This is a rarer gift than it first appears.

To sit beside a stranger at dinner; to find yourself in conversation with a fellow passenger on a train; to encounter a friend of a friend at a crowded gathering and ask them, with genuine interest rather than social obligation, who they are and how they came to be the particular and improbable creature standing before you; these things ought to be among the most ordinary experiences of life, and yet they are becoming, I think, increasingly unusual. We have become accustomed to treating other people either as obstacles to be avoided or opportunities to be exploited.

I realised this several months ago while attending a (though I despise the term) networking event in a room filled with exactly the people you might expect. There were polished shoes and polished conversations taking place in little circles around the room, with students exchanging LinkedIn profiles with all the romance of neighbouring kingdoms exchanging trade agreements. It was impossible not to notice how much effort was being expended in appearing interesting and how little in being interested.

Everywhere you looked, conversations were collapsing beneath the weight of self presentation. Questions existed just to create opportunities for answers. People were not speaking to one another so much as advertising themselves in one another’s direction.

Subscribe now

And yet, in the corner of the room, there was an older gentleman whose name I never learned. For almost an hour he spoke with a succession of strangers and not once, as far as I could tell, did he mention his own profession, achievements, opinions, travels, qualifications, investments, ambitions or connections.

He asked questions. Endless questions. Thoughtful questions.

When he discovered that I was, in part, studying urban planning, he wanted to know why certain cities seemed to possess personalities while others felt interchangeable. When a student mentioned learning Farsi, he became fascinated by the peculiar emotional atmosphere of languages and whether thoughts changed shape when translated. When somebody remarked that they had recently returned from Norway, he spent a while asking about winter darkness, not because he intended to visit Norway himself but because the phenomenon interested him.

What struck me was not that he appeared extroverted.

He did not.

Yet by the end of the evening more people had gravitated towards him than towards anyone else in the room.

Our modern conception of extroversion is theatrical. We imagine the extrovert as a performer; a person of inexhaustible confidence who enters rooms as storms enter coastlines, announcing themselves through force of personality.

The introvert, meanwhile, is imagined as their opposite; withdrawn, hesitant, private, content with solitude.

Entire industries have emerged around this distinction. Books have been written. Personality tests administered. Identities constructed. One is almost tempted to believe that humanity consists of only two tribes who happen to share a planet.

But I am not convinced.

There are people who can address an audience of five hundred and yet struggle to speak honestly with one stranger. There are people who claim to dislike socialising but who, when they encounter a topic they genuinely care about, will spend four uninterrupted hours discussing it with complete absorption.

Perhaps extroversion has been misdiagnosed.

Perhaps it is not a quantity of speech but a direction of attention.

This possibility interests me because it transforms extroversion from a personality trait into something approaching an art; and the arts, unlike traits, can be cultivated.

The greatest conversationalists I have known possessed an almost childlike relationship with the world. They retained the assumption that everything contained something worth discovering. Most adults lose this assumption. Familiarity settles over life like dust over furniture. We stop asking why things are the way they are, and this causes the world to narrow. People narrow with it.

Children, by contrast, are natural extroverts in the deepest sense of the word. This is not rooted in confidence, but in curiosity. They approach existence as explorers rather than judges.

What if the art of extroversion is simply the refusal to outgrow this habit?

I sometimes think that intellectual life and social life are far less separate than they appear. We tend to imagine curiosity as something exercised upon books, theories, historical events, scientific discoveries; yet people are perhaps the most fascinating subject available to us.

Every human being is a library assembled from experiences no one else has lived through. Entire continents of memory remain hidden behind ordinary faces. To meet somebody is to stand before an unopened book whose contents are inaccessible unless sufficient patience and interest are applied.

This, it seems to me, is where genuine extroversion lies.

It is not found in confidence.

Nor is it found in charisma.

It is certainly not found in networking.

But, it is found in fascination.

The fascinating thing about fascination is that it leaves very little room for self consciousness. The moment you become interested in another person, your attention departs from yourself. You cease monitoring your own behaviour because something more compelling has captured your attention.

And perhaps that is why the most socially gifted people often appear so effortless: they are not thinking about themselves.

The rest of us are.

I hope they can add value to yours.

1. Never leave something you love to chance. The world is a gallery, and every wall is already hung with someone else’s passion made visible. The things you love privately, the projects you turn over in your head on the walk home, do not exist for anyone until you put them out where they can be seen. Chance is not a curator. If you want the thing you care about to have a life beyond your own attention, you have to be the one who hangs it on the wall.

2. Build the identity before you feel you have earned it. You will wait forever for permission to call yourself the thing you are becoming. The writer who has not been published; the influencer with only two videos; the founder before the company works; the person who is not yet who they intend to be. The waiting is the trap. You become the person you want to be by acting as that person, and then one day the title catches up to the behaviour.

3. Choose your problems, because you do not get a life without them. The fantasy is a life with the problems removed. It does not exist. Every path comes with its own set of difficulties, and maturity is found in selecting which ones you are willing to keep. The right question to ask yourself is never “what do I want to enjoy” but “what am I willing to suffer for.”

4. Follow your curiosity even when it does not look useful. The most interesting people are assembled from a hundred apparently pointless fascinations that turned out, much later, to connect. You cannot see the connections in advance, so the instinct to prune everything that is not immediately productive is a kind of self-impoverishment.

5. Be willing to be a beginner in public. The fear of looking foolish keeps more people small than any lack of talent. Every skill you envy in someone else was preceded by a period of being visibly bad at it, and they survived that period because they cared more about the skill than about how they appeared while acquiring it.

6. Distrust the comfort of cynicism. Cynicism feels like intelligence because it is unfalsifiable and it never gets disappointed. But it is mostly a defence, a way of deciding in advance that nothing was worth hoping for so that nothing can hurt you. It is easier to sneer than to try. The braver and rarer thing is to stay capable of sincerity and to care openly about something that might not work, knowing the sneer is always available and refusing it.

7. Let your standards be high and your self-judgement be kind. These two are not in tension, though most people collapse them into one. You can hold the work to an exacting standard while holding yourself with patience, the way a good teacher does. The person who lowers their standards to feel better never makes anything good. The person who flogs themselves for every failure burns out before they make anything at all. Aim for the mountain summit; be gentle with the climber.

8. Stop optimising for a version of yourself you will not recognise. You are constantly making sacrifices in the present moment on behalf of a future self. This is often a wise thing to do, but it is pertinent to remember that self is not a stranger, and they will not thank you for a life spent entirely in preparation. Live partly for the person you are now, because the future is not some destination you can arrive at. It is just more present, later.

9. Remember that you are going to die, and let it sharpen rather than paralyse you. Everything above gets its urgency from a single fact: you have a finite and unknown number of mornings. Held wrongly, that thought is a weight. Held rightly, it is a knife that cuts away everything that does not matter and leaves you holding only the things that do.

keep those principles that resonate with you, discard those that don’t :)

Share

This is why you remember the email you never sent more vividly than the twenty you did. It is why the unread chapter sits heavier than the shelf of books behind it, or why the project you abandoned at sixty percent still occupies mental space while the ones you shipped have gone silent.

Completion does not feel like victory so much as disappearance. The thing that occupied you simply stops occupying you.

I want to take a look at your intuition here, because I think it sits on top of two separate truths: one about how the mind is built, and one about what a life is for, and they turn out to be the same truth seen from two sides.

In 1927 a Lithuanian psychologist named Bluma Zeigarnik, working under the Gestalt theorist Kurt Lewin, noticed something about waiters in a Berlin café. They could hold an unpaid order in memory effortlessly and forgot it the instant the bill was settled. Zeigarnik took this into the lab, gave participants small tasks and interrupted them partway through about half, and found that the interrupted tasks were recalled markedly better than the completed ones. Her explanation, in Lewin’s terms, was that an intention sets up a state of psychological tension that persists until the task is resolved. The open task stays loud because, in a literal sense, it is still running. Completion switches it off.

Notice what this means at the level of experience. The mind does not file a finished thing under “achieved” and feel satisfaction. It files it under “closed” and releases it. The reward you imagined would arrive on completion was already being paid out, in the form of attention and aliveness, for as long as the thing stayed open. Finishing does not deliver that. Finishing ends it.

Subscribe now

The effect is not evenly distributed, and the asymmetry is unkind. The American psychologist John Atkinson, working on the motivational side, found that the people who felt the pull of an unfinished task most strongly were precisely those who had most wanted to finish it. The tension scales with the investment.

So the person most haunted by open loops is the person most alive to their projects. This is the first hint that the Zeigarnik tension is not a bug in an otherwise efficient system. It looks instead like the felt signature of being gripped by something, and being gripped by something is not obviously a state we should want to escape.

There is a temptation, especially in the productivity literature, to read this cognitive desire to complete purely as friction, or some sort of tax on cognitive bandwidth to be minimised. But a tax is something you pay for nothing. This is not that. The attention an open project draws is the same attention that, while you are inside the work, you would call absorption or even flow. The noise after you close the laptop and the focus while you are at the desk are not two phenomena. They are one orientation, sampled at two different moments. To want the after-noise gone entirely is to want, without quite realising it, the orientation gone too.

Here the philosophy has to carry equal weight, because the science describes the mechanism but cannot tell us what it is worth.

Bernard Williams, in his 1973 essay “The Makropulos Case”, drew a distinction that maps onto all of this. Some of our desires are merely conditional: if I am going to keep living, then I want to fix the cavity, do the dishes, answer the email. These do not give me a reason to live; they presuppose that I am living anyway.

But other desires are categorical. They do not hang from the assumption of my own existence. They are the desires (to finish the novel, to see the child grow, to build the thing) that propel a person forward into the future and, Williams argued, supply the only real reason anyone has to go on at all.

A life is structured around these categorical desires and the ground projects they generate. They are what makes death an evil, because death would leave them unsatisfied.

Put Williams next to Zeigarnik and you get an interesting synthesis. The open loop that hums and will not quiet, meaning, the unfinished task that occupies you, is the psychological shadow cast by a categorical desire. The tension is the felt evidence that you are still propelled, still oriented toward a future you have not yet reached. The reason an unfinished project keeps you company is that it is still a reason to live, and the finished project has stopped being one.

This is why completion can feel like a small death. Williams saw the large version of the point: he argued that an immortal life would eventually become unliveable, because a person who outran all their categorical desires, and who finished everything, would be left with nothing pulling them forward, and would arrive at a boredom no new experience could touch.

The terror is not having unmet desires. The terror is running out of them.

the unfinished task is not the obstacle to the life. it is, in a small way, the evidence that there is one.

---

None of this means the noise is benign. The same tension that signals a live project becomes, past a certain volume, a kind of suffering. Each open loop demands its small allocation of attention, the demands run in parallel whether or not you can act, and the result is the background overload that has nothing to do with how hard you are working.

Worse, when an open loop meets dread, it feeds itself: the tension keeps the task active, the active task raises the load, the raised load makes the task more aversive, the aversion deepens the avoidance. Procrastination is being trapped inside Zeigarnik tension.

The obvious way out is to finish everything, discharge every loop, and arrive at silence. But this is exactly the Williams nightmare in miniature, and it is also impossible. You cannot finish the dissertation tonight to stop thinking about it, and if completion were the only release, the conscientious person would be condemned to the noise forever.

The useful finding comes from Roy Baumeister and E. J. Masicampo, in a 2011 paper titled “Consider It Done!”. They showed that the mind does not require a task to be finished in order to release its grip. It requires the goal to feel resolved, and a sufficiently concrete plan counts as resolution. When participants made a specific plan for when and how they would complete an unfinished task, the intrusive thoughts dropped sharply, though not one additional thing had been done. The mechanism connects to Peter Gollwitzer’s implementation intentions, the if-then structures that decide the moment of action in advance: a vague “I’ll get to it” leaves the loop open, while “when I sit down at nine, I open this first” gives the system the one thing it was actually asking for, which was never completion but the credible assurance that completion is handled.

This is the move that lets you keep the categorical desire without drowning in its shadow. You are not extinguishing the project, which would be a small death of its own, and instead you are giving the tension a place to rest so the desire can stay alive without screaming.

I should add the honest complication that the Zeigarnik effect has had a hard time in the replication literature; a 2025 systematic review and meta-analysis found no reliable memory advantage for unfinished tasks at all, though it confirmed the related pull to resume interrupted ones. The tidy textbook claim therefore wobbles. What survives is the part Williams was pointing at independently: that we are creatures oriented by unfinished projects, and that the orientation, not its discharge, is where the living happens.

So I have stopped treating the hum as a problem to be eliminated. It is the sound a categorical desire makes while it is still unmet, which is to say while it is still doing its job. The task is not to silence it but to keep it from becoming a scream, and the difference between those two is usually just a plan.

You do not actually want to finish everything. A person who has finished everything has run out of reasons to begin.

erin x

everyone is seeing through the ‘flaws in the conventional wisdom’.

everyone has the angle that the mainstream has constantly missed.

everyone has the willingness to say the thing that supposedly nobody is allowed to say.

It is slightly exhausting. But it is also more than exhausting: it is dangerous.

there is, amongst certain groups that I have inevitably found myself operating within, an almost universal self-identification as the contrarian. there is a deep desire to be the person who: thinks for themselves; doesn’t follow the crowd; has arrived at their positions through independent reasoning rather than social absorption; is willing to be the penguin with its heart set on the mountains (sorry if you’re not chronically online to understand that reference; I am referring to this from Werner Herzog’s documentary).

what this means is that independent thinking has become the crowd position. contrarianism is the new conformism, and most of the people performing it have not noticed this because noticing it would require them to apply to themselves the same scrutiny they’re busy applying to everyone else.

this would be merely ironic if it stopped there. but it doesn’t, because the mass adoption of the contrarian identity goes beyond the irony. it has made conformity into a dirty word at the exact moment when what we most need is a more true account of what conformity is actually for.

conformity has, for at least the last several decades, been the villain of almost every cultural narrative we produce.

the suburb as hell.

the 9 to 5 as tragedy.

the person who followed the expected path as the cautionary tale against which the adventurous individualist is defined.

too much groupthink, of course, causes genuine harm. people should not be encouraged to subordinate their better judgement to the perceived consensus as this inevitably produces bad outcomes for the majority.

the capacity to dissent from a wrong majority is valuable and rare.

conformity is not always good and dissent is not always bad.

however the version of conformity that has become culturally toxic is not groupthink. it is unfortunately something much more ordinary and much more necessary, which is the willingness to coordinate with other people and uphold shared norms even when defecting from them would benefit you individually.

this is the conformity that makes trust possible.

it is the conformity that makes institutions function, communities cohere, and collective projects get off the ground. it is currently caught in the crossfire of a culture war.

the lone wolf is the fantasy that sits at the heart of this. I don’t think this is actually a model of independence. it is a model of parasitism.

the person who opts out of all social norms and who goes their own way regardless of what the social fabric requires, is surviving on the conformity of everyone around them. they are free-riding on the trust built by people who kept their commitments even when it cost them something.

the lone wolf only exists because the pack does.

the penguin that leaves for the mountains is not pioneering a new form of life; it is a penguin that will die in the wrong climate while congratulating itself on its uniqueness, having contributed nothing to the colony that made its existence possible.

Subscribe now

what has accelerated all of this is, predictably, the architecture of the internet. platforms reward differentiation as it generates engagement in a way that considered agreement does not, because agreement is boring and disagreement is stimulating.

the result is a commercial incentive structure that selects for contrarianism regardless of whether the contrarian position has been arrived at through the use of one’s brain or just through the calculation that going against the grain will generate attention.

and so we have produced an ecosystem of people who have learned to perform independence.

dear reader, here is the thing i find most troubling about this: the social fabric is not a metaphor, much as your english teacher may attempt to convince you that it is. it is, instead, a description of the set of shared norms and mutual expectations and accumulated trust that allows people to cooperate and coordinate around shared problems that no single person can solve alone.

that fabric is built through countless small acts of conformity that produce an understanding that the norm serves a function that transcends any individual’s interest in defecting from it.

what i want to say, and i am aware this will sound unfashionable to the point of sounding naive, is that we should be more willing to conform. convention is not always right and institutions are always deserving of loyalty, but the alternative (which we are in the middle of) is not a society of free individuals but a society of isolated ones, each performing independence while the infrastructure of collective life that makes individual flourishing possible slowly loses the thing that sustains it, which is the willingness of enough people to act as if it matters.

the lone wolf is a compelling image. it is also a description of an animal that is alone and will remain alone, making no contribution to any pack and receiving none from one, until it dies.

that is not freedom. it is just a different and lonelier kind of conformity, to the idea that you owe nothing to anyone, which has somehow become the most conventional position of all.

I don’t mean your preferences for those general things you were born with some disposition toward, but rather i’m interested in the more specific ones: the aesthetic sensibilities; the values; the sense of what a good life looks like; the things you find yourself wanting.

those preferences were formed somewhere, and formed by something, and the question of what formed them is one of the most practically important questions you can ask about yourself. your preferences are not incidental to who you are. they are, in a very direct sense, what you are. in some sense, the capacity to form genuine values and to find some things beautiful and others not, is among the most distinctly human capacities there is. it is what consciousness is for.

this is the background as to why the current situation is worth more of your time than mere mild annoyance. a reasonable estimate holds that approximately 90% of the content that passes through your eyes online is an advertisement. these are not necessarily direct ads with a sponsorship tag and a discount code at the bottom, but content that exists within the commercial logic of the platform you’re on. what i mean by this is that the content was specifically produced by someone who has a financial incentive to make you want something in a way which operates on your desires and values to serve someone else’s interests before it serves your own.

the influencer posting her morning routine is selling you a version of life that makes her brand partnerships viable. the founder writing about his productivity system is building an audience that becomes a funnel. the wellness account making you feel insufficient is running the oldest advertising logic in the world, which is to produce a dissatisfaction that its product can then resolve.

the algorithm that selected all of this for your attention is itself optimised not for your genuine preference but for engagement, which tends to correlate with anxiety and desire and comparison rather than with anything that would actually make your life better.

none of this is hidden. people know, in a theoretical way, that they are being sold to. but knowing in a theoretical way is different from understanding what this actually means.

advertising, in its traditional form, was trying to sell you a specific product. it colonised public space first, then private space through the television, then intimate space through the phone in your pocket and on your bedside table. throughout this time it was always, at least, legible as advertising. you knew the billboard was trying to sell you something. the relationship was clear.

what has changed is not just the scale of advertisement, though it would be remiss to quantify the scale as anything but staggering, but the texture has also changed. what the creator economy and the algorithmic platforms on which it thrives have produced is an environment in which the commercial motive is woven through content that presents as personal expression, as friendship even. the girl showing you her apartment is trying to sell you her aesthetic, which is sponsored by the rug company and the candle brand and the coffee subscription, but the register of the presentation is intimacy.

the format is “this is my life” and the function is “this could be yours, if you want it the right way.” and you cannot fully filter these two things apart, which means that when you notice yourself wanting the life you just saw, you genuinely cannot be certain whether that want is yours.

it sounds a little paranoid when stated directly, but it is not paranoid. it is just an accurate description of how desire formation works when the environment that forms it has been optimised by commercial interests. you did not choose what you were shown, nor did you choose the aesthetic regime that presented certain lives as aspirational and others as insufficient. you did not choose the repeated exposure to a particular version of what is beautiful or what is successful.

and yet, those exposures have shaped what you find yourself wanting, in the same way that any sufficiently repeated experience shapes what you come to prefer.

the advertising industry understood this before the internet existed. the point of advertising was never only to sell a product; it was to shape what people thought a desirable life looked like, so that the product could be positioned as part of that life. what the internet did was scale this logic past any previous limit and remove almost all the friction that once existed between the commercial motive and your inner life.

dear reader, here is the part that i think actually matters: if the mechanism by which you form values is corrupted, then the values themselves are compromised. forming values is not a peripheral human activity. it is not something you do in addition to living your life. it is the structure through which you understand what your life is for, what is worth pursuing, and what deserves your limited and precious attention. if that process is being systematically shaped by entities whose interest in you is entirely commercial, then the self that results, the one doing the wanting and the valuing and the caring, is partly a product of those interests rather than of your own genuine encounter with the world.

this is what i mean by the destruction of individuality. the concern is not that you bought a face cream you didn’t need. the concern is that you can no longer tell with confidence what you would have found worth pursuing in the absence of an environment specifically designed to make you want things that benefit someone else.

the individual is someone who has their own relationship to value, arrived at through their own experience and reflection. that process requires a space from commercial pressure, which has, almost entirely, disappeared.

what you might argue in response is that humans have always been social creatures whose preferences were formed by their environment. culture has always shaped desire and culture has always been entangled with power. this is true. but there is a difference between preferences formed through social life and exposure to art and ideas and the lived experience of other people, and preferences formed through an environment that has been engineered to produce maximum engagement and maximum commercial value.

the first can develop something that resembles individuality, even if it is socially shaped. the second produces a convergence toward whatever version of life is most easily monetised, which turns out to be a fairly limited repertoire of aspirations.

there is also something especially corrosive about the intimacy of the format. when the mechanism shaping your desires presents itself as a friend, you lose the ordinary defences that protect you from advertising’s more obvious forms.

Subscribe now

i am not arguing for some imagined pre-commercial purity that never existed. the first choice available to you is to notice what you are inside of. this will not fully undo the shaping. but it creates a small gap between the stimulus and the response, and that gap is where something like preference formation can still happen.

the second thing worth saying is that what is at stake is not trivial. the capacity to form your own values, in real rather than engineered contact with the world, is not one capacity among many. it is what makes you a person rather than a demographic, or, more precisely, an individual rather than a surface for someone else’s commercial interest. the colonisation of that capacity is something you are allowed to find alarming. i think the appropriate response to it probably is alarm, and then something more useful than alarm, but alarm first.

this, i think, is exactly why the game metaphor breaks down when you apply it to life: life does not come with a win condition pre-installed, which means you have to choose one, which means the most important strategic question you’ll ever face is not how to play well but what you are even playing for.

most people never answer this question, not from a lack of curiosity or seriousness, but because it resists the kind of thinking that gets you somewhere with most other questions. you can work hard at a career question and get a dream job. you can research a decision and improve it. but sitting down and asking “what do i actually want my life to look like” tends to produce either a blank, or a list of things you suspect you’ve absorbed from somewhere external without having chosen them, or something that sounds suspiciously like a linkedin ‘about’ section.

to get around this, people optimise for proxies. (1) money, because money is legible and measurable and socially respected and, more importantly, because at some point someone told you it was what you were supposed to want. (2) status, for similar reasons. (3) career advancement, which is a proxy for money and status wrapped in a narrative. (4) a certain kind of relationship, (5) a certain kind of apartment, (6) a certain postcode. to be clear, these are useful proxies: money buys optionality; status opens doors; a good apartment does improve your daily life. but a proxy is only useful if it is reliably pointing at the thing you actually want, and most of the time people have not checked whether their proxies are still pointing in the right direction, or indeed whether they ever were.

Subscribe now

there is a technique economists use called preference revelation, which is the idea that what you actually value is not what you say you value but what your choices reveal that you value. so one way to answer the question of what you are optimising for is to look at what you sacrifice things for. more specifically, you should look at what you choose when the choice costs you something. that’s the actual optimisation target. when people do this exercise openly and honestly, the results are often somewhat embarrassing, because what you are revealed to be optimising for is frequently not what you would choose if someone handed you the question directly.

a lot of people are revealed to be optimising primarily for the avoidance of anxiety, as opposed to things we might traditionally associate with the elusive ‘meaning of life’ like happiness, meaning, or success. a lot of people have optimised for the relief of not having anything urgent to worry about. this produces a set of choices that look, from the outside, like ambition, but are actually structured around eliminating threats rather than building toward anything. you work hard to be financially secure, for example, not because you have a vision of what security is for but because financial insecurity produces a dread that you will do anything to avoid. you stay in the job that is fine, as another example, because leaving would require tolerating a period of uncertainty that is intolerable. the consistent goal is the anxiety going away.

this would be fine if people named it as their goal. instead they name the proxies. they tell themselves and others that they are working toward something, and yet all the while the gap between the stated goal and the revealed goal sits there unchanging. this goes a long way to explain why, so often, ambition rarely produces people who arrive at the thing they said they wanted.

the self-help question here is “how do you figure out your values and then align your actions with them?”, and it has produced an enormous market of journaling prompts but not noticeably improved anything else. the interesting question is: when you try to identify your actual end goal, the thing you want for its own sake rather than for what it gets you, what do you find?

most people find something like: i want to feel that my life was worth living. or: i want to be close to people i love. or: i want to feel that i did something that mattered. these are perfectly good end goals, but they have a property that makes them very difficult to optimise for, which is that they cannot be chased directly. happiness, meaning, connection, the sense that your life has been well-spent - these are all things that arrive as byproducts of doing something else, and they tend to vanish or become meaningfully hollow the moment you make them the direct object of your attention. you cannot pursue meaning the way you might pursue a promotion.

the people who spend their lives asking whether their life is meaningful are usually the same people who feel that it isn’t. meaning is what happens when you are absorbed in something outside yourself; it is incompatible with the stance of the person who is monitoring themselves for signs of it.

this is, dear reader, the genuine trap. you are probably optimising for proxies because the thing you actually want cannot be optimised for. so you find the nearest legible approximation and go after that instead, and the approximation may or may not deliver the underlying thing depending on factors you cannot entirely control, and whether it does or doesn’t, you were never quite doing what you thought you were doing.

i’m not sure this means you should abandon the proxies. there is something to be said for the person who decides that financial security is a reasonable proxy for freedom and then builds it, because at least they have a clear target and can make progress toward it and feel that progress, even if the connection between the proxy and the real thing is imperfect.

a life organised around a slightly-wrong goal is probably better than a life organised around no goal at all, which tends to produce the hovering state that most people recognise and hate in themselves: the sense of perpetual potential.

what i think actually helps is developing a habit of checking, periodically, whether the proxy is still pointing in the right direction. whether the thing you are working toward is still the most efficient route to the underlying thing you actually want, or whether it has decoupled and you are now working toward it out of momentum rather than direction.

this sounds obvious and is apparently very hard in practice, because the checking requires stepping back from the game while you are in the middle of playing it.

the version of this question i find most useful to ask people completely abandons the usual “what is your end goal” (because that tends to produce either paralysis or a recitation of inherited ambitions) and instead leans into something more like: in the moments when you have felt that your life was going well, what was present? and in the moments when it has felt most off, what was missing?

those answers are usually much more specific and much more honest than anything produced by the more direct question. they are also usually more actionable, because they point at concrete things rather than abstractions.

the win condition you actually have is almost certainly not the one you’ve been playing toward. that is not necessarily a disaster, because the game is long and you can adjust, but only if you are willing to ask the question and sit with an answer that might not sound as good on paper as the one you’ve been giving.

most people aren’t. but the fact that you’re reading this suggests you might be.

Share

michael young, a british sociologist, coined it in 1958 in a satirical novel called the rise of the meritocracy in which he envisioned a dystopia, not a blueprint, where a society sorted entirely by intelligence and effort produces a cold, contemptuous elite and a humiliated underclass with no excuse for its position. young spent the rest of his life watching the word he invented get adopted as an aspiration by the very people his book was warning about. he wrote an essay in 2001, three years before he died, begging people to stop using it positively.

they did not stop.

what is a meritocracy?

plato thought the best society would be governed by those most capable of governing, people he termed ‘philosopher kings’, selected through rigorous education and demonstrated wisdom. this is meritocratic in the narrow sense: allocate positions to those best suited for them. but it’s not what most people mean today, and plato wouldn’t have called it meritocracy because the word didn’t exist.

aristotle got closer to the modern idea with his account of distributive justice: give to each according to their contribution and their worth. this is recognisably meritocratic in that it proposes that reward should track something the person actually did or is.

john rawls, whose work underpins most contemporary liberal political philosophy, was explicitly suspicious of meritocracy. in a theory of justice, he argued that what looks like merit is partly just luck in the form of the natural lottery of being born with particular talents, temperament, and work ethic, none of which you chose. a just society therefore, for rawls, cannot simply reward what people happen to be good at, because what they’re good at is partly determined by factors entirely outside their control.

for me, though, i’m going to use a simpler definition, because i think the philosophical debate can obscure something that I find is actually quite intuitive: a meritocracy is a system where when you do better, you get more. Returns, whether those look like income or opportunity or position, track performance rather than birth, connection, or chance. you contribute more, you gain more. that’s the version of the idea i want to both defend and interrogate.

Subscribe now

i think the case for meritocracy, defined this way, is strong, and i want to make it before i complicate it.

the first reason is efficiency. a system that allocates roles and resources to those most capable of using them well produces better outcomes for everyone in it. you want your surgeon selected on surgical skill, for example; and your bridges designed by competent engineers; and the people making important decisions about your future, and that of your children, to be good at making decisions, not good at being born to the right family.

the second is fairness, or at least the most practically achievable version of it. a meritocracy doesn’t ask where you came from and instead it asks what you can do. for anyone who was not born into advantage, this is the best available deal that any system has offered: your position in it can, in principle, reflect your contribution rather than your circumstances (I understand this is a theoretical point and that there will always be frictions preventing this from being the case, but I will arrive at such objections subsequently).

the third is the incentive structure. if people believe that effort and capability are rewarded, they produce more effort and develop more capability. the belief itself is productive, and this translates in wider productivity as the motivational architecture of a society is partly determined by whether people think trying matters.

the fourth is social mobility. the historical alternative to meritocracy is not equality; it is aristocracy in the hereditary transmission of position and resource. meritocracy, at its best, breaks the grip of inherited privilege by creating a different mechanism for advancement. it obviously doesn’t eliminate advantage in its entirety, but it creates a competing channel.

the fifth is legitimacy. inequality is substantially more tolerable socially, politically, and psychologically when people believe it reflects genuine differences in contribution. systems perceived as meritocratic produce more social cohesion, less resentment between population strata, and more acceptance of outcomes. people accept that the person earning more has earned more, provided they believe the system of determination is fair (which is, in fairness, a rather large caveat).

the most serious intellectual challenge to meritocracy comes from michael sandel, whose book the tyranny of merit argues that meritocracy produces a specific and corrosive set of social attitudes. These are hubris in the winners (who believe they fully deserve their success) and what he calls the “demoralisation” of the losers (who, in a meritocratic system, arguably have no one to blame but themselves).

when you tell people the system rewards merit and they don’t succeed, the implication is that they didn’t merit success. this is psychologically brutal where older systems of inherited status were not: at least if you were a medieval peasant, your position wasn’t your fault. meritocracy makes failure personal.

but i think sandel’s argument is against a particular practice of meritocracy, not against the principle. a meritocracy practiced with humility (by which I mean one that acknowledges the role of luck in producing the capabilities that get rewarded and maintains dignity and decent floor conditions for everyone regardless of their relative merit) is not what sandel is describing. his target is a specific all-american version of the idea that has been weaponised into a justification for contempt.

rawls’s objection the meritocracy is, I find, deeper and harder to dismiss: if talent is partly luck, rewarding talent is partly rewarding luck, and there’s nothing particularly just about that. i find this philosophically compelling and practically insufficient. even if you accept that capability is partly unchosen, you still need some principle for allocating positions and resources, and reward-demonstrated-capability remains more defensible than any alternative i can identify. the solution to the luck problem is not to stop rewarding merit. i would instead argue that a solution looks like redistributing enough that the floor for those who don’t succeed is decent. even rawls himself didn’t fully argue against meritocratic allocation at the top.

the evidence that existing societies are genuinely meritocratic (in the sense that outcomes track performance rather than origin) is, at best, mixed.

raj chetty’s research at harvard, arguably the most comprehensive study of economic mobility ever conducted, found that the income rank of your parents is the single strongest predictor of your own income rank in the united states. 43% of people raised in the bottom income quintile remain there as adults. only 4% reach the top quintile. the economist miles corak demonstrated what he called the great gatsby curve: countries with higher inequality have lower social mobility. the us and uk, which most loudly advertise themselves as meritocratic, rank poorly on mobility relative to the nordic countries, which have more redistributive systems and, apparently, more actual meritocracy.

the old boys’ network has not disappeared. who you know remains a powerful predictor of outcomes in finance, law, media, and most elite professions. elite universities, which are often touted as the nominal laundering machines of meritocracy, are selecting increasingly for people who were already advantaged. This can be proxied through their selection of students whose parents could afford the tutors, the extracurriculars, and the schools that produce the grades that produce the place.

all this is to say the meritocratic mechanism is working as designed; it is just that the inputs to it are not equal, and the outputs therefore aren’t either.

the postcode lottery is not meritocratic either. being born in a deprived area correlates strongly with educational, career, and health outcomes. none of this reflects individual performance. all of this reflects circumstances of birth. a true meritocracy would need to equalise the conditions under which merit is developed before it could fairly reward the merit itself.

interestingly - and this is the part you should pay attention to - people consistently believe in the meritocratic nature of their system more than the evidence supports. surveys across western countries show that majorities believe success in their country is determined primarily by hard work and intelligence rather than by family background or luck. this belief persists in the face of data that would significantly complicate it, if people engaged with the data.

so meritocracy is both false as a description of how outcomes are currently distributed, and necessary as a belief that we hold about the system.

for the successful, the myth serves a useful psychological function. if your position is the result of merit, you deserve it. if you deserve it, you don’t need to feel guilty about it. the just-world hypothesis ( which is the well-documented human tendency to believe that people get what they deserve) is operating here. people need the world to be fair not because it is, but because the alternative is too threatening to live with.

for the unsuccessful, the myth is motivating in the short term and devastating in the long one. if the system rewards effort, effort is worth making. this is productive. but if the system rewards effort and your effort has not been rewarded, the failure is yours. the meritocracy myth turns structural inequality into personal inadequacy, and the people it does this to are those who were already least advantaged by the structure.

and yet. the alternative to the myth is the belief that outcomes are arbitrary and that effort is irrelevant. this produces something worse. it produces learned helplessness. the people most harmed by the gap between the meritocracy myth and the meritocratic reality are not helped by knowing the full extent of that gap. they are, in many cases, made less able to navigate the actual system by the knowledge that the actual system is unfair.

this puts us in a difficult position. the myth is motivating and false. the truth is accurate and partially demoralising.

i don’t think the answer is to stop believing in meritocracy. i think the answer is to use it as a standard the system is held to, rather than a description of the system as it is. we should be saying: this is what we are aiming for, this is how far we have fallen short, and here is the gap that needs to close.

young was right that the word has been misused. but the ideal behind it, that what you contribute should shape what you receive, and that your origins shouldn’t determine your ceiling, is worth keeping, even if the implementation has mostly let it down.

the myth is false. the aspiration isn’t.

okay, i’m being dramatic. but not by much. entry-level job postings dropped 38% between 2022 and 2023. new graduates made up just 7% of new hires in 2024, down 25% from the year before. 52% of the class of 2023 were underemployed a full year after graduating.

one headline i came across recently called it “the worst entry-level job market in 37 years,” which is a sentence that contains a lot of information.

at the same time, the cost of a basic life has moved considerably out of reach. in the uk, you now need a household income of £60,600 to buy an average home, against a median uk household income of £33,300. in the us, the income needed to afford a typical home has increased nearly 50% since 2020, to just under $117,000, while median household income sits around $81,000. the median age of a first-time homebuyer in the us hit 38 in 2024.

so: fewer jobs available, more applicants per job, and the traditional destination of the traditional route (that being job security, a family home, and the ability to pay for your, and your children’s, expenses) has become both more expensive and more distant.

this is the context i want you to hold in your mind, because I am going to argue that it explains something that’s being consistently misread as a cultural shift.

everyone is starting a company.

new business applications in the us hit 5.49 million in 2023 - the highest year on record. the number of startups increased 16% between 2019 and 2023, more than twice the rate of the preceding period.

the word “founder” has become, in the circles i move through at least, the prestigious answer to the question of what you’re doing with yourself. “i’m looking for a job” sounds like a vulnerability; “i’m at a company” is not treated as ambitious enough.

“i’m building something” is the answer that gives you status.

i want to make an argument here that is slightly uncomfortable (not least for myself, given the stakes I have in this): i think the explosion of interest in startups among young people is, at least partly, a recession indicator.

the core of my argument boils down to the fact that a meaningful proportion of the people now calling themselves founders are not founders because they have a burning problem they need to solve. they are founders because the alternative - entering a labour market where entry-level roles are disappearing and there is a need to accept salaries that in many cities make the basic architecture of adult life inaccessible - has stopped looking like the stable path it once was.

when the traditional route loses its promise, people look for a different route. and startups, right now, are the most available and socially acceptable alternative.

this is not, in my opinion, a small distinction that we can just gloss over as a shift in larbour market dynamics. it matters because it influencers who is building and why they are building.

i say i have stakes in the game here because i am actively building a startup. (the specifics are here, if you’re curious.) so i am, in the most literal sense, a case in point for the argument i’m making. i am doing the thing i’m interrogating.

I don’t wish to pretend that this doesn’t affect how i’m thinking about my argument, because the most honest version of this essay requires me to hold two things at once: i believe in what i’m building, and i also think that some of my reasons for building it are structural rather than purely visionary.

the line between “i have identified a genuine problem and i am solving it” and “the conventional route looked less appealing so i found a different frame for my ambition” is blurrier, for most founders, than the mythology around the ‘silicon valley tech genius’ perhaps allows.

compounding these adverse labour market outcomes has been the development of agentic ai. over the past, shall we say, two(ish) years, the barrier to entry has been lowered so substantially that building a product no longer requires a technical co-founder, significant capital, or a long runway before you have something to show. you can go from idea to prototype to something resembling a product faster than at any previous point in the history of technology. this is extraordinary and it is good. innovation is good. more people being able to attempt things is good.

however, as is the case with any scenario, especially those involving innovation and tech, there are some second-order consequences that are being substantially underweighted.

the first is the overwhelm problem. when the barrier to building drops, the number of products rises, and the signal-to-noise ratio drops with it. what i mean by this is best described through example: markets that used to have three viable solutions now have thirty mediocre ones. nobody can find the good thing, the good thing can’t find enough customers to survive, and an enormous amount of human energy is expended producing things that don’t make a meaningful difference to anyone. essentially, the matching systems between consumer and problem-solving product have broken down. the phrase “the best minds of my generation are building recommendation algorithms” is from 2011 (Jeff Hammerbacher). the best minds of my generation are prompting claude.

the second is the talent allocation problem. i’m aware that this sounds almost comically effective-altruist of me, and i say it knowing the argument can be levelled directly at me, but: many of the people building b2b saas tools [for the less chronically online/in the startup ecosystem this is businesses whose business model is selling software to other businesses] for workflow optimisation are people who could be doing things with considerably larger impact. this is not a critique of any specific choice: the economics of startups push almost everyone toward b2b, where there’s a clear buyer with a budget, and away from consumer products or public goods, where the monetisation is harder. the result is a startup ecosystem that is, in aggregate, oriented toward selling software to other companies rather than toward the problems that actually need solving. which brings me to the issue that the c doesn’t exist in b2b saas; the consumer is missing out on much of the innovation currently being produced.

the third is the failure problem. 65% of startups fail within ten years. 48% within five. the most common reason, cited in 42% of post-mortems, is that there was no market need for what they built. the founders were solving a problem that didn’t exist at sufficient scale, or that their potential customers didn’t experience as painful enough to pay for. this is what happens when you start with the solution and work backwards to the problem, which is much easier to do when the impulse to start a company comes primarily from the labour market rather than from a genuine, persistent, specific frustration with how something works.

most of the people starting companies right now will not make money from them. that is a real cost, and the culture of startups as aspiration does not discuss it honestly (of course, the argument that you learn a lot, even from failing, does hold water, but that’s not quite the point I’m making if you’ll indulge me in the nuance).

Subscribe now

there is a related conversation happening about whether to go to university at all. the thiel fellowship - which pays selected students $250,000 to drop out and build companies - has become a kind of totem for this argument. there are people for whom this is the right call. there have always been people for whom this is the right call, and they were outliers before the fellowship existed and they remain outliers now. the fellowship, and others like it, doesn’t make the path more generalisable.

here is my actual view: i think ai will increase rather than decrease the premium on degrees from elite institutions, and specifically on postgraduate ones. this will occur not because the knowledge inside them is irreplaceable (as much of it increasingly is replaceable) but because in a world where the baseline quality of machine-produced work rises, the signal value of a credential that says a human being with verified ability sat with this problem rises with it.

the elite degree becomes a trust signal in an environment of pervasive uncertainty about what was made by whom. i could be wrong about this. but it’s where i’ve landed.

the thing i keep coming back to is that the startup boom of the mid-2020s will, i suspect, look in retrospect like a combination of two things. some of this drive is genuine innovation comprising real problems being solved by people who had genuine insight and the tools to act on it faster than any previous generation could have. much of it, however, is what happens when a generation is priced out of conventional stability and offered, by the culture, an alternative story about what that means.

the solutions to the two problems are different. if the issue is that young people are locked out of a labour market that has stopped creating entry-level roles, building a startup is one response, but it is not a policy, and it does not solve the underlying issues. if a significant proportion of current startup activity is a response to labour market failure rather than genuine product insight, that has implications for how we think about what this wave of entrepreneurship actually represents and what happens to all those founders when, or if, the market opens back up.

i’m not sure i have a clean answer to any of this. i’m building something, i think it’s a real problem, and i’m also aware that i exist in the exact conditions i’ve just described.

*the data in this essay comes from myperfectresume’s analysis of entry-level job postings, the federal reserve bank of new york’s college labour market tracker, ons housing affordability data for the uk, bankrate’s 2025 housing affordability study, us census bureau business formation statistics, and failory’s startup failure rate research.*

the question is this: do i want the things i’m working toward, or do i want people to see that i have them?

and before you say “both” - i know.

i’m saying both too.

but i’ve started to notice that both is doing a lot of heavy lifting as an answer, and that the two parts of it are not equally weighted.

here’s what I mean: i notice a difference between how it feels to do the work and how it feels to tell someone about the work. and, if i’m being fully honest (which is the whole point of this) the second one often feels better. not always. but enough.

i think this is a more common feeling than people might let on. i also think it’s one of the most important things an ambitious person can look at and understand about themselves, yet almost nobody does, because the honest version of the answer to this question isn’t particularly flattering.

ambition, as a concept, doesn’t really invite this kind of scrutiny. it’s presented as almost entirely virtuous (“you’re so ambitious” is a compliment) and so the question of what the ambition is actually for tends not to get asked. we interrogate what people want to achieve and yet almost never interrogate why they want to achieve it.

i think that why contains almost everything.

there are two versions of ambition that look identical from the outside. one is wanting the thing itself. the other is wanting what the thing signals: specifically, the proof it provides that you are the kind of person you want to be seen as being.

both make people work very hard.

both produce people who talk fluently about their goals.

both can get you to the achievement, eventually.

what they don’t both produce is peace when you arrive.

if i look at my own ambition honestly, i can trace parts of it back to the satisfaction of being told i was “doing well”. i want my parents and teachers and friends to look at me with approval, and much of my formative educational years were spent in an environment in which being good at things was the primary social currency. at some point, fairly early, you therefore learn that achieving things produces something beyond the achievement itself, whether that be recognition, social credit, or just personal satisfaction. these outcomes are valuable enough that you start, at least partly, to work towards them.

this is not some personal failing of mine or yours. it’s pretty much universal (as far as my limited sample size stretches, anyway). the problem isn’t that the approval-seeking was there at the start. the problem is when it becomes the dominant engine with which you propel yourself, and your choices (in terms of what to pursue, what to value, what to call success) start being made primarily on the basis of how they’ll look rather than what they’ll actually produce in your life.

aristotle wrote about the difference between pursuing honour and pursuing virtue; by this he was referring to the difference between doing the excellent thing in order to be seen doing it and doing it because you actually care about it. his issue with honour-seeking wasn’t that recognition is worthless, but that honour is given by other people, which means it’s never fully within your control.

the person whose sense of self is built on being seen to achieve things has placed themselves in a permanently unstable position, because what other people think is not something you can ever fully secure.

the impressive version of ambition has some tells, i’ve found, once you know to look for them.

the first is that the achievement always feels dull. you get the thing you have been working towards and there’s a moment of relief (note this is not a moment of pride nor of satisfaction) and then the next thing assembles itself and the current one already feels insignificant.

this happens partly because all highs fade, which is just how hedonic adaptation works. but it also happens because, if what you were really chasing was the feeling of being seen to achieve something, that feeling is by definition brief: the audience moves on and so you need the next achievement to generate the next response.

the second is that you start choosing between options based on how they sound rather than what they are. i don’t think this happens consciously, or at least it didn’t for me. it happens in the evaluation of “would this be good” and “would this look good”, and when those two things diverge, the impressive version of ambition knows which one it picks.

the third is that if you’ve been running on impressive-ambition for long enough, the question of what you actually want can start to produce nothing. you’ve been deciding what to want based on what’s impressive to want for long enough that the original preference is nigh on impossible to locate.

i say this having experienced it and i say this not knowing exactly how to fix it.

the question i’ve found most useful (and i offer it as a question rather than an answer, because i don’t have a clean answer) is this: would i still want this if nobody could know i was doing it?

this is not the same as the question, “would i work in secret.”. it is more: if the thing produced no external signal (and by this I mean there would be no credential at the end and no visible proof that you’re the kind of person who does this kind of thing) would you still want the daily reality of it?

i find this question difficult.

harry frankfurt, the philosopher, made a distinction i like: there is a difference between merely wanting something and actually caring about it. caring, in his sense, is not just a more intense form of wanting, but it’s something that orients you and persists through difficulty because the difficulty is incidental to the thing you care about, rather than the price you’re paying for the reward at the end.

the impressive version of ambition produces wanting. only the genuine version produces caring. someone who cares about the work produces something different, in the long run, to someone who wants to have done the work, both in the quality of what they make and in whether they can look at their life and feel that it was actually theirs.

i’m not writing this from a position of having figured it out. i’m writing this because i’ve been sitting with the discomfort of not being sure that my ambitions are as genuine as i’d like them to be.

the truth, as far as i can access it, is that my ambitions are mixed.

some of what i’m working toward i want genuinely. some of it i want partly because of what it will say about me when i have it.

i think the honest version of this for most ambitious people, if they look carefully enough, is also mixed.

the thing i’m trying to do (and i don’t always manage it, if I ever do) is to know the difference. i want to be able to look at a goal and have some sense of which engine is driving it, and to make choices accordingly.

do you want it, or do you want other people to see that you have it?

i’m still working out my answer.

i want to be careful about how i say this, because my professors might be reading... i love the academia. i think my supervisors and lecturers are fantastic - they push me to understand my subject with a depth i expected and value, partly through the intellectual rigour of the supervision model (which, for those unfamiliar, is the cambridge tradition of sitting in a room with one or two other students and being asked to defend your thinking in front of your professor - at the other place [Oxford] they call them tutorials).

that part works. that part is what cambridge says it is. (and i promise i’m not being held up in a library and forced to write this!)

what i didn’t anticipate is everything outside the academics.

cambridge as an institution sells itself on intellectual rigour. the rigour is real - i’m not disputing that - but there’s something the prospectus doesn’t tell you, and it is something i’ve had to work out for myself over the last year and a half: intellectual rigour and intellectual curiosity are not the same thing.

and cambridge, for all its reputation, seems to attract far more of the former than the latter.

what i mean is this is that the people here are, by almost any measure, extraordinarily academically capable. they work hard, they understand their subjects, they perform well in the systems the institution has built to measure understanding. and yet (and i say this knowing it might be wrong, knowing it is based on my own limited sample of my own limited experience) i often find a narrowness that surprises me.

many people have a deep focus on the single subject, the thing that got them here, and a relative incuriosity about everything else. there is an abundance of people who know a very great deal about one thing and have not yet found themselves wondering, with any urgency, about the rest.

i thought more people here would be curious. i thought they would want to do things.

i realised quite quickly that a lot of people talk, and relatively few people do. i’ve sat in enough common rooms and enough formal dinners to notice the pattern of conversations that perform intellectual seriousness far more than they enact it.

the difference between someone working through an idea in front of you and someone demonstrating that they are the kind of person who works through ideas is, once you’ve seen it a few times, fairly clear.

cambridge, perhaps because of its reputation or perhaps because of what its reputation attracts, has more of the latter than i expected.

i think this is a byproduct of what academic culture, at its most elite, actually rewards. the posturing is not dishonest; it cannot be because most people doing it are not consciously performing. but the environment selects for it.

if the institutional signals of success are about demonstrating mastery, demonstrating intellectual seriousness, and demonstrating the right kind of engagement with the right kind of ideas, then you get a culture that is very good at the demonstration and rather less interested in what the demonstration is supposed to be for.

this matters enormously if, like me, you are trying to actually build something.

i am working on a startup. i will not get into the specifics here (though they are here ;) ), but the experience of developing it at cambridge has taught me something i didn’t expect to learn: that being surrounded by very academically able people is not the same as being surrounded by people who share an enthusiasm for getting things done.

i’ve found the obvious in that meaningfully building requires a certain tolerance for uncertainty, coupled with a willingness to make decisions with incomplete information and a deep-rooted interest in doing before you fully understand. this disposition, it turns out, is not what cambridge primarily selects for or rewards.

the people who have been most useful to me have not, in the main, been the people with the best exam results. instead, they’ve been the people who wanted to make something exist that didn’t exist before, and who were willing to be bad at it for long enough to get good at it. i’ve found some of those people here, but fewer than i expected, and finding them has taken more effort than i’d have thought in an institution supposedly full of exceptional minds.

forgive what follows, because i’m aware of how it sounds, but i think the university should operate more like an incubator. more like an academic y combinator: a place that identifies exceptional people who want to do things, gives them the resources and the community and the permission to try, and gets out of the way long enough for something to happen.

the intellectual formation can coexist with this. the two are not incompatible but right now, they sit in separate parts of the university’s identity and the formation gets the prospectus while the incubation is left with the occasional entrepreneurship society.

i’ve been trying to work out whether i’m making an argument about admissions, and i think i partly am.

cambridge, like most elite universities, selects primarily on academic performance. this makes sense as a proxy: it is largely an accurate depiction of one’s intellectual abilities in terms of future exam results (and, as a university, cambridge wishes to optimise for the number of first class graduates it can produce), and, importantly, it is measurable.

however, my issues is that it is a proxy for a specific set of capabilities summarised by the ability to acquire and demonstrate knowledge within a defined structure. the people who are best at that are not necessarily the people who are best at what comes after.

the people who are best at performing within the institution are not always the people who will do the most interesting things once they leave it.

the alternative is hard to operationalise, i absolutely understand that. “future returns,” in the broadest possible sense (what someone will build, create, contribute, change) is intangible, context-dependent, and almost impossible to assess at eighteen on the basis of a personal statement and a set of interviews. i am not claiming i know how to fix the admissions process. i’m not sure anyone does.

but i think there is a question is worth asking more directly than it currently is, and that question is: what is cambridge actually trying to produce? and is the process by which it selects people well-calibrated to produce that thing?

the deepest version of my frustration, i think, is this: there is a difference between knowledge acquisition and understanding, and elite academic culture has become extraordinarily good at the former while remaining much less reliable at producing the latter.

knowledge acquisition is knowing the content of a field. understanding is knowing what to do with all of that. the former can be tested in an exam. the latter mostly cannot.

the supervision system at cambridge pushes, at its best, toward understanding because the good supervisor does not want to know if you’ve read the paper, they want to know what you make of it. but outside this formal academic structure, the culture rewards knowledge acquisition.

from my perspective, this must change. not because knowledge acquisition is worthless (it is obviously not) but because what the world will actually ask of cambridge graduates (and graduates from other elite universities around the world) is understanding: the capacity to encounter a novel situation and think through it, rather than reach for a framework that has already been downloaded into their brain.

i am aware that i might be wrong about some, or all, of this. my experience is narrow, my sample is limited, and there is a version of this argument that is just a person in a specific environment failing to find their people and making it a systemic critique.

that is a valid opinion to have.

i’ve met people here who have changed how i think, and who have been genuinely curious. they exist.

but they’ve been harder to find than i expected, and i’ve spent more of the last eighteen months than i’d like feeling like the thing i care about most (which is, building something real in the world that didn’t exist before) is a minor subgenre of the cambridge experience rather than a central one.

what i wanted when i applied was a community of people who were curious about everything and wanted to do something with it. what i found was a community of people who are very good at their subject, some of whom are curious, some of whom want to do something, and almost none of whom make the connection between those two things automatically, the way i’d assumed they would.

perhaps that was always too much to expect. perhaps it’s exactly the right thing to expect, and the institution hasn’t figured out how to deliver it yet.

i’m still deciding which one it is.

and then you go back to your life, and the room is different permanently. the room is structurally different.

you find yourself, months or years later, thinking in a way that you can trace back to that encounter, using a lens that isn’t originally yours, but that you’ve borrowed so thoroughly it’s started to feel like it belongs to you.

most people have had this with one or two people in their lives. it doesn’t generally happen with the people who taught you the most facts. sometimes it happens with a teacher, sometimes a writer you’ll never meet, sometimes a person encountered briefly at exactly the wrong or right moment. the distinction i want to make is between the people who informed you and the people who transformed you.

transformation

informative encounters add to your existing structure: they give you new content, new context, or new data/knowledge to place within a framework that you already hold. you essentially come away from such an encounter knowing more about the same things.

transformative encounters change the structure itself, meaning you don’t just know more, but you’re actively operating with a different framework. the philosopher thomas kuhn described paradigm shifts in science as moments when the anomalies that couldn’t be explained by the existing model accumulate to the point where the model itself has to change. the transformative intellectual encounter does something like this at the individual level.

this distinction explains why you can read a thousand books and attend a hundred lectures and still find that there are only a handful of encounters, or sometimes just one, that you come back to as genuinely having changed how you operate. the rest, however valuable, worked within a framework that had already been set. it is only these rare transformations that change the framework itself.

vygotsky described the zone of proximal development as the space between what you can do alone and what you can do with support. this is the edge where personal and professional development happens with the greatest efficiency, and is where the next level of your own capability becomes accessible through what we might perceive to be ‘the right encounter at the right moment’. the transformative intellectual encounter almost always happens in this zone.

this means that encouraging and inviting such encounters is about you, and what stage you’re at, and whether your existing framework is under enough pressure to be permeable, than it is about the other person.

this is why the same person can be utterly transformative to one person and merely interesting to another. it is also why the mentor or teacher or writer you’d have dismissed at fifteen can be exactly what you needed at twenty-one. transformation requires a particular kind of readiness that looks like a mental framework already strained at the seams.

what makes a transformer

the people who transform how you think are almost never trying to. the transformation is a byproduct, and not a product in and of itself.

what these people tend to share is something about how they inhabit their own thinking. something particular that I have noticed is that they possess a wonderful ability to think out loud, incompletely, in front of you. watching a mind you look up to move through something uncertain, without pretending to certainty it doesn’t have, teaches you that not-knowing is a state an intelligent person can occupy without it being a failure.

it models a relationship to ideas that is different from the one most formal education produces, which tend to reward the performance of certainty over the genuine article of inquiry.

the other thing they tend to do, and this is the one i find most interesting, is give you language for something you already felt. the most transformative intellectual encounters often feel more like recognition than discovery. once you have the ‘word’, you can think about the thing, and once you can think about the thing, you can build on it, and something that was formless becomes available as a foundation.

if you are reading this, there is a version of you that exists in other people’s heads. you have had conversations that you don’t remember, that other people do. you have said something, offhandedly, that may have landed differently than you intended, and that a specific person, at a specific moment, received as exactly what they needed.

you don’t know who.

you don’t know when.

the encounters that have mattered most to people you know are almost certainly not the ones you’d have predicted.

the research on mentorship makes a distinction between what kathleen kram called career mentoring (which is mentoring of the instrumental, practical kind), and psychosocial mentoring (which operates more at the level of identity and involves how a person comes to see themselves in relation to their domain). the psychosocial is harder to observe, harder to measure, and considerably more transformative. importantly, it operates almost entirely outside the awareness of the person providing it.

you cannot engineer becoming someone’s transformative person.

you cannot decide to change someone’s cognitive register the way you might decide to teach them something.

harold bloom wrote about the anxiety of influence in literature: the idea that the deepest influences are the ones you have to struggle against and eventually depart from in order to find your own voice. the most transformative intellectual encounters are not necessarily the most comfortable ones.

for whoever is reading this

underneath the argument about mentorship and cognitive frameworks and psychosocial development, this essay is actually you something very simple.

it’s asking whether you’re letting people change you.

i, quite frankly, don’t care whether you’re open to being changed in the abstract, because most people will say they are without question, but i do care (and so, dear reader, should you) whether you are genuinely permeable to specific encounters with specific people. in plain english, i care about whether you are reading and listening and talking in a way that leaves the possibility open that you might think differently afterward.

being informable is easy. we receive information constantly.

being transformable is harder. it requires bringing your existing framework into contact with something that strains it and tolerating the discomfort of not quite knowing what you think while the change is happening.

the people who change how you think don’t offer you certainty. they offer you a better question. what you do with it - the months or years of working through what it implies - that part is yours.

- erin x

take a moment and actually try to answer it before reading on. notice that i am not asking you about the last time you had a day off, or watched a show you enjoy, or went for a walk outside, or lay down for a bit.

i am asking you when was the last time you genuinely had a break.

i’m guessing you’re finding it harder to answer than you expected. i find it harder to answer than i expect, every time i ask myself. i’ve started to realise that difficulty is a very important thing to pay attention to.

in order to answer the question, we need to define what a break actually is.

i suspect your definition, if you have one, is wrong - or at least incomplete. mine was.

the dictionary defines a break as an interruption in continuity, a pause in work or activity. that’s not what i’m talking about. the neuroscientists have their own version: a state of cognitive restoration and the activation of the default mode network, which is the system that activates when the brain is not directed at an external task, associated with mind-wandering and memory consolidation. that definition is closer, and it matters, but it’s still not quite there.

what i’m talking about is something so subjective that it is hard to explicitly name: a feeling of turned-offness in which there is no tab open in the back of your mind. it is a state in which you are genuinely, fully, unreachably not here, in the sense of not being available. not off duty but actually off.

you might recognise this as something you’ve had distantly. a particular afternoon from childhood, maybe. a moment on a holiday before taking out your phone. it is a qualitatively different state from rest and leisure, and i think we have almost entirely stopped accessing it, without quite noticing that it’s gone.

the problem is structural, and it starts with what we’ve built.

everything is available all at once, all the time, in the device that is almost certainly within arm’s reach of you right now (if not the very one you are reading this on). your work email is there. the news is there. the people you haven’t replied to are there. the content - the endless, infinite, inexhaustible content - is there.

at any moment, at three in the morning, on christmas day, in what used to be the unreachable privacy of your own head, everything is accessible. you can choose not to look, but you cannot be genuinely unreachable.

there is a study by ward and colleagues at the university of texas that found, remarkably, that just the presence of a smartphone on a desk (even if face down and untouched with notifications off) reduced the available cognitive capacity of people trying to complete tasks. the phone didn’t have to be used to have an effect. its proximity was enough to divert a measurable portion of attention toward the suppression of the impulse to check it.

we are not, even when we’re not looking, genuinely off.

this is new, historically speaking. not the feeling of having a lot to do, of course - that is ancient, but the structural impossibility of the off state. before the internet, the end of the working day was a genuine state change enforced by architecture: the office closed, the phone was attached to the wall, and the only way to reach you was to physically go to where you were.

the break was a condition of the world and you didn’t have to discipline yourself into unavailability because unavailability was the default.

that default has been reversed. availability is now the default, and unavailability is something you have to actively choose and defend, with considerable effort, against a world that is designed to continuously reclaim your attention. the break is no longer something that naturally arrives, but something you have to build against friction.

most people are not building it. most people are not even noticing it’s missing. Perhaps that is somewhat unfair: most people are noticing the symptoms without naming the cause. they feel the tiredness that sleep doesn’t resolve.

if everything is available all at once, everything becomes equivalent and equivalence erodes meaning.

meaning - the sense that something matters in the way that a particular moment is different from the moments around it - depends in large part on contrast and scarcity. the meal that matters is the one that’s different from the meals around it. the conversation that stays with you is the one that opened something different. the holiday that restores you is the one that was genuinely ‘other’ than the rest of the year in structural terms, not just geographic ones.

these things derive their weight partly from what surrounds them: from the ordinary that makes the extraordinary legible as extraordinary.

josef pieper, the german philosopher, argued in 1948 that genuine leisure (interestingly, this is what the greeks called schole, from which we get the word school, because they thought this receptive state was the precondition for learning) was not the absence of activity but the presence of a quality of inner stillness.

he argued that leisure is a mode of being in which you are not instrumentalising your time. You are not producing, nor consuming, nor managing, but are just present in yourself and your surroundings, without agenda. he thought this state was the foundation of culture itself: that art, philosophy, and genuine thought could only emerge from it.

i think pieper was onto something that we have, in the last fifteen years or so, made structurally scarce.

when everything is available all the time, time loses its texture. there is no before and after. this results in there being no genuine anticipation, because the thing you’re anticipating is already accessible. there is no genuine memory, in the sense of a period that is now over and cannot be returned to, because you can open the photos and videos and messages to reload the moment on demand.

the edges of experience (the beginnings and endings that give it shape )dissolve into an ambient stream of now that is always more of the same.

in that stream, things start to feel meaningless. meaningless in the same way that a piece of music feels meaningless when it’s played continuously at low volume until it becomes part of the background: it’s still playing and you can still hear it, but something essential has drained out of it.

i am not arguing that the solution is a digital detox or a phone-free retreat or a carefully scheduled weekly period of intentional rest, although i think intentional rest is better than no rest.

the issue is deeper than any of those interventions can reach.

the question is whether we, on brand society terms, are willing to take seriously that the turned-off state is not a luxury or a productivity strategy, but something closer to a precondition for a life that feels like it actually means something.

the world we have built, which has systematically removed the architecture that used to produce it automatically, requires us to do something effortful and countercultural to get it back.

blaise pascal wrote in the seventeenth century that all of humanity’s problems stem from man’s inability to sit quietly in a room alone. he was, as far as i can tell, being serious. and he wrote it before the room contained a device that connected you to everyone and everything you’ve ever known, all at once, at no charge, with infinite content and a notification system designed by some of the smartest engineers on the planet to make sitting quietly feel unbearable.

i don’t know when i last had a real break. i’ve been trying to remember since i started writing this.

i’m not sure that’s a trivial thing.

- erin

it’s 11pm.

you should be working, or sleeping, or both.

instead you’re watching a video about how to be more productive.

at some point while you are watching, the anxiety you feel about whether or not you are ‘doing enough with your life’ starts to ease. you feel like you are doing something. you feel like tomorrow will be different.

you feel, if you’re honest, pretty good.

you haven’t done anything. but the feeling is real. and that feeling is worth $32.5 billion a year.

that number is the most interesting thing about the productivity industry: we are spending more than the gdp of 45% of countries, every year, on content and tools and systems designed to make us more productive, and the market is growing.

if the product worked, you’d expect the market to eventually shrink because solved problems don’t tend to keep generating revenue.

the size of the thing

notion, a note-taking app that has since moved more into personal ai development, has 100 million registered users, a $10 billion valuation, and $400 million in annual revenue.

ali abdaal, a former doctor who now makes videos about productivity, has 6.5 million youtube subscribers and over 500 million total views.

the us self-improvement market is worth $13.4 billion.

the global market is projected to reach $90.9 billion by 2034.

these are not the numbers of an industry solving a problem. these are the numbers of an industry feeding one.

the claims and the reality

the productivity industry wants to sell improvement, which is mostly defined as concrete, measurable gains in factors such as your output, your focus, and your mental wellbeing. what it actually sells is something considerably more valuable to its own business model: the feeling of working on yourself.

this distinction matters enormously, because the feeling of working on yourself and actually working on yourself have almost entirely different relationships to continued consumption.

if you actually improved, you wouldn’t need the next video, the next system, the next book. you’d be done, or at least at a different stage of done.

but the feeling of improvement (which might look like the dopamine-hit-manifesting-as-increased-productivity that you get from a well-designed notion template or watching a video of me explaining the PQ4R technique or adding a new habit to your tracker) does not reduce your appetite for more of the same. it increases it.

the market for the next piece of content is created, in part, by the consumption of the last one.

a 2026 study followed a large cohort of adults over two years and found that people who used self-help products did not differ in personality or wellbeing trajectories from those who didn’t. usage intensity made no difference.

the productivity industry is, at its most basic level, selling a product whose primary measurable effect is a continued desire for the product.

the gap is the product

here is the economic logic: the productivity industrial complex requires a gap between where you are and where you feel you should be. it needs you to feel not-quite-there, meaning you believe you deserve better and that you are capable of improvement. These are positive feelings, no doubt, but the industrial complex needs to grapple with the notion that, if the gap closed, the market would close with it.

this is why the most effective productivity content doesn’t actually resolve the underlying problem. it names the problem, validates the problem, gestures toward what could be a solution, but ultimately leaves you sufficiently un-solved that you’ll come back for the next dose of advice.

the best productivity creators are not trying to fix you because fixed people don’t need more videos.

So, if i create productivity content, where do i fit in? This is where I have been very intentional in my discussions around productivity: offering a ‘tip’ or ‘hack’ as a panacea is not my style, and there is a reason why my content revolves so heavily around the neuroscience: i am not trying to hand productivity to you on a plate; i want to give you the tools to understand your brain so that you can work out what real productivity looks like for you.

Does that make me completely innocent? Perhaps not. But the effort i make is intentionally directed to avoid selling you the productivity industrial complex.

the productivity content is the procrastination

watching a video about how to study is what you do instead of studying. making the notion database for your revision notes is what you do instead of revising. the productivity content is not the preparation for the work. it is the procrastination with better cultural branding.

anxiety and procrastination have a well-documented cyclical relationship: anxiety produces avoidance, avoidance produces more anxiety, the anxiety makes the original task feel more threatening, and the avoidance deepens. what productivity content offers, in this context, is an activity that feels like addressing the problem while actually being another form of avoidance. it is procrastination that has convinced you it is self-improvement.

the reason it’s so effective at this is something called structural procrastination: the tendency to do lesser tasks in order to avoid a more important one, while still feeling like you’re being productive. reading about the best way to structure your notes is a lesser task relative to actually writing the essay. your brain accepts it as progress which causes the anxiety to briefly subside. but, the work remains incomplete.

the person who is genuinely in deep focus does not need to watch a video about deep focus. the person who is anxious and avoiding their work does, and will.

what is really being sold

there is a layer beneath the practical utility of productivity content: identity.

the aesthetic of productivity has become a subculture. think about the studygram or studytok pages you might have seen, or that you yourself follow: the satisfying notion dashboard, the journaling setup photographed on a curated desk, the 5am morning routine; these are not primarily functional objects. they are signals to an audience about the kind of person they are or, more likely, are trying to portray themselves as. the person who has a sophisticated productivity system is not necessarily more productive than the person who has a to-do list on a post-it note, but they visually have a more developed identity as someone who takes their own output seriously.

this identity is worth paying for. it is worth spending an hour on. and, it is specifically appealing to people who feel somehow out of control or overwhelmed by their work. the productivity identity is a response to the experience of not coping. the not-coping, rather than being addressed directly, is managed through the consumption of content that provides the feeling of response without the substance of it.

the market grows with the anxiety

the productivity industry did not emerge in a period of calm. it grew alongside the internet, and always-on work culture that this has allowed to emerge. it benefits greatly from the erosion of the boundary between work and not-work.

the market for feeling more in control of your time and output grows in direct proportion to how out of control of the broader world and economy people feel. this is the mechanism by which a person who is overwhelmed becomes a consumer, and by which a consumer remains one.

the productivity industrial complex is very good at answering the question: how should i organise my work? it is structurally incapable of answering the question underneath it: why do i feel, despite everything, like i am not enough?

that second question is not a productivity problem. it is not a systems problem. it is not a tools problem. it is not a problem that has a youtube channel. the industry cannot say this without dismantling its own market, so it doesn’t say it. instead it releases a new app, book, or framework, and you pay for it because the alternative of sitting with the original question is considerably less comfortable than watching yet another video that promises to change your life.

the argument i’ve just made (that is, that the content is the procrastination, and that the feeling of not-enoughness is the product) is itself a piece of content. you consumed it instead of doing something else. i wrote it instead of doing something else.

the satisfaction you might feel right now, having engaged with a serious idea about the failure of self-improvement culture, is not entirely different from the satisfaction you feel after watching a productivity video. it’s the feeling of having addressed something, without necessarily having changed anything.

i don’t mean this in the motivational poster sense, where the word “failed” secretly means “didn’t go as well as had hoped,” and where listing your failures is just a way of telegraphing that you survived them. i mean it in the actual sense: there are things i tried that really didn’t work, directions i pursued that were totally wrong for me, efforts i made that produced nothing commensurate with the effort itself, assessments i sat and performed worse than i wanted to on, to name just a few.

some of these i recovered from. some i changed course from. some i’m still not sure what to make of.

what i’ve learned, not from ‘surviving’ failure, but from watching how i processed it, is that almost everything we do instinctively in the immediate aftermath of failing is precisely wrong. not slightly suboptimal, but precisely wrong. the responses that feel most natural are the ones most likely to ensure that the failure teaches you nothing and costs you the maximum amount.

this is an essay about why that is, and what accurate failure-reading actually requires. it is not going to tell you that failure is secretly success. it is going to try to be useful.

what failure actually is

failure is information. i don’t say this as some sort of comforting reframe, just merely as a literal description of what it is. when something you attempted doesn’t produce the result you intended, the gap between intention and result contains a signal about the world. it tells we about the relationship between your current approach and the outcome you’re pursuing, and therefore whether the approach needs adjusting or the direction needs changing.

this is what makes failure valuable, when it is valuable. suffering is not a prerequisite to learn from failure, and the retrospective narrative in which “the hard times made you who you are” does not need to be worn as a badge of honour. the value of failure is in the signal, and the signal is only useful if you can read it accurately.

the problem, and this is the thing the motivational poster version never addresses, is that accurate failure-reading requires you to do something that is deeply counter to instinct: to stay in the discomfort of not knowing what the failure means, long enough to figure out what it actually means, rather than reaching for the most emotionally available interpretation and getting on with your life.

the two things failure can mean

not all failures are the same, and the most important distinction in processing one is this: does this failure mean i’m going in the wrong direction, or does it mean i’m not there yet?

wrong-direction failure is telling you something about fit or underlying assumptions. your approach is structurally misaligned with your goal, or the goal itself was based on a model of what you wanted that has turned out to be inaccurate. the relevant response here is to change course. the failure is data about direction, not about effort, and no amount of increased effort applied to a wrong direction improves the situation.

not-yet failure is telling you something about the gap between your current capability and the required standard. the direction is right, but you haven’t developed the skill or the knowledge or the judgment that the next level requires. the relevant response here is to continue, but with accurate information about what specifically needs to change, not simply with more effort applied to the same approach.

both types of failure feel, in the immediate aftermath, almost identical. this is what makes them so easy to misread. the wrong-direction failure produces the same sting as the not-yet failure. the double-sided urge to both persist and quit is the same. the sensation of failure does not come with a label indicating which kind you’re dealing with, and the most psychologically natural responses to failure (that are, to persist or to quit) are each the right answer to one type and exactly the wrong answer to the other.

the cultural bias here is worth naming: we have a strong preference for the story in which persistence was the answer. there is a social desire to label grit as a virtue and construct narratives around the people who succeeded being the ones who always kept going. this bias toward persistence is pervasive and genuinely harmful, because it treats “keep going” as universally correct advice when it is only correct in not-yet situations. the person who persists in a wrong-direction failure for years, because they’ve been told persistence is a virtue, is not being courageous. they are doing the same thing over and over again, expecting different results, which, colloquially, is the very definition of insanity.

there are things you should quit. there are directions that are not for you; goals that were based on a self-concept that has since updated; paths that made sense at seventeen that don’t make sense now. the failure that tells you this is not a lesser failure than the failure that says “not yet.” it is equally valid data. reading it correctly sometimes requires the courage to act on what it’s telling you, which is not always the more heroic-looking choice.

why we can’t read the signal

martin seligman’s research on explanatory style (that is, on how people explain the causes of events in their lives) identified three dimensions along which explanations vary: (1) whether the cause is permanent or temporary, (2) whether it’s pervasive or specific, and (3) whether it’s internal or external. the explanatory style you apply to failure is one of the most reliable predictors of what happens next.

pessimistic explanatory style applies permanent, pervasive, internal causes to failure: “i always fail at this kind of thing, i’m bad at everything, it’s who i am.” this produces what seligman called learned helplessness, which is the gradual extinction of effort in the face of repeated failure, because the model of why failure keeps happening offers no handle for change. if failure is evidence of fixed, global incapacity, there is nothing to do but stop.

optimistic explanatory style applies temporary, specific, external causes: “i failed this time, at this specific thing, for these specific reasons that i can address.” this preserves agency and motivation.

but here is where the research gets more complicated than the self-help version suggests: the optimistic style, applied indiscriminately, also prevents accurate learning. if every failure is attributed to bad luck, unfair circumstances, or temporary factors beyond your control, you never build an accurate causal model of what actually went wrong. the external attribution protects your self-esteem at the expense of the data. you leave the situation without having extracted the signal, and the next attempt is made with the same approach, because nothing has changed in your model of why the last one failed.

the failure-as-data approach is not optimistic or pessimistic. it requires something more specific: temporary and specific attributions that are nonetheless honest about what you actually contributed to the outcome. you don’t want to be constantly thinking “i failed because circumstances were unfair” and nor should you always default to “i failed because i am fundamentally not capable of this.” but, instead, find yourself somewhere closer to: “i failed because i did these specific things, used this specific approach, made these specific assumptions, which produced this specific gap between intention and result. what can i change?”

this is much harder than either defensive response, because it requires you to look directly at the failure and take it seriously as information about you and your approach, while simultaneously not taking it as information about your fixed capacity. holding both of those things at once is the cognitive and emotional work that accurate failure-reading actually demands.

failure can produce failure

defensive self-protection after failure produces more failure

when you fail at something and your primary response is defensive (meaning, you attribute the failure to external factors to protect your self-esteem, or reframe it as not-really-failure to protect your self-narrative) you exit the situation without having built an accurate model of what went wrong. the data was there, but you chose, understandably, not to process it in a way that was useful. instead you resolved the emotional discomfort and moved on.

the next time you encounter a similar situation, you bring the same approach, because nothing updated. the same failure is available. if it happens again, you’ve now produced a pattern, not of the failure itself, but of the failure-followed-by-defensive-processing-followed-by-no-change sequence. patterns, once they accumulate, start to function as evidence of fixed incapacity because you’ve now failed at this thing several times without any change in outcome, which begins to feel like data about what you’re capable of.

what began as a single failure that could have been one data point in a useful series has become, through repeated defensive processing, a self-fulfilling story about your relationship to this domain.

what accurate failure-reading requires

claude steele’s research on self-affirmation shows that when people are given an opportunity to affirm their core values before receiving critical feedback, they process that feedback more accurately and less defensively than people who receive it cold. the ego-threat has been pre-empted, which means the defensive response to the feedback is reduced.

the practical implication is not complicated, but it sounds a little like the self-help stuff you probably ignore (side note: sometimes, just sometimes, those self-help gurus really do know what they are talking about!): before you look hard at a failure, stabilise yourself in your values and your broader sense of self. your goals is to create enough psychological ground that looking directly at what went wrong doesn’t feel like an existential threat. because if it feels like an existential threat, you will not be able to look at it honestly, and you will leave without the data.

the questions that tend to produce accurate failure-reading are process questions, rather than capacity questions. instead of “am i capable of this?”, ask yourself “what specifically did i do, and what specifically didn’t work?”.

the difficulty here is that there is no clean algorithm for distinguishing wrong-direction from not-yet. there is no number of failures at which you should conclude it’s the former. that distinction requires judgment, and judgment requires personal engagement with the evidence of what you’ve tried and what has changed. it requires you to take the failures seriously enough to read them carefully, which is exactly what the defensive response protects you from having to do.

what i mean when i say i’ve failed more than i’ve succeeded

i mean that accurately accounting for failure requires being willing to not know, for a while, what a particular failure means. it requires that you sit with the question, “is this telling me to stop, or to continue differently?”, without resolving it prematurely into the most emotionally comfortable answer. this is uncomfortable in proportion to how much the attempt mattered, and the attempts that produce the most important data are exactly the ones that matter most, where the discomfort is highest and the temptation to reach for the defensive response is strongest.

the motivational poster version of failure offers you a story in which every failure was building toward something. sometimes that’s true. sometimes the failure was telling you something terminal and you needed to hear it. the actual value of failure is that it knows something you don’t about the gap between your model of the situation and the situation itself.

most of the time, we’re not willing to find out what our failure knows. and that - not the failure itself, but the unwillingness to read it - is where things tend to go wrong.

- erin x

this is not just a problem i face because i am what you might call a ‘content creator’. i’ve talked to people who’ve never posted anything more considered than a birthday message and they describe the same internal doubling: you notice that a meal you’re eating is especially beautiful partly because you’re already anticipating the story you’ll tell about it; you feel, even in the moment, that an argument is slightly staged because some part of you is watching it happen and already editing it for the retelling. the version of yourself that exists in other people’s heads has started exerting pressure on the version that moves through the world.

we did not develop language for this before social media existed, because the conditions that produce it are new. but the underlying mechanism is not; it was described very eloquently in 1902.

the looking-glass self

charles horton cooley was a sociologist writing at the turn of the last century who proposed that the self is not a fixed interior thing we carry around and occasionally reveal to others. it is something we construct, continuously and unconsciously, through a three-part loop: (1) we imagine how we appear to others, (2) we imagine the judgment others make of that appearance, and (3) we form our sense of ourselves through those imagined reactions. he called this the looking-glass self - the idea that other people function as a mirror in which we see, and come to know, ourselves.

what cooley was describing is not vanity, nor was it a form of social anxiety. it is far more structural than that: it is the claim that selfhood, at a foundational level, is an inherently social construction. you could not develop a self in isolation, because the self is built from social feedback.

this is, in important ways, true. the research on identity development bears it out: we need recognition to form a stable sense of who we are, we are exquisitely sensitive to social feedback from very early in our development, and the self is far more permeable to social influence than most of us like to think.

what cooley could not have anticipated is what happens when the mirror is always on, and when it is calibrated not to reflect the actual judgments of actual people who know you, but to amplify whatever provokes the strongest reaction in the largest number of strangers.

when the mirror becomes a metric

in cooley’s framework, the imagined reactions of others were exactly that: imagined. they were fuzzy and interpretable, coloured by your own projections and uncertainties. the people whose opinions shaped your self-concept were, for most of human history, the people who shared your physical life: your community, your family, the specific circle of people whose estimation of you you actually cared about.

social media replaces the imagined reaction with a number. a like count, a view count, an engagement rate, for example. the ambiguity of imagination collapses. you don’t have to imagine how the post landed because the post tells you exactly how it performed.

this sounds like it should be clarifying, even freeing. in practice it is neither of those things. because what the number measures is not what you actually want to know. it measures a specific kind of response (that is, one which is passive, momentary, and emotionally reactive) from an aggregated audience whose composition you can estimate but never fully know. the metric is precise in its number and almost meaningless in what it conveys. but your psyche doesn’t know that. it responds to the number the same way it responds to genuine human recognition, because the neural systems that process social approval were built for a world where a signal of social approval was, reliably, evidence of social approval.

what this means, in practice, is that the looking-glass self is now being shaped by a mirror that systematically rewards very certain things, with those things being legibility, emotional provocation, consistency of persona, the impression of intimacy, and the performance of relatability, to name but a few.

nobody is cynically optimising for these things, but they are so rewarded because these are the things that aggregate human attention, and the metric that tells you how you performed is, ultimately, a measure of aggregated attention. if you are checking that metric and letting it inform your self-concept, even unconsciously, then the mirror that is shaping who you are is one that was calibrated by an algorithm, not by the humans who actually know you.

the backstage problem

erving goffman, writing in 1959, proposed that social life is structured like theatre. we perform for ‘audiences’ by adjusting our presentation and playing the roles that different contexts demand. his most useful concept, for this conversation, is the distinction between front stage and backstage. front stage is where the performance happens. backstage is where you go when the curtain comes down and you drop your social character.

goffman thought of backstage as necessary and restorative because it is where the self that predates all the performances goes to exist without management.

social media compresses this backstage, sometimes almost to nothing. you post and view posts from the bed and from the kitchen and from the moments that used to be private by default. the physical architecture of privacy (things like the closed door) has not disappeared, but the psychological architecture has become much more porous. when you carry a device that connects you to an audience into every room, every moment becomes potentially front stage. the decision not to post is still a decision made in the presence of the possibility of posting. you are, in some sense, always aware of the curtain.

what concerns me about this is not that the performance is exhausting, although it sometimes is. it is that backstage (remember, this is the place where the unperformed self goes) requires practice to access. and if the conditions of contemporary life mean that you are rarely, if ever, genuinely backstage, then your ability to access it atrophies.

i don’t think most people have dramatically performed selves and secret authentic selves that are radically different from them. i think the performed version and the interior version are always in dialogue, each shaping the other. my worry is not that the performance is false. my worry is that you perform so much and so often that it is the only version of yourself that is left.

the creator version

when your online presence is not just consumption but production (especially when you have an audience who has developed expectations through having a sense of who you are and what you do and what you sound like), the loop cooley described tightens considerably. the audience’s response shapes what you produce. what you produce shapes who you present yourself as being. who you present yourself as being shapes who you become.

it is not necessarily negative. the audience for this blog has made me think more carefully and write more precisely than i would have done in private. the knowledge that these words will be read is not nothing for it has shaped what i think and how i think it in many ways i’m grateful for.

but it has also exerted a gravity of its own. the voice i use here is not not me. it is, in important ways, me: it is the version of me that exists in the act of writing to you, which is a real act performed by a real person. but it is not the full inventory of me, and i am aware, sometimes uncomfortably, of the distance between the version of myself that is most practised and the version that exists in the gaps between posts.

the question that sits underneath all of this, for me, is: do i know who that person is? the one in the gaps. the one that exists when there is no audience and no awareness of one. i think i do. but i am less certain than i would like to be, and i am more certain now than i was a few years ago that this certainty is worth caring about.

the question you can’t ask your phone

the looking-glass self, in cooley’s original formulation, was built through sustained relationships with actual people. the mirror that formed you was held by people who knew you in three dimensions. the recognition they offered was genuine recognition, which means it was also specific: they saw a particular person and responded to that particularity.

the mirror of social media is different in kind, not just degree. it does not see you in particular. it sees the signal you produce and aggregates the reactions of people who are responding to that signal, in a moment, through a screen, without the context of knowing you. the self-concept being shaped through that interaction is not being shaped by the people who know you most clearly. it is being shaped by the impressions of strangers, mediated by a platform optimised for engagement, and then quantified into metrics that are measuring something rather different from whether you are, in fact, being seen.

i don’t think the answer is to disengage, or to stop building in public, or to treat every moment of online presence as a psychological hazard. the answer, if there is one, is something more unglamorous: the deliberate maintenance of spaces where no performance is being offered or expected:

• conversations that go nowhere and won’t be written up

• experiences you don’t document

• states of being that aren’t legible enough to post

• thoughts that exist only in the time between having them and the moment they dissolve without becoming anything at all.

the self that predates all the performing needs somewhere to be.

the question i keep returning to is whether i’m giving it enough room.

and whether i’d know the answer, if i wasn’t.

- erin x

this finding has been replicated many times in many contexts: with origami, with lego, with drawings, with badly designed business proposals. the pattern is robust and spells out the result that personally investing our labour increases our perception of the value of the end product, largely independent of the actual quality of the result. norton and his colleagues called it the ikea effect. what it describes, in practice, is that effort is not a neutral input, but something which contaminates your ability to evaluate the thing you’ve worked on.

I chose that word on purpose. contaminate. not shape, not influence - contaminate. this is because the effect is not a gentle nudge toward overvaluing your own work, but a systematic distortion of your perception that, once understood, you start to notice operating almost everywhere in your life. and nowhere does it operate more insidiously than in the decisions that actually matter.

why effort makes things feel more valuable

the psychological mechanism behind the ikea effect runs through cognitive dissonance, which is leon festinger’s foundational insight that the human mind, when confronted with a gap between its actions and its beliefs, will resolve that gap by adjusting the beliefs rather than the actions, because adjusting the beliefs is easier.

festinger and carlsmith demonstrated this in a now-famous 1959 experiment. participants were asked to perform a profoundly boring task for an hour (which was turning pegs on a pegboard, one by one, over and over) and then to tell the next participant that the task had been interesting and enjoyable. half of them were paid twenty dollars to do this. half were paid one dollar. afterward, when asked how much they had actually enjoyed the task, the one-dollar group rated it significantly more enjoyable than the twenty-dollar group.

the counterintuitive logic here is precise. the twenty-dollar group had an external justification for what they’d done: they’d lied for a meaningful sum of money. the one-dollar group had no sufficient external justification. and in the absence of adequate external justification for their behaviour, their minds generated internal justification instead: the task must have been more interesting than it seemed, because otherwise, why would i have said it was?

effort operates the same way. when you have worked hard for something (and by this I mean really worked, in a way that cost you time and sleep and social plans and pieces of yourself) the psychic pressure to justify that investment is enormous. and the most available justification is not the retrospective audit of whether the effort was warranted that you might think would be optimal. to your neuropsychology, it is, instead, a largely unconscious revision of your perception of the thing itself: “this must be worth it, because i wouldn’t have worked this hard for something that wasn’t”.

what this does to how you think you understand things

for students, the most immediate version of this is about comprehension.

the essay you laboured over for three days, for example, feels more than the one that came quickly. it is not necessarily more polished, although it might be that too. what i mean is that it feels more hard-won. unfortunately, that feeling is not a reliable indicator of whether the essay is, in fact, any better.

aronson and mills showed this in 1959 with an initiation study. participants who underwent a severe, embarrassing initiation to join a group subsequently rated the group as significantly more interesting and worthwhile than participants who underwent a mild initiation or none. the group was the same group. the initiation changed the perception, not the object being perceived.

similarly, the knowledge you acquired through real struggle feels more solidly yours than knowledge that came easily. this is sometimes accurate: effortful processing does produce more durable encoding, and there is something to be said for the depth of understanding that comes from working through something hard. but this is not always the case. and the problem is that you cannot tell the difference from the inside.

the struggle makes you feel like you understand. it does not guarantee that you do.

the harder and less acknowledged version of this is about beliefs. the positions you have argued most strenuously and invested most identity into are not necessarily your most considered views. they are your most effortful ones. and the more effort you’ve put into defending a position, the harder it is to evaluate whether that position is actually correct, because the effort itself has raised the perceptual stakes.

being wrong about something you’ve argued loudly for is more painful than being wrong about something you held loosely.

so you become much better at finding reasons to maintain the positions you’ve invested in than at evaluating them honestly.

the part that actually matters

here is where this stops being a quirk of psychology experiments and becomes something that shapes the structure of a life.

the ikea effect does not just operate on furniture, or essays, or group memberships. it operates on relationships, career paths, and the image of yourself you have been effortfully constructing. and it does so through the same mechanism: the more you’ve put in, the more you’ve revised your perception of what you put it into upward, and the less able you are to evaluate it from anything resembling a neutral vantage point.

the relationship that has required the most work will feel more valuable than one that has been easy, regardless of which one is actually better for you because effort is what the brain uses as a proxy for value, and when you’ve invested enormously, the brain updates your sense of what you’re investing in to match. this is the mechanism that makes leaving hard relationships so much harder than it should be by any external measure: you’re walking away from the relationship as your effort has taught you to see it, which is more valuable than it actually is.

the same thing happens with career paths and academic directions. the degree you’ve spent three years working is three years of effort justification. it is three years of your brain quietly updating its assessment of whether this is the right direction in proportion to how hard you’ve been working. the harder it has been, the more compelling the case your own perception makes for continuing.

the sunk cost fallacy, the tendency to keep investing in something because of what you’ve already put in, is well known. what’s less well understood is why it’s so resistant even in people who know about it.

the reason is the effort paradox: the effort has genuinely distorted your ability to see the thing clearly. you can’t evaluate it the way you could have before you began, because you are not the person you were before you began.

the question you can’t answer from the inside

knowing about the effort paradox does not immunize you from it. this is worth being honest about, because a lot of psychology is presented as: here’s the bias, now you can overcome it. the evidence suggests this is largely not true for the ikea effect and its relatives. effort continues to inflate perceived value even in people who understand the mechanism in detail.

what knowing about it does allow, however, is the creation of deliberate distance before evaluating things you’ve worked hard for.

this is harder than it sounds, because the moment you most need to evaluate clearly (“should i continue? is this still right?”) is usually the moment you are most inside the investment, most unable to access a perspective that predates your effort.

the question that tends to cut through more reliably than most is not “is this good?”. the question that you should be asking is: “if i arrived at this for the first time tomorrow, having put nothing into it yet, would i choose it?”

not would i appreciate it, or would i see its value, or would i understand why past-me invested in it. would i choose it, now, for the future, starting from nothing?

the answer, for things that are genuinely right for you, is yes. the things that deserve your continued investment tend to survive that hypothetical.

the thing about the furniture

the ikea containers that people assembled in norton’s study were not actually better than the ones made by experts. they were identical. what changed was that the perceiver had sat down and worked through the instructions and pressed the tabs into the slots and turned the thing over in their hands.

we are all, in this sense, surrounded by our own flat-pack constructions: the beliefs we assembled argument by argument, the paths we built choice by choice, the versions of ourselves we put together from the available materials under time pressure, without always being sure we were following the right instructions.

we value them more than we should, not because we’re foolish, but because we made them.

- erin x

i’m not going to say that.

not because it isn’t partially true, but because as a complete statement it is the kind of thing that sounds liberating but is actually, for a lot of people in a lot of situations, actively harmful.

grades do matter. they open specific doors and close others in concrete ways, and the advice to stop caring about them is a luxury that not everyone reading this has access to.

what i want to say instead is harder but more useful: there is a relationship with grades that makes your life and ambitions very small, and i lived inside it for long enough to know what it costs.

the answer to that smallness is not to stop caring but to care differently: you must understand, with some precision, what grades are worth and what is worth sacrificing in order to achieve them. there is a line to be found between taking them seriously and letting them become the organising principle of your entire interior life.

that line, in my experience, is not where most people think it is.

the inside

if you’re reading this you probably recognise yourself in my description. for me, i was most caught in the “defining one’s entire worth through grades” loop during my gcses.

what i remember most clearly is not the studying itself but the way everything outside the studying started to feel irresponsible. the book i wanted to read that wasn’t on the reading list was frivolous. there were many conversations i cut short because i should be getting back to work.

without noticing it happening, the frame through which i evaluated what was worth my attention got very narrow. the question stopped being “is this interesting” or “does this matter” and became: “will this be on the exam”.

that is a very small aperture through which to engage with knowledge, and when you’re looking through a small aperture for long enough, you start to forget that the frame is a choice and not the natural shape of things.

goodhart’s law

in 1975, the economist charles goodhart articulated what has since become one of the most generative observations in the social sciences: when a measure becomes a target, it ceases to be a good measure.

goodhart was talking about monetary policy, but i feel the principle applies well to academic assessment. grades were designed as a measure: they are a compressed signal of understanding and of engagement with the quality of thinking that a course of study was supposed to produce. a good grade was supposed to be a byproduct of actually having learned something.

when the grade becomes the target rather than the byproduct, the whole system inverts. you are no longer optimising for understanding and producing a grade as evidence of it. you are optimising for the grade and producing a performance of understanding as a means to it.

these can look almost identical from the outside. from the inside, and in terms of what they leave you with, they are completely different.

the student optimising for the grade learns the mark scheme. they learn which arguments the examiner is likely to reward, which examples are safest, and which essay structure produces the most consistent results. they learn to write for the reader they’ve been given rather than to think for themselves. this produces grades.

the grade produced by this optimisation is no longer measuring what it was designed to measure; the signal has been corrupted by the act of targeting it. the grade says: this person knows how to get a good grade.

it no longer reliably says: this person understands the material and has developed the intellectual capacities this course was supposed to build.

you can spend three years at university optimising for a grade and emerge with a very good degree and a very thin education. the two are not the same thing.

Subscribe now

the opportunity cost

here is the question i wish someone had asked me during my most intensive revision periods: what is your revision costing you?

not in time, though it does cost time; in what the time would otherwise have been. in the reading you’re not doing, the project you’re not starting, the conversation you’re not having, the skill you’re not building, the thing you’re not making.

your revision, like anything else, has a real and specific opportunity cost that almost nobody accounts for when they decide that more studying is always better.

the research on the predictive validity of grades is, in this context, worth knowing. grades predict some things fairly well: performance in further academic study and first-year job outcomes in credential-dependent fields, for example. they are a legitimate signal for these purposes, but their predictive power decays significantly over time.

after five years in almost any career, grades explain very little of the variance in performance, advancement, or satisfaction. better predictors, according to a substantial body of research on job performance, are factors such as intellectual curiosity, communication ability, the capacity to learn new things, resilience in the face of setbacks, and the ability to work with and influence other people.

these are, almost without exception, things that grade optimisation tends to crowd out. or, at minimum, they are things that are not cultivated by the act of revision itself.

curiosity is built by following interesting threads, not by reviewing correct answers. communication is built by practising it, in contexts that require you to actually persuade or influence someone. resilience is built by attempting difficult things that might fail, not by perfecting things you already understand. the skills that will matter most in five years are largely developed in the time that intense revision consumes.

but grades do matter

i want to resist the version of this argument that becomes, at its conclusion, a sophisticated justification for not doing the work.

grades matter. they matter as a threshold signal in a world that is sorting people at scale and cannot evaluate every individual on their full complexity.

a competitive graduate programme, a scholarship, a training contract, a first job at a firm that uses degree classification as a filter - these are gates that grades open or close, and the people on the other side of those gates are not wrong to use the signal, however imperfect. they don’t have a better one.

for students looking to enter through those gates, the dismissal of grades as merely conventional or symbolic is not liberating. it is a luxury position that they can’t afford, delivered by someone who already got through the gate.

the question, then, is not whether grades matter. the question is: at what point does caring about grades start producing worse outcomes than it prevents? at what point does the anxiety, the narrowing, the crowding-out of everything else, cost more than the additional grade increment is worth?

and that point, in my experience, arrives earlier than most people expect, as is often the case where diminishing marginal returns take effect.

there is a tangible difference between a person who spent three years developing only for the exam and a person who spent three years developing genuinely: the former has a better grade. the latter often has a better education.

what the obsession costs

the thing about grade obsession that i think is hardest to see from inside it is that it doesn’t feel like a loss. it feels like virtue. it feels like effort and discipline and taking things seriously. and it is those things. but it is also, simultaneously, a narrowing of what you allow yourself to care about, and that narrowing has a compound cost that doesn’t fully manifest until you’re out the other side of it.

edward deci and richard ryan’s research on intrinsic versus extrinsic motivation found something that should give every grade-obsessed student pause: when you introduce extrinsic rewards for activities people already find intrinsically interesting, you reduce their intrinsic motivation to engage with those activities.

the grade is an extrinsic reward. and the more your relationship with your subject is structured around the grade, the more you risk eroding whatever intrinsic interest you arrived with.

the students who leave university with their curiosity intact are not, in my observation, the ones who cared least about grades. they are the ones who refused to let the grade be the only thing they cared about. they actively worked to keep the intrinsic thread alive alongside the extrinsic one.

what a better relationship with grades looks like

it doesn’t look like not caring. caring is appropriate. grades are a legitimate signal in a world of signals, and treating them with indifference is not sophisticated.

a better relationship looks like this: take grades seriously enough to do well. do not take them seriously enough to let them become the boundary of what’s worth knowing. do not take them seriously enough to sacrifice the reading, and the curiosity, and the risk-taking that the grade cannot capture but that the years of education are uniquely suited to produce.

the floor matters. the floor opens doors. optimise for the floor.

and then use the space above it, the space you create by not spending every available hour in pursuit of the ceiling, to become someone more interesting than your transcript can describe.

transcripts are read once, usually by someone who doesn’t know you, in a filtering process designed to reduce you to a number that can be compared to other numbers. that is a function they perform reasonably well and that matters for specific purposes at a specific moment.

but you will live inside your own mind for the rest of your life. the quality of that mind is not built by revision. it is built by everything that revision crowds out when you let it consume the whole available space.

don’t let it consume the whole available space.

erin x

a feeling of having decided that tomorrow, or monday, or the new term, or after this one last thing, you are going to become a different kind of person.

you are going to become more disciplined. more focused. more consistent.

you can feel that person. they exist, vividly, just ahead of you. and in the moments when you’re imagining them most clearly (like in making the plan, setting up the systems, watching the video about the morning routine, buying the notebook) there is something that feels almost like satisfaction. it feels like some part of you has already crossed a threshold.

and then you don’t cross it. not quite. you negotiate with your plan. your system lasts four days. your notebook stays mostly blank.

rather than sitting with what happened, you find yourself, almost without noticing, back at the beginning of the loop: planning again, imagining again, feeling, again, the comfort of being about to become someone better.

the possible self and what happens when you live inside it

in 1986, psychologists hazel markus and paula nurius introduced the concept of possible selves: the representations we hold of who we might become, both aspirational and feared. this includes the ideal selves we hope to realise, as well as the selves we’re anxious about becoming, and the selves we believe we could be if circumstances were different or effort were sustained. possible selves, markus and nurius argued, are not idle fantasy; instead, they are cognitively active, motivationally significant structures that shape how we interpret our current behaviour and what directions of change feel available to us.

this is the constructive function of the possible self: it gives you something to move toward. the student who can vividly imagine themselves as competent and engaged has a cognitive scaffold that organises effort in ways that the student with no such image doesn’t.

but there is a shadow function that the research opened up: when the possible self becomes load-bearing in your identity (i.e. when “i could be disciplined”, “i could succeed”, “i could be the person who shows up consistently” becomes not a direction of travel but a description of who you currently are) something structurally dangerous happens. the possible self stops being a goal and starts being a possession. the issues lies in the fact possessions, unlike goals, have to be protected.

action threatens the possible self. specifically: the action of actually testing whether you are, in fact, the person you believe you could be is threatening because the test might return a negative result. and if it does, you lose not just the attempt but the identity that was resting on the belief in the attempt’s potential success.

so the psychological architecture whose primary function is maintaining a coherent and positive sense of self does what it always does when the self is threatened: it avoids. it finds reasons to delay and it produces an endless succession of preliminary conditions that need to be met before the real work can begin, because the preliminary conditions are safe and the real work is not.

why imagining it feels almost as good as doing it

kent berridge, a neuroscientist at the university of michigan, spent decades untangling what had been assumed to be a single thing and turned out to be two completely different systems.

the assumption was that dopamine, the neurotransmitter most associated with reward and pleasure, was primarily responsible for the experience of enjoyment. the loop was thought to be: you do something pleasurable → dopamine releases → you feel good → you’re motivated to repeat it.

berridge found that this was wrong. dopamine is not primarily a pleasure neurotransmitter. it is a wanting neurotransmitter: dopamine is the neurochemical of the motivated pursuit of a reward that has not yet arrived. the system that produces the feeling of enjoyment (what berridge called the “liking” system) runs on different substrates entirely: opioid and endocannabinoid mechanisms, not dopamine.

wanting and liking are separable. you can want something intensely without liking it much when you get it. you can like something without particularly wanting it. the wanting system (the dopaminergic seeking circuit) fires not in response to the reward itself but in response to stimuli that predict the reward.

this means that imagining yourself as the disciplined, successful person you could be fires the seeking system in ways that are neurologically similar to, and sometimes stronger than, actually being that person.

the wanting is activated by the vision.

the liking, when and if the work is done, is less immediate than the wanting was.

this is why the planning high is real. it is also why the visualisation session produces something that feels like progress, and why watching a video about productivity generates a motivational feeling that is genuine and neurologically grounded and still leaves you no closer to doing the thing. the dopaminergic reward is delivered by the anticipation of the identity, not by the work required to earn it.

a brain that is receiving the reward without the effort has, from a purely motivational standpoint, limited reason to undergo the effort.

Subscribe now

the productivity content trap

i want to apply this specifically to something that i think is one of the more insidious features of the current information environment (and that I, admittedly, can be part of).

the internet has produced an ecosystem of content whose explicit purpose is to help you become more productive and more disciplined. it is, individually, often genuinely useful - I always try to ground my content in real research and produce real insight, for example.

the issue lies in what consuming it does, at a neurological level, when consumed as a substitute for the behaviour it describes rather than as a supplement to it.

watching a video about discipline is a performance of the identity of someone who takes self-improvement seriously. it activates the possible self: the version of you that has the morning routine and the consistent study habit and the organised system. and it delivers a dopaminergic signal that is the neurological signature of progress.

it does this without requiring any of the things the actual behaviour requires.

the consumption of productivity content at scale trains the brain to associate the idea of self-improvement with reward. the actual self-improvement, which involves difficulty and failure and imperfect execution, produces a much weaker and more delayed signal. the content is a shortcut that bypasses the effort while delivering the feeling of progress, and a brain offered a shortcut to reward will, with impressive consistency, take it.

if you have spent significant time consuming content about how to be more productive and disciplined and focused, and you have not become noticeably more productive and disciplined and focused, this is not because the content was bad or because you are irremediably lazy; it is because the content was doing something to your brain’s reward architecture that made the actual change less neurologically necessary.

the ego protection mechanism

claude steele’s research on self-affirmation established that when the self-image is threatened by failure, by evidence of inadequacy, or by the possibility of either, the brain mobilises resources to restore its integrity. it does this not primarily through addressing the source of the threat and instead looks to affirm the self in other domains.

in other words: the ego is not especially interested in accuracy. it is interested in coherence and positivity, and it will go to significant lengths (including elaborate self-deception about why now is not the right time to begin) to maintain a self-image that the test of action might complicate.

the person who believes that they could be disciplined if they really committed (i.e. the person believing that they have the prerequisite potential, intelligence, and capacity) is holding a self-concept that action would risk. this is not because action would necessarily disprove it, thought that may be the case, rather because action introduces uncertainty, and uncertainty means the verdict is no longer in your control. Losing control of the verdict is, to the ego’s accounting, worse than not seeking a verdict at all.

the potential self is, in this sense, an extraordinarily well-defended position. it is unfalsifiable as long as it remains untested. and the ego, which is in the business of maintaining positive self-regard, has every incentive to keep it that way.

what this produces, at the behavioural level, is a sustained high opinion of one’s own potential alongside a persistent gap between that potential and what’s actually being produced.

what actually breaks the loop

the wrong answer is more motivation. motivation, as i’ve written before, follows action rather than preceding it.

the wrong answer is also better planning, for the obvious reason that planning is the drug. more elaborate planning produces more of the dopaminergic anticipation signal and less of the actual behaviour change.

what actually works, according to the implementation intentions research by the same peter gollwitzer whose symbolic self-completion work is elsewhere in my essays, is the specific removal of the decision point. gollwitzer’s implementation intentions are if-then structures: “if it is tuesday at 3pm and i have finished lunch, then i will open my notes before i open anything else.” the if-then formulation bypasses the moment of choice (which is the moment in which the ego can produce reasons to delay) by making the behaviour automatic rather than deliberate.

the cognitive mechanism is the transfer of initiation from the effortful, ego-involved intentional system to the relatively automatic if-then system. the starting was decided in advance, in a moment of lower emotional investment, and then the trigger just fires.

the second thing that works is a specific reframe of the identity question. james clear, drawing on a body of research on identity-based change, argues that the most durable behaviour change doesn’t come from goal-setting (things like “i want to become disciplined”) but from identity claims grounded in evidence: “i am someone who did this thing today.”

this means you aren’t constantly defining only the possible self. you are, instead, defining the self who acted, however imperfectly, and can therefore make a specific evidenced claim about who they are.

the potential self is not real

the fully-realised version of yourself that the potential self promises does not exist and cannot be interacted with, protected, or inhabited.

the only self that exists is the one performing actions in the present tense, and the only evidence about who you are comes from those actions accumulated over time.

the addiction to potential is an addiction to a feeling that substitutes for the messier, less neurologically spectacular, more durable experience of actually moving. it is a comfortable place to live. it is also, structurally, a place from which nothing gets made and nothing about the actual self changes.

you are allowed to have potential. you are not allowed to mistake it for an achievement.

erin x

you can watch this piece on youtube :)

what do i mean by post-literate? i’m not making the argument that people are completely ceasing to read. they are not. more words are produced and technically consumed today than at any point in the history of civilisation. by the raw measures (things like words per day, text messages sent, articles published, characters typed) we are the most verbally active human beings who have ever existed.

what i’m actually describing is something subtler and, i think, more consequential than illiteracy. it is a change in the relationship between human minds and written language; and this looks like a shift in the cognitive habits and expectations that govern how most people encounter, process, and retain information expressed in words. you can be functionally literate in the conventional sense - able to read, able to write - and still be post-literate in the sense i mean. the question is not whether you can read. it is what happens inside you when you do, and whether that process still produces the effects that reading, historically, was responsible for.

what literacy does to the mind

reading is not a natural cognitive capacity. this is the foundational insight of maryanne wolf’s decades of research on the reading brain.

unlike language, which is encoded in dedicated neural architecture that develops in virtually every human being exposed to social contact, reading has to be learned. and not merely as a skill, in the way that swimming or driving is learned, but as a neurological technology: a set of processes that, when practised with sufficient depth and regularity, physically rewire the brain. the “reading circuit,” as wolf calls it (the complex network of neural regions recruited for deep reading, including areas originally evolved for object recognition, language processing, and motor planning) does not come pre-installed. we have to assemble it, slowly, through years of practice, and it can be disrupted by changes in that practice.

what this circuit, once assembled, makes possible is extraordinary. not just decoding words, but the full cognitive architecture of what wolf calls deep reading: the capacity to make inferences beyond what is explicitly stated, and to inhabit a perspective radically different from your own, and to slow down inside complex argument and follow it to its conclusion. these are, wolf argues, the cognitive foundation of critical though and of the kind of sustained reasoning that produces both scientific advancement and moral progress.

and they are specific to reading. not to language, to reading. the oral transmission of information, which was the dominant mode for most of human history, produces different cognitive habits. the shift from oral to literate culture was therefore a reorganisation of how minds engaged with ideas.

walter ong, the philosopher and literary theorist who spent his career mapping this transition, showed how the internalisation of alphabetic writing transformed not just what people knew but how they thought. it produced the capacity for sustained solitary thought and the linearity and sequential logic that literate cultures came to regard as rationality itself.

what is changing now

marshall mcluhan argued that “the medium is the message”, by which he meant that the form through which information reaches us shapes our cognitive habits far more fundamentally than the content does. television, he thought, was producing minds structured for television: fragmented, image-based, emotionally immediate, hostile to the sustained linear argument that print demanded.

he was right, and he was writing in the 1960s. the technology he was describing looks quaint now.

what the current moment has produced is not quite what ong or mcluhan anticipated, though both of them got closer than most. ong described what he called “secondary orality”, which he viewed as a return, through electronic media, to some features of oral culture: its communal participation, its present-tense focus, its emotional immediacy. but the secondary orality he described was still largely chosen and directed by humans. what we have now is something ong’s framework doesn’t capture: an information environment that is not only oral-participatory in character but algorithmically curated, in which the specific content any individual encounters is selected not by their own judgment or by a human editorial process but by a recommendation system optimised for engagement.

the result is a relationship with information that is structurally different from anything that has preceded it, and different in a specific way: it is optimised for the reader’s existing preferences and emotional reactions rather than for the expansion of them. the literate reader, moving through a book or an argued essay, is required to follow a sequence of thought not of their choosing. the algorithmic reader encounters, with every scroll, a succession of content items specifically selected to feel immediately relevant, emotionally engaging, and satisfying to encounter. it is content calibrated, at scale, to produce the neurological signature of interest without the cognitive labour that used to be required to generate it.

this is the post-literate condition as i understand it: not the absence of words, but the presence of an entirely new relationship to them; one in which the expectation of difficulty has been removed and the patience required for sustained linear argument has been systematically not reinforced. the reading brain, like any neural circuit that is not exercised, is slowly dismantling itself.

the evidence from the reading brain

what i find most striking about this is not a study about screen time or attention spans. it is an observation from wolf herself, one of the world’s leading experts on the neuroscience of reading, about her own mind.

after a year of immersion in digital reading (think: emails, articles, feeds; the normal texture of academic life in the twenty-first century) wolf sat down to read hermann broch’s the death of virgil, a novel she had read and loved, and found she could not do it at the pace she once had. the reading brain she had spent decades developing had been, through shifts in her daily practice, disrupted. the circuits were intact but the fluency had degraded.

she describes reaching the end of a page and realising she had retained almost nothing. she describes the experience that i think millions of people recognise from their own reading but rarely articulate clearly: the growing difficulty of staying inside a text that does not immediately reward you.

naomi baron at american university found similar patterns in her research on digital versus print reading: students reading on screens were more likely to skim, to skip to conclusions, to resist the effort of following sequential argument. when asked about their own reading behaviour, they reported knowing that they weren’t reading deeply but feeling unable to sustain the alternative. the habit and expectation had changed. the text hadn’t.

what is being lost at the population level

i want to try to be quite precise here, because i think the imprecise version (“young people don’t read anymore”) is both factually wrong and analytically useless.

what is being lost, at the population level and at a pace that i think we do not yet have adequate language for, is the specific cognitive infrastructure that deep literacy built: the expectation that understanding requires effort and time and that difficulty is not a sign that the material is wrong for you but a necessary condition of genuine engagement with it.

this infrastructure is the foundation of the capacity for sustained critical reasoning. it allows for the epistemic patience that makes it possible to genuinely understand a position you disagree with rather than simply encountering a summary of it. these are not, and should never be viewed as academic luxuries, but instead as the cognitive machinery of democratic self-governance and scientific reasoning.

neil postman, in “amusing ourselves to death,” argued in 1985 that the shift from a typography-based public discourse to a television-based one had already produced a culture in which the complex, sequential, argued proposition (of the kind that could be expressed in a book or a serious essay) was losing its authority to the image-based, emotionally immediate, entertainment-oriented logic of television. And it has only worsened since.

the irony that sits at the centre of all of this

the post-literate society is not a society without text, but a society drowning in it. more words are produced daily than the entire written output of several centuries combined and more information available to any individual than any previous generation could have imagined. we have access to more “content” than any one person could consume in a thousand lifetimes.

the problem is the quality of the relationship.

a society can be saturated with words and still have a post-literate cognitive culture, in the same way that a person can spend an hour in an art gallery looking at their phone and still claim to have visited the gallery. the engagement that the exposure was supposed to produce is largely absent.

what has happened, i think, is that the form of information has changed faster than our understanding of what form does. we have treated the shift from print to digital, and from chosen to algorithmically curated, as a change in delivery mechanism as if the difference between reading a book and scrolling a feed were roughly the same as the difference between a hardback and a paperback. it is not. the medium is the message. the form is producing minds structured for the form. and the form is not structured for the cognitive capacities that we continue to demand in our universities, our institutions, and our public life.

why i think about this in the context of learning

i make content about learning. my entire project is, in some sense, built on the premise that the appetite for genuine understanding is still there and that it can be reached, engaged, and sustained, even in a media environment optimised for the opposite.

and i believe that. i have seen enough evidence of it in the responses i get, in the things people write to me, in the communities that form around serious intellectual content, to believe it genuinely rather than as an act of faith.

but i also think that the effort required to reach that appetite (by this I mean the effort to bridge between the cognitive habits that the information environment reinforces and the cognitive habits that genuine learning demands) is greater than it has ever been. and that it will continue to increase. the students i write for are the first generation whose primary relationship with information was formed in the post-literate mode, and they are being asked to perform acts of sustained, effortful, patient reasoning in institutions that were designed for minds formed very differently.

this is not a reason for pessimism. wolf is clear on one thing that i think matters enormously: the reading brain, once formed, is not simply lost.

it is disrupted. and disrupted circuits can be rebuilt.

the neuroplasticity that made deep reading possible in the first place - the capacity of the brain to rewire itself through practice - does not disappear in adulthood.

the question is whether enough people understand what is at stake to make that effort consciously, rather than discovering what they’ve lost only when the task demands it and they find themselves, like wolf with broch, at the bottom of a page with almost nothing to show for it.

erin x

”my experience is what i agree to attend to.”

the five words at the end of that sentence are what i want you to sit with for a moment. he doesn’t say “what happens to me.”, nor “what i’m given.” he says what i agree to attend to. james was making a claim about the relationship between attention and experience that is so radical, and so plainly correct once you see it, that i’m not sure i can improve on it: your attention is not a tool you use to engage with your life. your attention, directed and sustained over time, is your life. the things you attend to are, in the most literal sense, what your existence consists of.

i want to start here, because i think many of the conversations about the attention economy start in the wrong place: they start with screen time and productivity and the number of times you check your phone. those things matter, of course, but they’re downstream of the more fundamental question, which is: what kind of thing is attention, and what is actually happening when a system is designed to extract it from you at scale?

what the attention economy actually is

the phrase “attention economy” was coined by the psychologist and nobel laureate herbert simon in 1971, and it describes something architecturally different from previous economic models.

in a conventional economy, a company produces something and sells it to you. you are the customer. in the attention economy, the product is not something sold to you. it is something extracted from you (whether that be your attention, your behavioural data, your psychological profile, or some similar quantity) and sold to a third party. instead of the customer, you are the raw material from which a product is manufactured.

interestingly, the product is not (not precisely, at least) your attention itself. it is your predictability and the extraordinarily detailed model of your psychology (inclusive of everything from your anxieties, to your desires, to your susceptibilities; all building up an image of the specific combination of stimuli that will reliably keep you engaged) that is assembled from your behaviour at scale and sold to anyone who wants to influence it. what is being sold is the ability to predict and shape what you do.

this business model operates transparently within the terms of service you agreed to without reading, and it is the foundational logic of the most widely used communications infrastructure in human history.

never before in the history of human civilisation has there existed a system whose explicit commercial purpose was the continuous, large-scale extraction and monetisation of human attention. the romans had bread and circuses. they did not have a team of two thousand engineers optimising the circuses in real time based on which specific gladiators made each specific audience member most likely to return tomorrow.

the asymmetry problem

the people designing these systems (or, more accurately, the systems themselves as they are increasingly powered by agentic AI) have access to, and have, over twenty years, developed, the most sophisticated applied understanding of human psychology that has ever existed. they know, from hundreds of millions of behavioural data points, exactly how variable reward schedules produce compulsive checking behaviour. they know the relationship between social validation metrics and dopamine release. they know which notification timing maximises the probability of re-engagement. they know that the autoplay feature that begins the next video before the current one ends removes the decision point that would allow you to stop, and they know exactly how many additional minutes of engagement that produces per user per day.

skinner’s variable ratio reinforcement schedule is the design principle behind these systems. this is the mechanism behind slot machine addiction, in which rewards are delivered at unpredictable intervals and in unpredictable quantities, producing the highest rates of compulsive behaviour of any reward structure ever studied.

on the other side of this asymmetry exists you. your willpower is finite and depletes under cognitive load. your awareness of this problem does not, as the research consistently shows, reliably protect you from it.

this is not a fair fight, and i think recognising that clearly, without euphemism, is the necessary first step: you are a human brain encountering a system specifically engineered to overcome human brains.

what it takes from you that isn’t time

most of the discourse about the attention economy focuses on numerical metrics: hours per day, screen time reports, the aggregate lifetime hours spent on particular platforms. those numbers are significant, of course, but i think the focus on time metrics misses what is, neurologically and experientially, the more important cost.

sophie leroy, a management professor at the university of washington, introduced a concept she calls attention residue, and this is the finding that, when you shift your attention from one task to another, a portion of your cognitive resources remains attached to the previous task.

gloria mark at uc irvine found, in a study that i talk about a lot on social media, that after a digital interruption it takes an average of twenty-three minutes and fifteen seconds to return to a state of deep focus on the original task. twenty-three minutes. and this is not just for a long distraction, it applies for any interruption, however brief. this means that just one notification leading to one moment of context-switching will cost you nearly half an hour of cognitive depth.

what this means, practically, is that the attention economy does not steal two hours of your day in one transaction. it steals it in increments of twenty-three minutes, dozens of times, in a way that really feels like nothing each time it happens. so, the study session that should have produced three hours of deep work, instead produces forty-five minutes of work and two hours of interrupted, attention-residue-compromised near-work that generates the feeling of effort without much of its output.

what gets crowded out

there is a mode of mental activity that is among the most important things a mind can do, and that the attention economy is systematically eliminating from the daily experience of an entire generation.

neuroscientists call it the default mode network, and this is a set of brain regions that become highly active during states of undirected, internally-focused thought (which essentially means daydreaming and mind-wandering). for a long time this network was called the “resting state” network, because it activates when you’re not doing anything specific, and, at the time, the assumption was that “not doing anything specific” meant the brain was doing nothing of importance.

the research has since established that this was entirely wrong. the default mode network is not idling. it is doing some of the most cognitively significant work undertaken by a brain: consolidating memories, processing emotional experiences, generating creative connections between disparate pieces of knowledge, and constructing and updating the narrative self. these are foundational functions to the capacity for deep thought, self-knowledge, and original work.

the default mode network activates specifically in the absence of external stimulation. translation: it requires boredom, or something close to it. it just cannot run concurrently with the task-positive network, which handles focused, externally-directed engagement. the two networks are in something close to neurological opposition.

what the attention economy has done, at scale, is eliminate the conditions under which the default mode network can operate. the smartphone, and social media specifically, means that the unoccupied moment no longer exists. think about it; every space that would previously have been available for the undirected, internally-focused cognition that the default mode network performs - every queue or commute or interval between tasks (or even waiting for the microwave to finish!) - is now filled with algorithmically selected stimulation.

the gaps are gone. and with them, a significant portion of the cognitive processing that used to happen in them.

the consent problem

i want to say one more thing, perhaps a little more controversial: you did not consent to this in any meaningful sense of the word consent.

you were handed a phone, probably in early adolescence, at a stage of neurological development when the prefrontal cortex (the region responsible for impulse control, risk assessment and planning, and the evaluation of long-term consequences) was still years from maturity.

the terms of service you agreed to were tens of thousands of words long, written in legal language, designed to be unread and practically unreadable.

the product you were receiving had been specifically engineered, by teams of experts in behavioural psychology, to be as habit-forming as possible.

the social infrastructure around the platform had already become essential to your social life, so the opt-out was not realistically available.

this is not voluntary participation in an economic transaction, but is probably something closer to the gradual assumption of a condition you were never meaningfully asked about. i think understanding that clearly, not as a reason for outrage or helplessness, but as an accurate description of what happened, is important context for the self-compassion you extend to yourself about how difficult it is to change the relationship.

you didn’t choose this dynamic freely. you can still choose to engage with it differently from here. those are both true simultaneously.

what actually helps (and it’s not deleting the apps)

the research on digital wellbeing consistently finds that restriction alone (covering things like deleting apps, setting timers, and locking the phone in another room) produces temporary change that is difficult to sustain, because it treats the symptom rather than the underlying structure. this structure is that the default relationship with these platforms is passive, reactive, and algorithmically directed. changing this default relationship is more durable than trying to limit it.

what that looks like in practice is something more like reclaiming what you might call attentional sovereignty: the deliberate exercise of choice about what you attend to, when, and for how long, in a way that is structured enough to function as a real alternative to the default.

doing this means designating conditions under which you do not have the phone available at all, not because you’re disciplined but because the option has been architecturally removed. it also means embracing choosing what you will consume rather than accepting what the algorithm selects for you. most importantly, it means reintroducing protected periods of genuine boredom, like walks without audio and meals without screens, so that the default mode network has the conditions it needs to operate.

and it really means understanding, with some clarity, what is at stake. william james again: your experience is what you agree to attend to. the attention economy is, in the most literal sense james intended, making a claim on your experience, on what your life consists of, moment by moment, at scale, for as long as you are a user.

please do not misunderstand me: this is not an argument for rejection. it is an argument for understanding the nature of the thing you’re engaged with, so that your engagement can be a choice rather than a default.

your attention is the substance of your experience. the question of what you give it to is not a productivity question. it is a question of how you want to spend your life.

erin x