Unlearning
I got my first phone at thirteen or fourteen, later than anyone I knew, and spent the better part of middle school relying on other people’s parents and my rigid schedule to get to and from school.
My phone lived in a cubby in the mudroom. It wasn’t allowed in my room, the family room or the kitchen, so if I wanted it, I had to go to the mudroom to use it. For the next several years I was at war with the cubby. I snuck the phone upstairs, got caught, got yelled at, and snuck it up again. My parents didn’t budge. As a senior I was still sneaking it up to my room and still getting caught. Before that, I’d watched my sister leave hers in the cubby when she was home on break from college.
Nobody told us why, and our parents never offered. Ours was a house with rules for the dishes, the trash and the common spaces, and the cubby was just one more.
I don’t regret fighting the cubby. But in all those years of sneaking the phone upstairs, I never once asked why the rule existed.
The rule finally made sense in college. On movie nights in the common room, my friends would last twenty minutes before a screen lit up in somebody’s lap, then another, until half the room was watching two things at once. Hardly anyone made it through a whole movie. Somewhere in there I realized I didn’t feel the pull of that six-inch screen the way everyone around me did. My phone had spent most of its life in a cubby, and a habit never had a chance to form.
It wasn’t only movie nights. Phones came out at dinner, under the desk in class, halfway through someone else’s sentence. A lot of conversations ran alongside a dozen text threads. Nobody treated it as rude, because nobody had been taught that it was.
Looking back, I think the cubby had less to do with phones than with manners, though I doubt my parents would have put it that way. They were teaching me that the person in front of you gets your attention, and they taught it the only way that works on a fourteen-year-old, by taking away the alternative.
For years my mom and I had the same argument about the phone in my pocket. She thought people would look back on what phones did to kids the way we now look back on cigarettes in the 1970s and '80s. I always scoffed and called it ludicrous. How could my phone be anything like a cigarette? Then this summer Meta agreed to pay more than $16 billion to settle states’ claims that it had hooked kids on its platforms, and to change how its apps work for children. I realized she had a point.
I deleted social media from my phone a while ago, and the cubby stayed, years after anyone stopped enforcing it. The rule never changed. I did.
The cubby isn’t special. You’ll be told things your whole life that you don’t agree with, by parents and teachers and bosses, and you’ll do plenty of them anyway. That part is rarely your choice. What you do with them afterward is. Some prove right years later, the way the cubby did. Others prove wrong, and those are harder to drop, because by then you’ve forgotten anyone told them to you.
At Harvard, salary came up in the first week and never went away. I learned which firms paid “one-twenty” and which paid only “one-oh-five,” and I learned to say “only.” When someone landed a summer trading job at a top quant fund that paid thirty-five thousand, everyone who heard about it wished it were them, me included. By senior year I could tell you roughly what every bank and consulting firm paid. I couldn’t have told you what anyone in those jobs did all day.
I wanted the biggest number I could get, the way a kid flips a graded paper straight to the score and never reads the comments. Unlike my cubby, this lesson had no parent behind it to resent and nobody yelling when I broke it. It came from classmates comparing offers, from recruiters, from alumni I barely knew, so it felt less like a rule than like the way the world worked. It was comfortable, too. One figure told me how I was doing, which saved me from deciding what doing well meant.
In September of my senior year I published a column called Metric Mindset, arguing that people treat salary as a stand-in for meaning and that a number with no ceiling leaves you permanently unsatisfied. I believed it while I wrote it, and it changed nothing. Whenever a job came up, the first question anyone asked was what it paid, and I answered it like everyone else. Maybe that’s capitalism. It was also just how everyone talked.
My parents told me constantly that pay wasn’t the point, and so did my mentors. I heard it on calls home and over coffee with people a decade into careers I admired, and it sank in the way the cubby had, slowly.
The decision came during a stretch of 2 a.m. nights doing busywork one person kept handing me, with no time or room for anything else. For something I’d gone back and forth on for months, the conversation itself didn’t take long. I told one of the people who ran the firm where I'd spent two summers, one of the sharpest people I’ve ever met, that I wouldn’t be coming back, and that was that.
The rest of senior year was harder than I expected. Every dinner seemed to include someone asking where I’d landed, and for months I didn’t have an answer. The pressure only grew, and I started to think I’d made the wrong move. Mostly I kept quiet about it.
I took a job at a startup instead. Nobody there hands me a process, so I have to work most things out myself, and I’ve learned more in a few months than I did in those summers of private credit. I’m glad I made the change.
Writing that column hadn’t been enough. I could explain exactly why the number was a bad way to choose a job and still choose it. Knowing a lesson is wrong is the easy part. Actually dropping it took months, and the same few people telling me the same thing until I believed them.
Nobody had to order us into the number. The people around us took care of that, mostly without meaning to. There was an unspoken rule against admitting you weren’t working, or that you were bad at something, or (worst of all) that you had no idea what you wanted, and a good offer was the cleanest proof that you were fine. I kept to it in public as well as anyone. That, in fairness, is how a rule like that lasts. It doesn’t need anyone to enforce it.
The real cost is that you can’t unlearn something you can’t say out loud. “I don’t know if I want this” is where changing your mind starts, and it was the one sentence I never heard anyone say. So the number answered for us.
My parents’ control was a different animal. It came from people who knew me, and at fourteen I needed it. That kind of control is good for you right up until you know enough to choose for yourself, and the hard part is that no one marks the moment. The people profiting from you have no reason to, and the people who love you tend to notice late. You have to decide on your own that you’ve outgrown a rule, and it’s always easier to put that off.
When I finally decided, it didn’t feel brave. The cubby made sense slowly, over years of movie nights, and the number took longer than that. Both needed time and a tolerance for not knowing, which is exactly what we’re running out of.
In 1784, the German philosopher Immanuel Kant published his answer to a question raised in a Berlin journal: What is enlightenment? Part of his answer was that he did not live in an enlightened age but in “an age of enlightenment,” one where the obstacles to people thinking for themselves were slowly coming down. Almost 250 years later, those obstacles are growing back. Any question I have can be answered in the time it takes to type it, usually before I’ve spent a minute wondering about it myself.
Machine learning is the clearest example. A model learns by being shown the right answer over and over until its output matches, and the best ones have gotten startlingly good at it. In September, OpenAI said one of its models had produced a proof for the Navier–Stokes problem, one of seven math problems with a $1 million prize attached.
But people chose the problem, and the method it built on came from human mathematicians who had been working on it for years. A model can answer almost anything put in front of it. It never asks a question of its own, or doubts the rules it was trained on. If I had only ever obeyed my parents’ rule, that’s what I would have learned too: the habit without reason. And a model like that is now where a lot of people go first for answers, and most of them treat what comes back as correct.
We’ve started rewarding the same thing in people. The jobs that pay people my age the most reward doing what you’re told, quickly, without fuss. I spent a good part of two summers formatting credit memos, updating spreadsheets and checking figures that someone more senior would check again. A chatbot can already do most of that work, and a person who has only been trained to do it is easy to replace.
Kant had a word for whoever makes dependence comfortable. He called them Vormünder, the German term for a child’s legal guardians, and he argued that most people stay in their care because it’s easier than thinking. If an app picks where you eat and a map tells you every turn, you can get through a day without deciding much at all. As human consciousness matures, the feed and the chatbot are the most patient guardians anyone has built, because unlike a parent, they never want you to outgrow them.
Open a feed and it shows you more of whatever people like you already liked. Ask a chatbot something and it gives you the likeliest answer to what you typed. Both hand you the average, and they hand it over instantly, before anything has a chance to develop in you. Ask one what you should want from a career and you’ll get the consensus, beautifully phrased. It’s the same lesson I spent years unlearning, now available on demand.
You can see the result without looking at a screen. Spend a Saturday walking around New York and count the matching puffers and sneakers that pass you, until the faces start to blur together. The people who don’t fit are still out there, and they matter more than they look, because they tend to be the ones who notice something first.
My friends show it too, even at Harvard. You’d be surprised how few of them read books, and fewer still have something of their own they’re deeply into. Screens take most of the day, and between AI and social media, a lot of us come away drained and a little more alike. People are less sharp than they used to be. We’re thinking less and valuing the wrong things.
Kant had one more hope: the public. Give people the freedom to argue, he said, and their enlightenment becomes almost inevitable: a few independent minds start thinking for themselves, and the rest follow. His bargain was to obey at work and argue freely in public, in your own name, to anyone who would read you.
At Harvard, class was supposed to be that kind of public, and it rarely was. Most people kept quiet rather than risk being wrong in front of everyone, because whoever said something half-formed would get picked apart behind their back after class. The ones who did talk often hadn’t done the reading, and they would try to convince the room that what they were saying was fact, with no real idea what the book argued. So most discussions said very little. The strange part, I think, is how much people want the opposite. Almost everyone I know is tired of hearing the same opinion repeated back to them, and would rather hear someone say something different and stand behind it in their own name.
What gets said in public instead usually sounds like a sermon. Kant held that a clergyman preaching to his congregation isn’t reasoning publicly at all, however large the crowd, because he speaks on someone else’s commission. Every fall of college my feed filled with the same post: thrilled to announce, humbled and grateful, a logo and a start date, in a voice that belonged to Claude. Kant would have recognized the pulpit. Our audience rewards one voice, platforms decide who hears it, and at work we still do as we’re told, so his bargain has failed at both ends. Nor is there a single public left to wake. There are thousands of small ones that never hear each other, and nobody today can wake them all at once.
No one can save us, so it falls to each person to do this on their own, though on your own doesn’t mean without help. The best lessons I’ve had came from my parents and my mentors, and I still ask people who know more than me constantly. Kant wanted us to think without anyone’s direction, and I doubt anyone ever has. What I can do is pay attention to who is teaching me and what they get out of it, and give the lessons I don’t understand yet some time. It helps to know what I’m protecting, and right now that’s the people in my life and my time.
It took me about eight years to understand a rule about where my phone went at night. In a few years I’ll reread this and find something in it I have to unlearn too. I can’t say what yet. If I could see it, I’d have dropped it already.
A while ago I told my parents they were right about the cubby. They laughed, said they’re always right, and asked whether I had any idea how hard I’d pushed on those rules. I do now.
Thoughts? Reply by email.