> ## Content Index
> Fetch the complete content index at: https://under-revision.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# On the Cost of Fixing
- URL: https://under-revision.ghost.io/on-the-cost-of-fixing/
- Published: 2026-10-03T12:09:06.000Z
- Updated: 2026-10-04T22:39:48.000Z
- Author: Dean Hatton

> *The stone that the builders rejected has become the cornerstone.*  
> Psalm 118:22 (ESV)

## I.

Not long ago I came across an ad for the Horowitz Andreessen Academy, a new school in San Francisco for young people who would rather build things than spend four years in college. It admits students ages sixteen to twenty-two. No grades, no degree. There are compute credits, courses in inference engineering and the craft of selling, and lectures from the people who run the largest technology companies in the world. Among its founding partners is Meta.

I found the ad strangely satisfying. It took just 90 seconds to confirm a prejudice I'd been carrying: most of our troubles with technology began with a division of knowledge. We taught some people how to build powerful things and left a shrinking group of others to ask what those powerful things are for. And it seems the builders rarely wandered into the other building. The questions that AI now makes urgent are ancient ones: what a flourishing life requires, why we want what we want, and what we owe to a person who must be treated as an end and never as a means. The builders, I assumed, had never been introduced to them.

Facebook was my proof. Tocqueville imagined a power that would spare citizens the trouble of thinking, attend to their pleasures, and keep them in perpetual childhood, and I could think of no better description of the algorithmic feed. Here was the tutelary power he feared, built by people who had never read anything like him. And here was an academy, with Facebook's parent among its sponsors, proposing to train the next generation of builders in a way that will limit how long they stay with a problem before turning it into a project.

## II.

The trouble with a confirmed prejudice is that it stops you from checking. When I finally did, the proof fell apart in my hands. Mark Zuckerberg studied Latin and Greek at Exeter, has spoken of his love for the Aeneid, and admires Augustus enough to have named a daughter after him. Peter Thiel, the first outside investor in Facebook, studied philosophy at Stanford, where he came under the influence of René Girard, and he has credited Girard's theory of mimetic desire with shaping his bet on the company. Girard taught that we learn to want what others want and that this imitation breeds rivalry and envy. Thiel saw that a network built on watching one another want things would grow very large.

So the builders had been introduced. They had read the books, some of them closely, and at least one had taken a profound account of human desire and found in it an investment thesis. The division of knowledge I had blamed was real enough, but these men had crossed it and returned unchanged, carrying what they found as a tool with a commercial purpose.

That left me with a harder question than the one I started with. If reading the right books does not form a person, then the difference must lie in how the books are read and in what the reader is willing to let them do to him. I had assumed that exposure was the thing missing. Exposure, it turned out, was cheap. What mattered was a way of attending to a text that I could not yet name. Before I could judge anyone else's reading, I had to admit how poorly I had understood my own.

## III.

For most of my life I understood attention as the instrument of usefulness. You looked at something to find what was wrong with it, and you stopped looking once you knew what to do. I was good at this. From my first job out of college, I sought to turn fluency into advancement, using the tools of statistical process control to solve quality and productivity problems, first for my workgroup and before long for an entire business. Attention served the fix, and the fix was its reward.

I brought the same attention home. I took providing more seriously than nearly anything, and I believed it included fixing. When one of my children struggled, I looked for the cause and the remedy. When my wife was troubled, I listened for the problem inside the trouble, the part I could fix. I thought of this as love, and it was a kind of love. It was also a way of seeing the people closest to me as conditions to be improved.

A good remedy can follow close, loving attention. But my readiness to offer a remedy often meant I stopped looking too soon. What I could not see was that fixing required me to decide in advance who they were, since to fix something you must already know what it is supposed to be. Every remedy carried a picture of the person I was remedying, and the picture was mine; anything in them that departed from it registered as variance to be reduced. Because my diagnoses were often correct, I had little reason to doubt them. A provider who solves problems well rarely hears that he has misunderstood the people he is solving them for.

I also cleared paths. Some instinct stopped me where clearing would have made them dependent, but within that limit I removed a great many obstacles, and it seldom occurred to me that some were theirs to meet.

This was the attention of a builder, and it shared its essential gesture with the attention that built the feed: a way of looking at people as problems to be anticipated and solved. I had faulted the founders for a way of looking that I practiced every day at my own table, even if I never carried it as far. I did not find the words for it on my own; Iris Murdoch gave them to me a few years ago.

## IV.

Murdoch took the word attention from Simone Weil and defined it as "a just and loving gaze directed upon an individual reality." Only after several readings did I understand that its weight falls on those last two words. An individual is precisely what a fixer cannot see, since a person's particular features are what his remedies are designed to correct. A reality is something that exists apart from my purposes for it. Elsewhere Murdoch wrote that love is "the extremely difficult realization that something other than oneself is real." I had always assumed the difficulty of love lay in the effort it demanded. She placed it in perception.

I now count Murdoch’s modest illustration among the most important passages I have read. A mother, M, dislikes her son's wife, D. She finds her pert and brusque, a little vulgar, tiresomely young. M behaves impeccably throughout, and no one would guess what she thinks. Then, over time and by deliberate effort, she looks again. She asks herself whether she has been jealous, snobbish, or unfair. Gradually D comes to appear spontaneous where she had seemed undignified; refreshingly simple where she had seemed vulgar; and delightfully youthful where she had seemed tiresome. Murdoch adds that D may have moved away or died by the time this happens. Nothing M does in the world changes. Nothing is fixed. Murdoch insists that M has done something of real moral importance, because we can act well only within the world we can see.

When I tried M's discipline on my own family, the first thing I noticed was how seldom I had looked at them without a purpose. Once I stopped fixing, I saw a willingness to take chances that had been there all along. At twelve, my daughter announced that she would go to the best school possible, wherever it was, and she did, though it meant leaving home. At the time, part of me treated her decision like I treated most things: as a set of problems to be managed. How will she handle the separation? What if she needs us? Only years later, reading Murdoch, did I look at the decision itself and see what it was. A twelve-year-old had chosen distance and difficulty because she wanted the best of something, and she had learned that way of choosing by watching me. We had moved the family across the country and back in pursuit of opportunity, and I had considered those moves to be my risks, borne by them. She had taken them as instruction. My remedies had never left room to notice it.

I could not have seen them this way at forty. It took decades of remedies, many of them correct, before I could recognize what they had kept me from seeing. Nor would it be honest to say the change was easy. Watching someone you love take a chance you could have removed and possibly fail at it has been the hardest discipline I have tried to practice as a provider. Fixing, I came to understand, had been a way of sparing myself that watching.

## V.

When I first called on Tocqueville, I quoted him selectively. The passage I had in mind contains a comparison I passed over because it seemed to concern someone else. Such a power, he writes, would resemble a father's authority if, like a father's, its aim were to prepare people for adulthood. Instead, it seeks to keep them in perpetual childhood. On that line, at least, I think I stood on the right side. Whatever my appetite for fixing, it stopped where fixing would have made my family dependent, and my children's independence bears that out. But Tocqueville's test measures what care produces, and Murdoch's measures how well it sees. I passed the first by instinct. It took me most of my life to notice I was failing the second.

The feed fails both. It is the fixer's attention with the love taken out and the scale made limitless. It watches each of us more closely than any parent could, for the same reason I watched my children: to anticipate our needs and supply them before we reach. Its designers speak of removing friction, and friction is their word for every obstacle a person might have met and grown by meeting. Murdoch's attention treats a person as a reality to be seen. The feed treats attention as a resource to capture and the person as the ground where it is found.

Mark Zuckerberg has recently published his own account of the future, and its diagnosis could have come from Tocqueville. He warns that hoping an absolute power will benevolently provide for humanity has rarely ended well, and he concludes that no one should decide for others what the good life is. His remedy is to give everyone a personal agent that understands their goals and works continuously to improve their relationships, health, career, and home: the tutelary power is distributed instead of abolished. He offers an example from his life. His daughter loves to bake, so each weekend his agent plans recipes for them, orders the ingredients, and offers suggestions as they work. I have no reason to doubt his love for her, and I recognize the impulse; I would once have called it providing. The agent improves the baking. What it displaces, though no one intends it, is the attention a father and daughter might have spent on the problem and on each other. He also promises every student a tutor with unlimited patience, which would relieve the student of the patience he is meant to acquire.

The danger I take most seriously is narrower than his promises. An AI can supply an answer the moment a question is asked, before the question has done its work on the one who asked it. That work matters more than it seems. A question held for a while, turned over and left unresolved, shapes the person holding it; it is how we discover what we actually want to know, and often what we had been assuming without noticing. These systems are, by design, tireless fixers. I use them, and I do not think the answer is to refuse them. But a young person who learns to accept the supplied answer as a matter of course may reach adulthood with little practice in the discipline Murdoch thought moral life requires: staying with what is there before deciding what to do about it.

The only counterweight I know of is formation, and formation begins long before anyone can understand what it is for.

## VI.

The ad promised ambitious young people a faster road. Skip the four years, it said, in effect, and start building now. I understand the appeal and owe its authors a concession I was slow to make myself. Aristotle warned that the young are poor students of ethics because ethics works on life experience, and they have not yet had much of it. He was right about me. Had someone handed me Murdoch at sixteen, I would have read her as a theory of perception, filed her away, and gone on fixing. The material her words needed took fifty years to gather, and much of it was my own failure: the problems I solved for people I loved before I had seen who they were.

So the Academy has a point, and the university should hear it. Four years of seminars cannot give a young person the understanding I am describing. Only life can do that.

Yet something did happen at eighteen, and I have come to think it mattered more than anything I understood then. As a freshman, I read Montaigne's *Essays* and wondered how a grown man could fail to understand himself. Self-knowledge seemed to me a problem one solved, like any other, and then moved past. Somewhere in those pages was his essay on the education of children, in which he asks the tutor to let the pupil trot ahead so he can watch his pace. If I noticed it at all, I took it as a note on teaching. I kept the book for five decades before I understood that Montaigne's uncertainty was an achievement and my certainty was the thing that needed examining. When the life finally arrived, the words were there to meet it. The texts wait for the life that can read them.

An education that skips them leaves nothing waiting. The builder trained only to build will reach forty or sixty with the same material I gathered, the same failures, and the same people misunderstood, with no language to see any of it. He may still learn, but he will learn alone and slowly, and perhaps too late for the people he was trying to provide for.

This is why I distrust the promise of speed, and why AI sharpens that distrust. Our tools now compress nearly every kind of work, and the Academy extends that logic to formation itself. Some understanding refuses compression. What an education can do is put the right words in a young person's keeping and teach the discipline Weil described: the patient attention formed even by the geometry problem one fails to solve. That discipline is the capacity to submit to a text before one can use it. It is also, I now think, the capacity to let another person be what they are.

This brings me back to those who provide for the young, whether they are parents, teachers, or academy founders. We cannot give them the understanding. It arrives when it arrives, on a schedule that no curriculum can control. We can give them words they will not need for decades and practice submitting to what they cannot yet use. I kept Montaigne for fifty years without understanding him, and the keeping mattered, because the book was there when a man in his sixties had finally gathered the life to read it. The practice matters as much as the words. A student required to stay with a text that baffles him, or a proof he cannot finish, is rehearsing the posture he will one day need before a loved one he does not, cannot, fully understand.

---