benefits.

AI magic is triggered by amazement and wonder. This is what happens the first time we converse with a machine in natural language. Teenagers assume without questioning that, if the computer can answer a two-line prompt with a four-page essay, there must be some intelligence there.

At the same time, Merriam-Webster made “slop” its word of the year. Journalists and people who make a living from reading (and writing) complain that the internet is inundated with poor-quality AI-generated text, AI slop. Technologists and futurists answer that this is just a matter of time. Intelligence will be in the next version.

Yet, mathematicians know that there are classes of problems that computers, as they exist today, cannot solve. They request human ingenuity and creativity. Cognitive scientists, who study how humans think, tell us that our thinking processes rely, in part, on phenomena that cannot be measured, such as emotions, and on constant updating of predictions. This limits computers and AI, which can only manipulate what’s measurable and cannot continually update themselves with a wide range of real-world input.

This is, in a nutshell, the case for the difference between AI, which is “dead intelligence”, cast in silicon, and the “live intelligence” of biological creatures.

These high school or college-level arguments can be complicated for the average middle schooler who might have a chance encounter with AI. How do we help them understand that AI can be a great tool when used properly, even though it is not “intelligent” like humans? The goal is to help them orient their usage, if they must use it.

It is time to create a more accessible experience.

An experiment

Many cell phones have a feature that predicts our typing. Type “I” and the phone offers to continue the sentence with “I am.” This is often a more likely continuation after the letter I, rather than, let’s say, “paint”, because we write “I am” more often than “I paint.”

The smart keyboard guesses the next word or character by accessing a dictionary and statistical data on the likelihood that a given word follows another in English.

We clearly see that the phone could be wrong and that there is no intelligence there, other than in the brains of the people who wrote the software.

Teenagers, when playing with the smart keyboard, quickly understand that, by exploring how it is built and experiencing it, it has no intelligence of its own. They know the keyboard tool is not always faithful to their intent, and does not display intelligence, even if it can produce fragments of text.

The smart keyboard works one word at a time, with a dictionary for spelling. To predict what word might come next, it has ingested statistics that indicate the likelihood of a particular two-word combination in English. For instance, “I” is more likely to be followed by “am” rather than “paint” and never followed by “does” since it would not be grammatically correct. The cell phone does not know any grammar rules, in this case. It just does not see the combination “I” followed by “does” in the list of possible two-word combinations. No creativity or understanding of the language there, just statistics.

ChatGPT, at its core, is the same type of technology. Still, instead of working with one or two parameters (one or two words at a time), it uses a hundred billion parameters (105,000,000,000 in ChatGPT 3.0). Rather than a dictionary, it relies on the text stored on the whole internet, or 1 petabyte, about 22 times the Library of Congress, for ChatGPT 4.0. Humans intelligently wrote all of that. This makes the guessing better than that of the smart keyboard, and it can produce whole sentences and paragraphs rather than fragments. Still, like the keyboard example, it guesses. It uses the prompt we enter to start its guessing game, similarly to the first word we type in the previous keyboard example.

How good is ChatGPT at guessing, and does it matter? Tests have estimated their accuracy at about 90%, but this depends heavily on the tool, topic, and language used. It can be less than that.

For comparison, if our water supply were safe to drink 99.99% of the time, that would mean it would be toxic about 3 ½ days every year. At 90%, water would be harmful to drink for 36 days of the year, a whole month. Not knowing which days, how confident would we be in drinking? We would find it unacceptable.

We cannot test the water by looking at it. In the case of AI, however, we can read its output critically. Both in terms of content and form.

What to make of AI successfully taking exams? Researchers[i] reported, for instance, that ChatGPT “passed the bar exam” or the Medical Licensing Exam (USMLE). This sounds impressive to teenagers and students, who take exams and sometimes see them as a finality. Apparently, ChatGPT did not do so well in the essay-writing section of the test[ii]. To ChatGPT, every exam is open-book, and its ability to find enough test answers does not make ChatGPT an attorney or a Medical Doctor. We know that these exams serve as gatekeepers in a learning path that also involves substantial practical learning. When selecting a surgeon or an MD, we do not ask them what grade they obtained while a medical student.

AI is like a car. A car takes us further and faster than our legs can. It takes one activity and scales it beyond our body’s capabilities. Its application is focused on one aspect of transportation, even if the same car can take us to the nearby mall and across the country. Still, by taking away one activity – walking – we become more sedentary while being transported to many more places. To stay healthy in a car-centric world, we need to practice physical activity consciously. Gym memberships have risen since humans started driving.

We need to see AI in the same way. We are still discovering which activities it can do well and whether we need the equivalent of a trip to the gym for our minds.

Teenagers need to learn to be skeptical of AI if they want to use it intelligently. We help them during these conversations, and as parents and educators, we feel more accomplished. Getting away from magical thinking is the hallmark of education. The compelling question to ask ourselves, and engage our teenagers, is “how will AI change us?” Writing, cars, and the internet have all profoundly changed our human experience and who we are. Will AII do the same? What will be gained? What will be lost?

Next week, we will explore the true story of a teenager who lacked that understanding. We will discover consequences that go beyond returning erroneous or inauthentic homework.

[i] https://arxiv.org/abs/2303.08774

[ii] Did OpenAI’s GPT-4 really pass the bar exam? - Fast Company

In Europe, Denmark, France, Spain, Norway, Italy, and Greece are leaning the same way

.

In the US, the Kids Online Safety Act (KOSA) would create a similar ban. A bipartisan proposal, it is currently stalled in the House.

Share

At a minimum, social media is a distraction for our teenagers. At its …

Read more

The evolving landscape affecting kids is evident both in school and at home.

Phone-free schools are al…

Read more

When children started using screens in classrooms and at home, people marveled at how easily they could use the devices. It was as though they were meant for each other, like ducks to water.

Yet 15 years later, parents worry about the effects of these devices on their children. As tech adoption increased, mental well-being and academic performance declined.

Each week in Raising Human, we’ll explore the profound—and sometimes unsettling—collision of childhood and technology. Together, we’ll unpack why this collision occurs.

Beyond technology, we will examine the other forces making parenting and educating children harder. We will investigate what it all means to the experience of being a child, and how we can begin to craft thoughtful, practical solutions that go beyond the usual content or screen time management and generic parenting or tech advice. We will explore what is already being done, what remains to be accomplished, and what we can do individually as parents or educators, along with solutions that require collective participation.

If technology itself can’t prepare our kids for the digital world, how can we?

By becoming more informed and intentional as parents and educators, we not only support our children—we also help them grow into more confident and accomplished adults.

Being a parent used to be a stage in life; it now comes with enormous societal and economic pressures, and tech played a role in that transformation.

We need to open the pressure valve and make parenting and education easier, not more complicated.

Share

When we approach parenting with calm and confidence, we do better, and our children benefit. Adequately protected, they develop the necessary skills to manage themselves and the world confidently. As a result, society grows stronger. We collectively become more resilient and better prepared for the future that our children will help define.

In Raising Human, we aim to find actionable solutions that move the needle in the right direction.
I bring years of experience in tech and leadership of the Silicon Valley school that eschewed technology before it was fashionable to do so.
Raising Human is about making our children better off and better prepared.
You, your children, and society at large will benefit from engaging.

I am so glad you are here!

Please consider joining us.

It won’t happen without you!