Sometime around 2012, the teenagers stopped being okay.
You can track it across every dataset that monitors adolescent mental health. Depression for teenage girls was flat through the 1990s and 2000s, stable enough that nobody was publishing alarmed papers about it. Then NSDUH data shows major depressive episodes among 12–17 year old girls increasing roughly 60% between 2010 and 2019. CDC numbers on self-harm hospitalizations for that same group tripled. Suicide rates went up too, and in ways that don't fit the usual explanations about better reporting or shifting definitions.
Boys showed similar patterns, less severe, but with the same inflection point: around 2012, accelerating through 2015, plateauing at levels that would've been a national emergency in any prior decade.
What happened in 2012?
Smartphone adoption crossed the critical mass for American teenagers. Instagram launched in 2010 and was reaching widespread adoption. Snapchat came in 2011 and The Infinite Scroll feed became the default of adolescent social life, which is to say that the machinery designed to capture attention at scale was now working on them exactly as it was intended to.
This is part 4 of the What Went Wrong series
Part 1: We Forgot We Were Animals
Part 2: The Thirty Months That Broke Everything
Part 3: How Families Disappeared
I notice this stuff in myself, and I think that matters for the argument. I'm an adult with a fully developed prefrontal cortex and reasonable self-control and I still check my phone something like 100 times a day. I catch my hand reaching for it during any pause. Waiting for coffee, drinking coffee, stopped at a red light, the few dead seconds between finishing one thing and starting the next, and (too often) when I should be doing something else.
The capacity I used to have for uninterrupted reading, sitting with a book for three hours without my mind drifting toward a screen, that's gotten noticeably worse over the past decade. I used to do it easily. Now it needs effort and sometimes the effort isn't enough.
I mention this because the common framing treats the attention economy as a youth problem. "The kids and their damn phones." But it hits everyone. I see it in myself, in my parents, in my grandma. But the kids are just the ones who are more vulnerable to it, less equipped to absorb the damage, and less practiced at hiding what it does to them.
Begin with the economics, because the economics are what produce the engineering.
Facebook, Instagram, YouTube, TikTok, 𝕏, they all make money when and if you use them. More money the longer you stay. The most money when you use them compulsively, without thinking about it, in every gap between every other thing you do with your day. Once you strip away the mission statements and congressional testimony, their business model is captured attention sold to advertisers. Everything else is window dressing around that one transaction, and that's really all there is to it.
Once you understand this, what you're looking at is a superstimulus. The attention economy exploits evolved attentional systems at intensities those systems were never built to handle.
Every feature of every platform gets A/B tested against alternatives. Thousands of experiments at once. Variants that increase time-on-platform survive. Everything else gets killed. This is literally how the product gets built, and it has been running continuously for more than a decade now on billions of users. What you see when you open Instagram is the output of an evolutionary process optimized for a single variable: keeping you there.
The kind of engineering that comes out of a process like this works on your nervous system in ways that you did not choose and that you largely cannot override. Your phone buzzes and your attention redirects toward it. That's the orienting response, an automatic redirect toward novel stimuli that evolved because novelty might mean predator, food, or mate. The response is involuntary. You can't opt out of it any more than you can decide not to flinch at a loud noise. Every notification and every refresh of the feed pulls that trigger, and the platforms generate novelty constantly because that's what keeps the circuit firing.
Variable-ratio reinforcement holds you once you're there. This is the reward schedule behind slot machines: you don't know whether this particular pull-to-refresh will show you something good or nothing at all, and that uncertainty is itself more motivating than either outcome alone.
The maybe is the hook. The actual content you end up seeing is almost beside the point.
Social comparison is the thing that runs underneath all of it. We are social animals. Humans are always tracking where they stand relative to the people around them, and this works fine in small groups where the hierarchies are stable and you can actually do something about your position. Social media quantifies the whole process with likes, followers, comments, shares, makes your standing visible to everyone, measurable at all times, updating constantly. You can never fully stop monitoring because the numbers might have moved since you last looked. And then reciprocity traps you further. Someone liked your post, courtesy says acknowledge it. Someone commented, silence reads as rude. Read receipts make avoidance visible. The social obligation machinery keeps generating engagement even when you had every intention of putting the phone down and doing something else with your time.
Chamath Palihapitiya, who ran Facebook's growth team, described what they'd built in 2017: "The short-term, dopamine-driven feedback loops that we have created are destroying how society works." Aza Raskin invented infinite scroll and has called it addictive by design. These are the men who built the thing. They are telling you what it does.
None of the design is accidental. Infinite scroll removes stopping cues. There's no bottom of the page, so you keep going until something in the physical world interrupts you. Autoplay starts the next video without you having to do anything, so continuing costs nothing while stopping requires a conscious choice. The red notification badge was A/B tested until they found the exact shade and shape that captures the most attention. That slight loading delay before content appears is manufactured anticipation, and it turns out that anticipation is where the dopamine is actually doing its thing.
People check their phones roughly 96 times a day. Once every ten waking minutes. Most smartphone users experience phantom vibrations — their pocket vibrating when nothing actually happened — which if you stop and think about it is genuinely insane as something happening to hundreds of millions of people. There's a study from Ward et al. (2017) showing that merely having your phone on the desk reduces cognitive performance. Powered off and screen down. You don't even have to be using it. It's just sitting there and you are already paying a cognitive tax for having it near you.
Context-switching is where most of the cognitive damage accumulates, I think. Every time your attention redirects there's a transition cost. Time lost, performance degraded on both tasks, a mental friction that compounds across dozens of switches per hour. We aren't good at multitasking, it turns out, or at least not in the way we like to think we are. What we're actually good at is rapidly alternating between things while getting worse at all of them.
Sustained reading is deteriorating. I don't have a clean citation for this because it's hard to study rigorously, but anyone who reads seriously has noticed it in themselves. You start a long article, your attention scatters after three paragraphs, your hand moves toward the phone. It just does. The capacity for extended focus atrophies through disuse the same way anything does. Complex cognitive work, the kind that requires you to hold your attention in one place for hours at a time, degrades when interruption is the default state of your waking life. Add to it that most college kids graduate without having ever read a book cover to cover and the picture gets even more grim.
Knowledge work takes the worst of it. Writing, analysis, programming, reading that requires actual thought. All of it depends on sustained attention, and all of it breaks down when your deep work periods are punctuated by context switches you never invited. You work more hours and produce less of value. You can feel that something is off without being able to put your finger on what it is, and that feeling itself is a symptom of the fragmentation.
Relationships get worse in ways that are hard to quantify but easy to observe. Two people in a room, both semi-present, attention partially elsewhere, conversation stays shallow by default. The experience of being truly seen by another person, which requires their full and sustained focus, is becoming rarer. Loneliness goes up even when you are in the same room as people you love. Strange thing to write. But it is what it is.
Adolescents are where all of this turns genuinely dark.
37% of American teenagers had a smartphone in 2012. By 2015 it was 73% (Pew Research). The mental health inflection tracks these adoption curves with uncomfortable precision across multiple independent datasets. You can argue causation if you want to, and many people do, but nobody has produced a competing explanation that fits the timing and demographics equally well.
Girls are more affected than boys by a significant margin, and the reasons for that are plausible even if they're hard to isolate cleanly. Female social competition tends to run through appearance and relational positioning, and Instagram provides infinite material for both of those. Likes and followers make status hierarchies visible in ways they never were before. Male aggression tends to be physical. Female aggression is relational: exclusion, reputation attacks. Social media is a near-perfect weapon for that kind of conflict. It follows you home, there is no escape from it, and the evidence lives on your screen forever. Girls also use social media more than boys do. Boys are more likely to use phones for gaming, which has its own issues but doesn't seem to produce the same comparison spiral.
There's a developmental timing question that worries me more than the rest of this. Adolescent brains are still undergoing major pruning and reorganization, especially in prefrontal regions handling self-regulation. Capturing attention during this developmental window might produce effects that aren't reversible the way they'd be for an adult whose brain is already finished. We don't have the longitudinal data yet. The worry is real and the precautionary case is strong.
Maybe something else happened around 2012 that disproportionately affects girls, coincides with smartphone adoption curves, and shows up in objective measures like hospitalizations and completed suicides. It's possible. I suppose it could be something else. But the people making that argument have yet to tell me what that something else actually is.
There's also the individual variation objection: most teenagers with smartphones are fine, so if phones were truly toxic, shouldn't the effects be more universal? This is fair. Some kids are more vulnerable than others, by temperament or pre-existing conditions or circumstances. The environment shifted in a way that is harmful on average without being harmful to every single person in it. Like urban air quality. Most people who breathe city air don't get lung cancer. That doesn't mean the air is safe for them to be breathing.
I'm not sure "addiction" is even the right word for most of this. Addiction implies a substance you can choose not to consume. Your phone is your alarm clock, your map, your camera, your connection to every person you care about. Opting out entirely isn't really an option nowadays. And that makes the whole question harder than it would otherwise be: what kind of use do you actually let yourself have? That's less satisfying than abstinence as a way of thinking about it, and also more honest.
So what do you do?
I'm genuinely uncertain how much individual action can do against an infrastructure this powerful. But I've been doing some experimenting with my own setup and there are a few things that seem to actually work.
Paying attention to what it is that captures yours is the obvious starting point. Which apps trigger compulsive checking. Which notifications you respond to reflexively. Document it honestly. Most people are surprised by what they find.
I pulled the addictive apps off my phone about a year ago. Social media, news feeds, anything with an infinite scroll. You can still access all of it on a computer when you want. And the difference is crazy, because on your phone it exploits every mechanism I just described, but on a computer it's just a website you can close whenever you want to.
Notifications: turned off everything except messages from people I want to hear from. The default state is silence, 24/7. Very few things genuinely merit interrupting whatever you're doing.
Physical spaces matter. Bedroom without the phone, dinner table without it, first hour of the morning without it, gym without it. I'm not perfect about this because you can't perfectly plan around life. But the zones exist and when I'm in them my sleep and my conversations are measurably better, and the quality of my morning is different, so are my lifts at the gym. The moment you start negotiating with yourself about exceptions is the moment you lose them.
For work you need structural barriers because willpower alone is inadequate against something engineered to be compulsive. Website blockers during focused periods, scheduled email checks, protected deep work time that you actually defend against interruption. Systems over discipline.
If you have kids, the urgency is different. Delay smartphones as long as you can. Every year of development without constant attentional capture matters. Dumb phone if they need one. Curate the technology environment like you'd curate their diet, because that is really what it is — an environment that they are growing up inside of. The whole "but their friends already have it" is an insane cope, or just an excuse to be lazy. If their friends were doing coke, you wouldn't let them.
Practice sitting with nothing to do. Get bored again. No phone, no book, no computer, nothing. The discomfort is withdrawal and it passes. On the other side of it is a nervous system that can actually settle down and do what it's supposed to do.
Try going screen-free for twenty-four hours sometime. The first time is difficult in a way that tells you something about where you are.
The attention economy was built to capture and monetize human attention without any regard for what the capture costs the person on the other end of it. The engineers knew what they were making. Some of them said so publicly, on the record. They made it anyway.
The teenagers stopped being okay around 2012. That's when the machinery reached scale. The adults aren't doing much better either, for the most part. They are just more practiced at living with the dysfunction. More used to mistaking scattered attention for what life is supposed to feel like.
The machinery won't dismantle itself. And waiting for someone else to come along and fix it is, if you think about it, itself a form of captured attention. You're directing yours at hoping instead of doing anything about it. So what does that leave you with?
What remains is defending your own cognition. Your attention is the only resource you have that actually determines what kind of life you end up with. The environment that you live in by default is adversarial by design now. You can call that a paranoid way of looking at it if you want to. I'd call it reading the room.
