Hormesis
Here is a fact that should be stranger than it is. Put a healthy person in a bed and take care of their every need: food, warmth, no stress, no load. They get weaker. Not maintained at some baseline. Weaker. Muscle goes first, then bone. Astronauts lose bone density in orbit not because space is hostile but because it is too kind. Nothing pulls on the skeleton, so the skeleton, sensibly, stops paying for itself.
The bone is not damaged by load. The bone is built by load. Stress it, within a range, and it overcompensates, laying down more material than the stress strictly required, because the body reads the stress as a forecast of stress to come. Remove the load and the same logic runs in reverse. Comfort, to a bone, is a signal that strength is no longer needed.
This is not a metaphor I am stretching. It has a name, hormesis, and it shows up anywhere a system can adapt. Muscles. The immune system. The liver’s handling of mild toxins. The way a little fasting triggers cellular cleanup the well-fed cell never bothers with. The tree that never feels wind grows shallow roots. The comfortable system is not preserved. It is quietly disarmed.
Once you have seen the shape of this you start seeing it everywhere, and the seeing is uncomfortable. Skill works this way. You only get better at the edge of what you cannot quite do, which is by definition the part that feels bad. Courage works this way. Judgment works this way. You acquire it almost exclusively from decisions that cost you. Every faculty I actually respect in another person turns out to have been laid down the way bone is laid down, by being loaded right up to the point where it nearly failed.
The thing that reliably prevents all of it is comfort arriving too early.
The cellular version of this claim is uncontroversial. Any doctor will tell you a sedentary patient is in trouble. The version that gets you in trouble is the same claim moved up one level, to the scale of a human life: that the discomforts most of us spend our adult energy trying to avoid, work that is too hard, expectations we cannot quite meet, the sense that we have not yet arrived, are not bugs to be eliminated. They are the load. Take them away and the same thing happens to a life that happens to a bone. It looks fine for a while and then quietly fails to be anything.
You can see the same shape at a different scale again if you look at who built the things in your life you would not give back. The country you live in, in many cases the country that took your grandparents in. The companies whose products you use every day. The art that has actually moved you. The medicine keeping someone you love alive. None of it was built by comfortable people. The American republic was built by people who were not having a good time. The soldiers who freed Europe in 1944 had cold feet and wet socks and were, in many cases, eighteen.
The immigrants who founded the tech companies whose products you are reading this on did not arrive comfortable.1 They arrived with nothing and a great deal to prove. Almost every founder I have met whose company I would actually use was kicked around as a kid in some specific way.2 Bullied. Outsider. Broke. Wrong religion in the wrong town. They were running from something, and the running is what got them where they got.
There is an obvious bad argument in the vicinity, and it is worth naming so I can refuse it. It goes: great people suffered, therefore suffering makes greatness, therefore we should arrange suffering. This is a sampling error of the most embarrassing kind. You see the bullied kid who became a founder because becoming a founder is what put him in your dataset. You do not see the much larger number of bullied kids who became anxious, or addicted, or simply smaller versions of who they would have been, because nobody writes essays about them. Damage is not dosed discomfort. A hurricane is not a set of heavy squats. The same load that builds the bone, applied all at once and without consent, shatters it.
So I am not arguing that suffering is good, and I would never want a child beaten or bullied or starved. That is not a disclaimer. It is the actual content of the argument. Hormesis is a story about dose. The moment you drop the dose you do not have a bolder version of the theory. You have a different and false one.
The milder, harder claim is that the historical pattern and the cellular pattern point at the same thing. The shape of strength, at every scale we can look at, is load applied on purpose. The voluntary part is not a footnote. It is the whole mechanism.
I am aware this cuts against almost everything you will read. The reigning advice is to rest, to be gentle, to honor your limits, to be kind to yourself, to celebrate small wins, to give every child a participation trophy so nobody feels left out. Every reputable scholar writing on motivation will tell you that the harsh version of any of this is bad. None of them are going to publish the opposite.
I think this is happening for a specific reason worth naming. Most advice is written for the median, and the median person is not overworking. The median person is not within striking distance of building something that did not exist before. So rest more, be kinder to yourself, you are enough is correct on average and exactly wrong for the specific person it is failing: the one who has confused the average prescription for a personal one. If you are reading an essay called Hormesis voluntarily, on a weekend, you are probably not the median. The advice that is good for the population can be exactly the advice that is quietly capping you.
I am not claiming the research is wrong. I am claiming it is aggregated, and you are not an aggregate.
The corollary of all this is the part most people will not say out loud, because saying it out loud is what gets you labelled. If discomfort is what builds, then the voice in your head whose job is to remind you that you are not yet where you said you would be is doing useful work. The modern view of that voice is that it is a wound. That self-compassion is what actually correlates with performance. That the people who run on self-contempt mostly burn out.
I think this view conflates two voices that feel identical from the outside and do opposite work. There is a voice that says you are worthless, and a voice that says the draft is not done. Both phrase it as not good enough. The first concludes therefore stop. The second concludes therefore continue. The advice “be kinder to yourself” is the right cure for the first. For the second it is a sedative, and the sedative is what you reach for when you have quietly stopped believing you can do the thing.
What the research is mostly measuring, I think, is people whose harsh voice is the first kind. Of course they underperform. Their voice is telling them to stop. The people whose harsh voice is the second kind mostly do not show up in the studies, because the second voice is not a complaint. It is an operating principle, and the people running on it are too busy to write papers about themselves.
I write all of this partly to myself, because I have noticed that the current is strong enough now that even keeping the distinction is something I have to do on purpose. The temptation, when you have worked for years and finally have a little room, is to read the room as a finish line. To start being kinder to yourself in the specific way that, six months later, you realize was the thing that quietly closed the gap you had been keeping open on purpose. The closing felt like rest. It turned out to be surrender, dressed as wisdom.
The practice, as far as I have one, is just a series of small refusals. Almost buying yourself something you have not earned and then closing the tab. Catching yourself about to call a draft finished when it is not, and writing one more version. Noticing that the easy way to handle a problem would also be the way that quietly admits you have stopped trying, and not taking it. None of these refusals are large. The largeness comes from how many of them you stack up over years, and from the fact that nobody is watching when you make them.
The hardest place to apply any of this is children. I should be honest that I do not have any yet.3
The cartoon version of the argument ends with kids being denied affection and called names on the theory that hardship builds character. That version is wrong, for all the reasons in the dose section. Love is not the load. Safety is not the load. The load is something narrower, and confusing the two is exactly the failure I have been describing in every other domain.
There is a softer version of the same mistake, and the culture has gone all in on it. The mechanism is the steady cushioning of small losses, on the theory that losing is harmful. A skinned knee gets a reaction calibrated to a broken arm. A lost game gets reframed before the kid has felt the losing. A bad grade gets blamed on the test. None of these is dramatic by itself. Stacked up over a decade, they remove every form of losing the child was supposed to learn from.
The bone in zero gravity does not get stronger. It gets weaker. A child who never loses does not become resilient. They become a person who has never had to come back from anything, and who therefore has no model for what coming back even feels like. The first real loss arrives in their twenties, uncushioned, and a generation that should be at the height of its drive is instead trying to figure out why a routine setback feels like the end of the world. It feels like the end of the world because they have no calibration. Nobody let them lose when it was small.
The kind move, the one I think most parents and teachers believe they are making, is to spare the child the small losses. The actual effect is to remove the only training set the child was going to get on a problem they are guaranteed to face later, at scale, alone. You meet the second voice early or you meet it late. Meeting it late, for the first time, in adulthood, with real stakes, is the version that breaks people. Cushioning the small losses isn’t kindness. It is the bone in orbit, applied to a person.
Where this fails. The first failure is dose. Turn the volume to maximum and stay uncomfortable stops being hormesis and becomes the thing that produces burnout, injury, and the quiet breakage that does not announce itself until it is done. Load is not damage only within a range. The range is real. Ignoring it is how this philosophy hurts the very people most attracted to it, which is the conscientious ones who will happily run themselves into the ground if you give them a noble reason to.
The second failure is that not everything in a life is a bone. Some things really do break under load and do not come back stronger. Sleep is not a muscle you can train to need less of. It is a constraint, and treating it as weakness is just stupidity with good posture. The same goes for the people you love. A relationship is not improved by being harder on it. There are domains where the right move is the load, and domains where the right move is care, and the skill, the actual skill this whole essay is in service of, is telling which is which. I get it wrong sometimes. I have, more than once, read a real limit as a weakness to push through, and paid for it.
The honest version of this philosophy is not everything must hurt. It is narrower and harder. Seek the discomfort that builds. Refuse the comfort that disarms. Never confuse either one with damage or with a real limit. Most of the difficulty is in the telling-apart, which is unglamorous, which is exactly why the cartoon version, just suffer, suffer always, suffer more, is the more popular one. The cartoon does not ask you to think. It just asks you to hurt, which some part of us finds suspiciously easy to agree to.
None of this is advice I have finished taking. I have not built the thing I am trying to build, and I am not writing from any high ground. I am writing it down because I have noticed that the days I spend trying to be kinder to myself are the days I get less done, and because I have come to suspect that the gap between who I am and who I am trying to be is the thing I should be protecting, not closing. The day I stop having to remind myself of any of this is probably the day I have stopped doing it.
Andy Grove arrived in New York from Hungary at twenty, deaf in one ear, speaking little English, with relatives he had never met. Sergey Brin’s family left the Soviet Union when he was six because being Jewish made his father’s academic career impossible. Jensen Huang was nine when his parents sent him from Taiwan to a boarding school in Kentucky that turned out to be a reform school. Jan Koum was on food stamps in Mountain View. The pattern is not that every founder had this story. The pattern is how often the story is there when you look. ↩︎
I am hedging on purpose. Almost every is not every, and the company I would actually use is doing a lot of filtering. Plenty of comfortable people start companies. Plenty of those companies are fine. The claim is about the smaller set whose work I find myself returning to, and even there the correlation is not a law. Counterexamples exist. I have just met fewer of them than the other kind. ↩︎
Which means everything in this section is theory. I have watched other people’s children and I have been a child, and that is all. I would not be surprised if a parent reading this thought I had underestimated something obvious. If so, the part most likely to be wrong is the dose, not the shape. ↩︎