Henrik Karlsson · blog Escaping Flatland, on Substack under his own domain, henrikkarlsson.xyz · Swedish, born July 1989, now living on a farm on a small Danish island in the Baltic with his wife Johanna and their two daughters; many essays are written with her · Still active: the archive starts in October 2021 and holds 118 posts as of September 2026; only 9 posts in 2021–2022, then a jump to roughly 30 a year from 2023; writing became his full-time job on 23 November 2024 · About four in ten posts are paid-only (48 of 118); this page is based on the public part. Dostoevsky as lover, which he marked as his 2023 favorite, is paywalled and was not read in full · This page is based on 14 public essays spanning October 2021 to May 2026: 3 from the earliest fifth of the blog's life (2021-10, 2022-09, 2022-10), and both sides of the two turning points (the 2023 jump in output, and going full-time in late 2024; three essays from 2025-04, 2025-07, 2026-05) · Selection: the four essays he himself pins on his About page come first; then internal backlink counts (across 70 public posts, Looking for Alice is linked from 16 others, the most; Everything that turned out well… and Relationships are coevolutionary loops from 10 each; Cultivating a state of mind… from 9); plus The Third Chair, which he marked "My favorite" in his 2024 year-end summary
Henrik Karlsson's most-read essay is dating advice. In the first paragraph, he admits he doesn't know much about dating.
What he does know is how he found his wife. In 2011 he gave a poetry reading in a bookstore, saw Johanna, and assumed she was there to see him. She was not. A month later, hungover in a café, they met again. He suggested they do something someday; she smiled and said yes. He told her he had drunk his phone away, so he asked her to take his number. She did. She did not text him. For the next eighteen months he stopped to talk with her every time he saw her — in the supermarket, at clubs, in the street — while working hard to make her feel he wasn't obsessed. Then one night in April 2013 they happened to walk the same path and reached the river. He lived on one side, she on the other. "Where are we going?" he asked, trying to sound as if he meant geography. She nodded toward her block.
In her room, piled with books and half-finished canvases, he did the thing they still joke about: he started talking about how ashamed he was of himself and how out of control his life was, with long references to St. Augustine. By any dating manual, this is suicide. Johanna was mesmerized. She had never met anyone who took his own thoughts with such earnestness and was so unashamed of his pain.
The advice he drew from that night is the best way into this whole page:
That is perhaps the most solid dating advice I have, by the way—show the inside of your head in public, so people can see if they would like to live in there.Looking for Alice, 2023-01-17
Notice the words "in public." This isn't only about dating. It is the operating principle of his blog.
The year before, he had written an essay about Ivan Illich and systems thinking — a topic he had never found anyone else intrigued by, and which magazines thought below a rejection letter. He put it online. One morning he woke up to find that people who wanted to talk about exactly this had found him. "I had conjured a minor conference!" The first few times a stranger arrived in his inbox, excited by exactly the same things, Johanna asked him what was wrong. He was crying in the kitchen.
Those were tears of homecoming.A blog post is a very long and complex search query…, 2022-10-05
Hence his best-known metaphor: a blog post is a very long and complex search query — one that finds people, not pages. Standard advice tells you to write accessibly, to cut digressions and obscure references so more people can follow. He says that cuts exactly the things that would help your people find you. Someone once told the composer Morton Feldman he should write for "the man in the street." Feldman went to the window and looked out. Who did he see? Jackson Pollock.
Write for Jackson Pollock.Ibid.
Write like that and almost no one will like it. But almost no one, multiplied by the population of the internet, is plenty — if you can find them.
Why does he care so much about who finds him? This is the genuinely original move. He thinks humans are animals that cannot avoid being shaped by the people around them: two-and-a-half-year-olds are about even with chimpanzees at solving problems alone, and pull far ahead only at watching others and copying them. Whatever surrounds you is what you become, and you can't switch it off. As a young poet he gave about a hundred readings and learned to predict when the room would laugh and when it would cry. Then the poems went hollow: "The poems, in conforming to their laughs and tears, hollowed out." To write what he needed to write, he spent the better part of a decade isolated.
So for him writing isn't self-expression. It is choosing who you will become. What you write decides who comes to you. Who comes to you decides what flows into your head. What flows in decides who you are. He read forty-two biographies of geniuses and found that the most consistent pattern wasn't talent or teaching method. It was that they grew up surrounded by exceptional adults. When Pascal's father found that his son had secretly re-derived Euclid's proofs, the family moved to Paris so that father and son could join Mersenne's mathematical salon. Karlsson boils it down to one line:
It is by changing your milieu that you change yourself.First we shape our social graph; then it shapes us, 2022-09-07
Looking back, that night in Johanna's room was the smallest version of the same move: put the strangest part of you out first, let the wrong people leave, let the right person stay. He married the one who stayed. Then he did it again, with a blog, to the whole internet.
One winter morning in late 2021, Henrik Karlsson woke before sunrise and noticed that something had changed.
He had recently written an essay about Ivan Illich and systems thinking. In all his years of writing he had never met anyone else intrigued by the topic, and magazines thought it below a rejection letter. He put it online. Then, overnight — "as the fields lay frozen and waiting and the hedgehog slept in its pile of leaves" — the internet quietly rearranged itself around him:
I had written for 15 years, but never before had this happened. I had conjured a minor conference! And I hadn't even known that you could do that.A blog post is a very long and complex search query…, 2022-10-05
After that, strangers started appearing in his inbox, excited about exactly the same things. He would drop his most obscure references and they would reply: read that, love it. The first handful of times, Johanna asked him what was wrong. He was crying in the kitchen.
A man who had been writing for fifteen years found out for the first time that writing could summon people. Nearly everything he wrote over the next four years answers the question that morning raised: if words can summon people, and people shape you, who should you summon?
Born in Sweden in July 1989, "of the last generation who will remember the time before the internet." Before becoming a writer he was a programmer, a factory worker, a poet, a teacher, a lab technician and an art gallery coordinator. As a quality controller at a medical factory, he once had to spend four nights, from 7 pm to 7:20 am, in a windowless room reading detailed instructions on how to safely wash his hands. It later became his favorite image for what goes wrong in education.
He met Johanna in a bookstore in 2011. Both later worked as teachers, and both concluded they would not put their own children through what they had seen. Homeschooling is illegal in Sweden, so they spent two years travelling to four countries to talk to homeschooling families and people who ran democratic schools. In 2020 they left Sweden for Denmark, found an island, rented an apartment there for nine months, and finally bought a farm. They moved in one February and had simply assumed there would be firewood. The former owner had left ten logs. They pulled the sofa up to the fireplace, burned one log at a time, and slept in beanies, watching their breath rise white above them.
The blog began in 2021. There were only 9 posts in its first two years, then about 30 a year from 2023. On 23 November 2024, writing became his full-time job. He says that much of what he publishes grows out of conversations with Johanna, "about 10,000 hours into our conversation."
He ran into the problem that would occupy him for a decade on a stage.
In his early twenties he passed himself off as a poet for a few years. He could recite Tomas Tranströmer's collection Baltics from memory. A director at the Royal Dramatic Theater took him under his wing and taught him how to move an audience, and the invitations to read kept coming. Because the stanzas were fixed, he couldn't make the split-second correction we all make before speaking. He had to reveal his innermost thoughts and "stare into the mouth of a woman yawning." People left in the middle of a sentence, and he had to finish it. Over a hundred or so readings he learned to model those reactions in advance. That changed the writing, and it changed him:
The problem was rather this: the gravity field of this particular audience did not align with my ethics and aesthetics: their expectations pulled me away from the places my thoughts needed to go. The poems, in conforming to their laughs and tears, hollowed out.First we shape our social graph; then it shapes us, 2022-09-07
To write what he was meant to write, he needed a different audience. He didn't yet know how to curate a milieu, so the only option left was to isolate himself for the better part of a decade. He calls it "bleak but necessary."
That is his root problem, and it can be stated plainly: people can't help being shaped by those around them. So how does anyone grow into themselves instead of being flattened into the shape of other people's expectations? At this point the blog's name explains itself.
The problem pulls in two directions, and his writing moves back and forth between them. On one side, your milieu makes you, so you should go out and find, summon and filter the right people. On the other, your milieu can swallow you, so you have to learn to be alone and protect the ideas that aren't ready to be seen. In his 2024 year-end summary he put the two forces side by side:
People who emphasize how to get things done, and move faster, and reach higher levels of mastery—agency without introspection—tend to get pulled into fierce yet boring status competitions. People who emphasize introspection and emotions but without a problem-solving mentality tend to get stuck, and painfully aware of being stuck.A summary of what I wrote in 2024 (quoting his earlier Thoughts on agency)
He says he doesn't write self-help. He writes about self-cultivation in the classical sense, bildung, a distinction Johanna pointed out to him. He describes his writing through a parable by Borges: a man spends his life drawing the world, its provinces, bays, ships and fish, and shortly before he dies discovers that the lines trace his own face. Every December he skims what he published that year and finds that essays he wrote on a whim "are shadows thrown at different angles from the same object." The four ideas below are four sides of that object.
In July 2021 a daughter was born. He watched the nurses carry away this small creature, smeared in white-yellow fat and remembering nothing, and what came to mind was a question from anthropology. Chimpanzees are born into the habitat their genes expect and get by largely on instinct. Humans are born into a world their genes have never seen: the orange glow of sodium street lamps, an oxygen mask in a green-clad nurse's hands. How do we survive?
Through what anthropologists call cultural learning. In an earlier essay he had cited a comparison: two-and-a-half-year-olds and chimpanzees or orangutans are about even at handling tools and solving problems on their own. The only clear difference is in watching others and copying them. Small children extract knowledge from adults just by watching them move about a room. They come to want what the people around them want, and they shift their dialect to match their peers. And you can't turn it off:
This ability is not something you can turn on and off. You are always internalizing the culture around you. Even when you wish you didn't. So you better surround yourself with something you want inside—curate a culture.First we shape our social graph; then it shapes us, 2022-09-07
The usual view is that people grow through effort, through grit. He thinks that gets the unit wrong. He draws a person's milieu as a directed graph. Books, friends, the accounts you follow, even the architecture of a Gothic church, all send you input. The readers you write for and the five-year-old watching you write receive your output. What you become depends on the total flow in and out. He borrows a passage from Nadia Asparouhova and then adds his own twist:
Grit is the node's eye's view. You are struggling against the graph. Agency, on the other hand, is the view from the graph. You are the graph. By changing it, you are changing yourself.Ibid.
His main example is the guitarist John Frusciante, who writes songs by picking a set of guitarists and playing along to their records until new ideas start to come. He carries a map of guitarists in his head and knows he is trying to bridge Jeff Beck's lyrical technique and Kurt Cobain's reckless abandon. Other musicians play whatever catches their fancy. Frusciante deliberately tunes what he takes in.
But input is only half of it. The other half is what he learned on the poetry stage: the people you write for pull you toward them like gravity. So the question became how to get the right people to receive your output.
The answer was that morning in late 2021. He studied the "conference" like an alien artifact. He went through the traffic data, talked to readers, and kept "gleefully and semi-randomly pushing every button I could find to see what happened." He came away with two findings.
The first is about how to write. The more precise and niche the words, the better the internet matches you with people you can build real relationships with. Mass media teaches you to be broad and a bit bland. He didn't want a general public. He wanted a specific set of people whose names he didn't know, only that they existed. So an essay had to work like a query that finds those people and, if necessary, filters out everyone else. How do you decide what to write? Ask what would have made you jump off your chair six months ago. Write that, and don't dumb it down, because you weren't stupid six months ago; you just knew less. He says that's how he wrote his essay on using GPT-3 to augment learning: 3,000 words, long real dialogues, a bit literary and flamboyant, "which you are not 'supposed' to do... at least not if you want the average LessWrong reader to upvote it." A few thousand people turned out to feel the way he did.
If you follow the advice above, you will write essays that almost no one likes. Luckily, almost no one multiplied by the entire population of the internet is plenty if you can only find them.A blog post is a very long and complex search query…, 2022-10-05
The second is about distribution. He had pictured it as a corporate ladder: first a few readers as obscure as him, then slowly levelling up until larger accounts noticed him. It turned out to work nothing like that. He used a tweet as a crude A/B test and traced where it went. He had about a hundred followers. Two slightly larger accounts retweeted it, accounts an order of magnitude larger saw it, and within an hour or two it had reached the largest accounts it would ever reach (one with 84k followers retweeted it, one with 681k liked it). Spreading back down to other small accounts took much longer.
The social structure of the internet, he says, is shaped like a river. Big accounts are the mouth of the Mississippi, fed by countless tributaries. Messages rush up the stream order and then trickle back down. "The larger accounts are not larger without a reason, but, at least in part, because they spend more time routing information in the network!" None of this is new. In the 1600s, a scholar who wanted to reach another scholar would send the letter to Marin Mersenne, who knew everyone who knew anyone. So the strategy is simple: hand your writing to the node closest to the kind of person you want to reach, whether that's a subreddit, someone you got to know in the replies, or a direct message. Once you've built up enough connections you become a small hub yourself, and good material starts flowing to you. He says he especially enjoys being attacked by intelligent people, because he then invites them to comment on his upcoming drafts.
He takes the metaphor as far as it goes at the end of the essay. LessWrong was summoned by a sequence of search queries written by Yudkowsky and Hanson, strong enough that the community outlived the departure of its original nodes. In 1963 J. C. R. Licklider wrote a memo about an "intergalactic computer network":
We're all living inside his search query now.Ibid.
If you read only the first idea, you'd take him for a networker: connect more, publish more. What he did next pushed his own conclusion toward its opposite.
In 2022 he was working on an essay about the childhoods of geniuses, and his method was crude. He wrote down twenty names that "a random Swedish person can recall... and think, Sounds like a genius to me" (von Neumann, Tolstoy, Curie, Pascal), ordered piles of biographies, read them and took notes. To reduce the bias, he asked whether he could think of anyone exceptional who didn't fit the patterns. He could, so he ordered more biographies. "This kept going for an unhealthy amount of time." He ended up skimming 42.
The first pattern was expected and confirmed the first idea: more than 95 percent grew up among exceptional adults. Virginia Woolf never went to school, and her father could invite Henry James, Thomas Hardy and Tennyson to dine with his children. From the age of three, J. S. Mill sat at his father's desk while his father wrote a ten-volume History of British India, allowed to interrupt every time he hit a Greek word he didn't know. Of the Pascals, Karlsson writes:
The instinct was to curate a culture, not to teach, not primarily.Childhoods of exceptional people, 2023-02-06
The second pattern also covered about 95 percent, and it pulled against the first: they had huge amounts of unsupervised time, and they were often bored. Russell was kept on his grandparents' slowly decaying estate by a grandmother so afraid he would overwork that she kept his lessons short, and his most important hours were spent wandering the garden alone. Maxwell, after his tutor hit him over the head, ran into a lake and refused to come out until his parents fired the tutor. Descartes was often bedridden as a child. The fields these people later became famous for, Karlsson writes, often came to them "almost like a wild hallucination induced by overdosing on boredom." He also tosses out a "spicy hypothesis" that he says he won't test: socializing too much with other children may simply not be good for your intellectual development.
That raises a real question. Your milieu shapes you, yet good ideas seem to need distance from your milieu. Which is it?
He went looking for the answer in working notes written for no one else. With Johanna he read Alexander Grothendieck's Récoltes et Semailles (more than 2,000 pages, by some accounts) and the workbooks Ingmar Bergman kept from 1955 to 2001. The Grothendieck passage is close to an answer in itself. After the war he ended up at a university that wasn't a center for mathematics. Annoyed that his teachers couldn't even define length properly, he spent the years from 17 to 20 alone, reinventing measure theory and the Lebesgue integral. Only in Paris did he learn that all of it had been done decades earlier, and his mentors told him he had simply wasted his time. His Paris classmates were quicker, more brilliant and better trained. Looking back after thirty years, he wrote that none of them had left a deep mark on mathematics, because they had worked inside circles drawn by someone else. To break out, they would have had to rediscover a birthright: "the capacity to be alone."
Karlsson reads one layer further:
This might look wasteful if you think what they are doing is research. But it is not if you realize that they are building up their ability to perceive the evolution of their own thought, their capacity for attention.Cultivating a state of mind where new ideas are born, 2023-07-26
That reconciles the two patterns. Solitude isn't for avoiding people. It lets you learn what undiluted curiosity feels like, so that inside any group you can tell whether you're excited because something has real potential or because it will raise your status there. Grothendieck did join the mathematical world, with mentors and collaborators, but he used that world from inside his own framework instead of being absorbed by it.
Then Karlsson divides this kind of discernment into three levels, the best passage in the essay. Level one is noticing that you find an idea exciting. Level two is the singer Kristian Mattsson, who puts the guitar down the moment he feels pure excitement while writing, because excitement means he is playing something he knows others will like, a path he has already been praised for. The songs he wants are the ones "he's ashamed of liking." Level three is Tarkovsky. Before Solaris, a sound engineer Tarkovsky considered brilliant told him not to use Bach, because "everyone is using Bach in their films at the moment." The diary records no reply. The music in the film is Bach. He saw that being popular had nothing to do with being right, in either direction.
That is very hard, Karlsson says, so most creative people stop at level two and learn that popularity is a trap. Some good ideas die needlessly that way, but more would die if popularity were allowed to kill unpopular ideas. He sums it up:
The zone is a place at the margin of society. But strangely enough, this fragile margin is where the ideas that make our society possible come from.Ibid.
The third idea starts with a failure. Applying to university, he looked through the catalog and thought: wouldn't it be cool to be a diplomat! He spent half a year getting into the political science program and going through "all sorts of schlep." Eighteen months later he realized he didn't like it at all. He panicked and dropped out, and then came up with a new vision.
Years later he noticed that everything that had turned out well in his life — his marriage, some of his essays, his career — had followed a different path. He took its name from the architect Christopher Alexander: unfolding. In a 1964 book Alexander defined good design as fit between form and context: a glove is well designed if it fits the hand. Karlsson extends this to a life. A relationship is good if it fits the people in it. An essay is good if it fits the truth, the writer's intellectual needs and the reader's mind. The useful thing about this definition is that it tells you where to look:
The context is smarter than you. It holds more nuance and information than you can fit in your head. Collaborate with it.Everything that turned out well in my life followed the same design process, 2024-08-05
To find a good relationship, you don't start by saying "I want a relationship that looks like this." That's starting from the form, the wrong end. You say: I'll pay attention to what happens with different people and iterate toward what feels alive. Observe the context, form a guess, let the guess decide your next small step, take in the new information the step produces, update, repeat. The opposite is "an unfounded vision": a dream that comes from fantasy rather than from understanding the context, the hope that if you could just get to some other place your problems would disappear. The diplomat plan was a vision.
His most convincing example is buried in a footnote: how the family came to the island. They needed a big leap, since homeschooling wasn't possible in Sweden. But the leap was unfolded. It came out of years of reading and working in schools. Once they had ruled out Sweden, they spent two years talking to homeschoolers and democratic schools in four countries. They found the island, visited twice, rented for nine months, and only then bought the farm. The control group was the other families he watched arrive with a romantic vision of island life and homeschooling. The first thing they did was buy a house; it was in the wrong part of the island, and moldy. A year later they fled to another island, or divorced, or burned out.
The biggest enemy of unfolding is your own preconceptions, and he offers a concrete tool against them: unbundle. When we think about careers we think "a job," but a job is a bundle of salary, identity, relationships, status and meaning. Only once you pull them apart can you see which parts you actually care about. A year later, in "On agency," he turned the tool on himself. At twenty-two he found that writing essays made him feel more alive than anything else, and Johanna told him to write more. He said there was no point. He knew the Stockholm literary world, and Swedish magazines didn't pay for essays. "No one has ever made a living as an essayist," he said. He wrote novels instead, because publishers wanted novels. He never cared about novels enough to get good at them, and his writing fizzled out.
I had conflated "being a writer" with "having a publisher" and "getting a salary from my writing." These are not the same thing.On agency, 2025-07-07
He could have taken a decent day job and used it to fund his essays. It was that simple. Philip Glass paid for his music with a moving company, a plumbing business and a stint driving a cab. Today Henrik and Johanna feed their family by writing essays, and she likes to tease him about it.
The part most worth noting is that he later corrected himself. In a footnote to "On agency," he admits that an essay urging you to define clear goals sits awkwardly next to an earlier one urging you to follow the context wherever it leads. Walking up and down the road by the farm, he concluded that it's a false dichotomy. A bad goal specifies what the solution should look like ("sign with a major Swedish publisher"). A good goal specifies only what the solution must achieve: food on the table, good friends, 30 hours a week to write, "but, as long as I get that, I'm agnostic about what the specific solution is." The second kind of goal fits with unfolding and may even be what unfolding needs. He says plainly: "I think I was a bit confused about this in the essay from last year."
The essay on unfolding ends with my favorite line on this page:
You will have lost track of all the experiments and insights that led you to a fit. But the good news is that you don't have to remember. The form does.Everything that turned out well in my life followed the same design process, 2024-08-05
The title of the "Looking for Alice" series comes from a story. Someone asked Gertrude Stein if she was a lesbian. She answered no, I just like Alice (her partner, Alice B. Toklas). Many people think Stein was lying. Karlsson thinks she had it right:
But hers is the right attitude: you do not like a category. You like individuals. And you're not born knowing which kind.Looking for Alice, 2023-01-17
He noticed that his friends who kept dating the wrong people tended to think in categories. After a relationship went up in flames, instead of updating ("avoid traits A, B and C"), they would say, "oh, you know . . . girls." At seventeen his own crushes were random, "as if I was an untrained neural network trying to predict which image was a cat." He was disappointed a lot, which was great, because by his early twenties he could roughly say what he liked.
The method of the series carries the "search" from the first idea into the most private part of life: extract as much information as possible, as fast as possible. Don't go to a café to talk safely; skip straight to the strange parts. How? He points to Werner Herzog's documentary Into the Abyss. Herzog is interviewing a reverend who, fifteen minutes later, will accompany a boy to his execution by injection. The reverend is talking about the Lord working in mysterious ways, words he has said before, words that protect him. In passing he mentions that he sometimes meditates on the beauty of squirrels on the golf course. Standing in a graveyard of nameless crosses, with mad Bavarian seriousness, Herzog says:
"Please describe an encounter with a squirrel."Herzog, as retold in Looking for Alice
The reverend's voice lifts and he starts talking faster, no longer repeating himself. About ten seconds later he is crying. Asked in interviews how he knew to say it, Herzog answers: "I know the heart of men." Karlsson says that's remarkable under time pressure, but it's also fairly simple: ask about something the person has rich experience of but has never drawn an answer from, or ask two or three increasingly detailed follow-up questions.
The third lesson is one he nearly failed himself. Great relationships are often hard to compress into a sensible string of words, and we are social creatures. When you describe your lover and your friends look confused, you feel bad. When they get excited, you feel a surge of affection for your date. So the field tilts toward relationships that are easy to explain. He remembers "with a cold sweat" that he almost turned Johanna down, because he couldn't explain what they had and his friends weren't excited, while they were excited about more legible people who wanted to date him. His witness is War and Peace. Prince Vasili keeps seating Pierre and Hélène together at parties, the guests get excited about what they take for the start of a great romance, and Pierre, confused, decides everyone must be right. They marry. "And it is, of course, the type of relationship that makes you want to commit suicide-by-duel."
You will not be able to explain how this fusion works. So don't do it.Looking for Alice, 2023-01-17
And once you've found the person? Part three, "Relationships are coevolutionary loops," opens with Darwin. In 1862 he received an orchid from Madagascar whose nectary was as long as his forearm. "Good heavens," he wrote, "what insect can suck it?" He inferred that a moth must exist with a tongue longer than any ever observed, and that moth and orchid had evolved in dialogue. When Karlsson meets someone who is comfortably, fully themselves, he says, he assumes someone gave them room to unfold.
He treats a relationship as a system you can iterate on, and admits that this is "a strange and rather mechanical way of talking about relationships." After moving into his grandparents' old house, he and Johanna spent every Sunday listing everything that had gone wrong that week on a blackboard they had painted on the wall of his study. It was their bug log. He wanted to get up at 5 a.m. and work for three hours, but kept sleeping in. Johanna asked why. He wasn't sleepy at night. He gave up coffee, tweaked things for a few weeks, and the mornings clicked into place. Early on, he thinks, iteration speed mattered more than anything else: adapt to each other too slowly and you grow apart.
A 2025 essay about his friend Torbjörn applies the same idea to friendship. They went to middle school and high school together, and Henrik saw a very funny farm kid who listened to Rammstein and drove an orange Volvo 240. For fifteen years Torbjörn was an extra in his life. Then, around 25, Henrik wanted someone to talk to about software engineering, took the train out to Torbjörn's place, and they talked for hours while walking on the lawn. It was Torbjörn's wife, laughing at them from the window, who pointed out later that they had been walking in circles. He realized he had misread him. Torbjörn wasn't funny; he was responsive. In a group that showed up as riffing on shared jokes. One-on-one it meant finishing each other's sentences and getting straight to the heart of things. Henrik's mistake was twofold: he had looked for shared interests instead of character, and he had seen Torbjörn in only one context and assumed that was who he was.
If someone seems boring to you, or a bad fit, it might be that you don't know how to prompt them, that you haven't seen them react to the context that brings out their full being.Sometimes the reason you can't find people you resonate with…, 2025-04-23
The latest turn of the loop is from 2026. In their first spring on the farm, Johanna fell in love with gardening; her head was full of Latin plant names, and he found the flowers boring. He cut down half an acre of shrubs, learned to use a chainsaw and sharpened poles for a fence. He did it out of love, but he couldn't connect with what she cared about. They lived "a thousand miles apart in the same house." One night in June 2023 she was talking about a flower that looked like a raw egg and whether its yellow would work against the yew hedge, and he noticed he didn't care either way. So he ran an experiment: he would close his eyes and keep them closed until he could see what she was seeing. He asked why she cared so much about this yellow. She explained ideas from Goethe's theory of color, and he got excited, because her insights about the design process applied directly to writing essays. A bridge was built. Later he sat beside her as she studied Monet. One evening at twilight he walked into the woods, and the pines he had been so bored by had every needle shimmering, distinct. The red of their trunks was the red of Monet's boat.
If you love someone well, you have to run to keep up, growing your heart to hold what they are growing into.Love is about being invested in someone's continual expansion, 2026-05-20
From 2022's "change your milieu and you change yourself" to here, one claim has come full circle. The milieu that changes you most is the person you love, and loving her well means letting her change so much that you have to change too.
Write only about things that can hit you on the head. His rule is to choose topics with practical consequences for his own life or for people he cares about: "if I'm wrong, reality will hit me over my head as I practice the ideas I write about." In areas like politics, where he can't affect outcomes directly, he says he doesn't trust his judgment and doesn't write. That explains why nearly everything he writes is about love, parenting, writing and moving house. It isn't a narrow outlook. He only bets where there is feedback.
Use many real people as data, then go looking for counterexamples. To work out how to raise children, he didn't read parenting books; he read 42 biographies. To understand where ideas come from, he didn't read creativity research; he read Grothendieck's and Bergman's working notes and Tarkovsky's diary. He admits the samples are unsystematic, but he has a habit: pull out the pattern, then ask "So whom does this not apply to?" and add whoever comes to mind.
Make ideas rigid so they can break. This comes from Lakatos's Proofs and Refutations, a book of mathematical philosophy. In your head an idea stays fluid, and you can always say "I didn't mean it like that." On the page it stiffens, and once it stiffens you can see the cracks:
Seeing your ideas crumble can be a frustrating experience, but it is the point if you are writing to think. You want it to break. It is in the cracks the light shines in.How to think in writing, 2024-04-25
First make the sharpest claim you can. Then "stretch it thin": ask why it might be true, and spread the conclusion into a chain of premises so criticism has more to aim at. His example is his six-year-old, Maud, insisting in a shop that they buy a pink plastic toy teapot. He couldn't come up with a compassionate counterargument, so he asked why she thought it was so great. "Because it is so beautiful. And I need one in plastic so it doesn't break. I would use it all the time." "See—trying to prove her point, she had given me three times as many claims to attack!" Then he separates two kinds of counterexample. A local one knocks out a premise but leaves the conclusion standing, so the explanation needs repair. A global one cracks all the way up to the conclusion. That is the one he hopes for, because it leaves a hole where something deeper can go.
Don't dismiss the faint sense that something is off. A slight tension across the chest, eyes fogging over on a reread: he says these signals are the easiest to brush aside and the most likely to lead somewhere new. He learned this from Johanna, who will sit with a draft for hours, neither writing nor editing, just putting into words why something feels wrong. Their best essays came out of those sessions, he says. Friends who write and publish quickly "generate texts filled with hidden doors and move on before they've opened them."
Write what would have made you jump off your chair six months ago, without dumbing it down. As described in the first idea, this is both how he picks topics and how he finds his people.
Learn who you are by doing things, not by staring at yourself. He quotes Nick Cave's reply to a reader: you don't need to know who you are to become an artist. His addition is that Cave isn't rejecting introspection. Cave introspects a different way: make a song, notice how it feels, as if running a million small experiments. "I get anxious and confused when I meditate on myself, and I discover myself in my work." He adds an honest caveat. After a year of intense work he fell ill and spent three weeks in bed, and got up a new person. "Looking for Alice" was the first thing he wrote afterwards.
The last one is the most characteristic: every December, look back. He says he follows his interests at random and drops a topic for something unrelated once it bores him. Only at year's end can he see that the essays are shadows of one object. His year-end summaries aren't report cards. They are a search for that face.
First, his most-cited piece of "research" is far weaker as evidence than it is popular. "Childhoods of exceptional people" rests on an unsystematic sample of 42 biographies, and the figures are his own estimates. On the share of people who were tutored, he writes outright:
I'm basically making these numbers up; it is an informed guess.Childhoods of exceptional people, 2023-02-06
When the essay was crossposted to LessWrong, the comments made sharp points. Srdjan Miletic raised selection effects: "those with unusual educations are more likely to have written about them," while successful people with ordinary educations may never have made it onto his shelf. Ege Erdil argued that even a correlation would show little: to judge whether tutoring mattered, you'd at least have to control "for obvious confounders such as family wealth and genetic endowment," which "is impossible to do simply by reading Wikipedia articles about people who have been successful." matthewp put it most bluntly: he couldn't see much "to rule out an alternative summary: get born into a rich, well-connected family." Two hundred years ago, private tutoring was the default for the gentry. In the essays read for this page, I found no direct response from Karlsson to these points.
Second, "unfolding" has a self-confirming shape. The essay's title is "Everything that turned out well in my life followed the same design process," so the sample starts out containing only things that turned out well. What about things that went through the same process and didn't? His counterexamples (the families with moldy houses who later divorced) are other people, and they are explained as failures to unfold. When a method's successes are credited to the method and its failures to not using it properly, it becomes very hard to falsify. To be fair, he patched one seam himself in his 2025 footnote, admitting the tension between following the context and setting goals and calling his earlier self "a bit confused."
Third, his examples come from the extreme tail, while his advice is aimed at everyone. Grothendieck, Bergman, Tarkovsky, Herzog, von Neumann: he draws patterns from them and then encourages readers to build a capacity for solitude or to act like Herzog. In "Childhoods" he did honestly include a section titled "They were gifted children," and pushed back on Erik Hoel for overstating tutoring. In other essays that caveat tends to disappear. A related blind spot is circumstance. The kind of "unfolding" open to a family where both parents work from home, earn a living by writing and are willing to change countries for their child is not open to every family, even though he also writes that the move left them in relative poverty for four years.
Fourth, there is a gap between what he says and how he publishes. He defines a blog as a search query for summoning your people, yet about forty percent of that query now sits behind a paywall, and several essays end with a note that he is "not yet fully funded." This isn't an accusation. Writers need to eat, and the public sixty percent works fine as a query. But for someone whose core method is to show the inside of your head in public, the ratio is worth noticing.
The sharpest criticism is one he wrote himself. In his 2024 summary he admits:
This bias toward the local means my writing borders on self-help—and I've been confused by that since I don't like self-help.A summary of what I wrote in 2024, 2024-12-12
In the same piece he says his earlier essays framing relationships as a way to grow and discover yourself "felt instrumental in a way that didn't feel true to what I wanted to say," and that he had moved toward treating relationships as valuable in themselves. The "iteration speed" and "bug log" framing is exactly where a reader would grow suspicious, and he got suspicious first. Perhaps the test for him is whether he keeps doing that: whenever a beautiful frame is about to harden into a method, make it rigid himself, and let it break.