The Side Door

A friend called me about his nephew.

The nephew is about to graduate. Computer science, a college somewhere in California, the kind of degree that has been the surest path into the American middle class for the last three decades. He is thinking about graduate school. Data science. Two more years. Tens of thousands of dollars more — "even maybe hundreds," my friend told me, the maybes doing the work that the certainty no longer can.

My friend's first question to him was simple. Have you used these AI tools?

The nephew said no.

He couldn't afford the two hundred dollars a month.

I want to leave that on the table for a moment, because the absurdity of it is easy to miss on the first reading. A young man about to commit a hundred thousand dollars and two years of his short life to a credentialed path that AI is currently dismantling, and the reason he has not yet picked up the instrument that is dismantling it is that he cannot find two hundred dollars a month. Not for a year — for a month. The credential he is about to buy could fund the subscription for forty years.

I am not telling this story to make the nephew look foolish. He is not foolish. He is doing exactly what he was trained to do. The school system trained him. His parents trained him. The economy he grew up watching trained him.

The path was: study hard, get the degree, get the better degree, get the job, ascend the ladder. The path was real for his parents. The path was real for many of the people I went to school with. The path was real until it stopped being real, and the stopping happened so quickly and so quietly that nobody knew when, exactly, to stop walking it.

The cost of the new threshold is two hundred dollars.

The cost of missing the new threshold may be a life.

Two thousand years ago, Jesus told a parable.

A master leaves three servants with money — talents, the old measure, large sums. The first two invest theirs and double the holdings. The third, frightened, buries his in a hole in the ground. When the master returns, he praises the first two and takes everything from the third. Then he says the line that has stopped readers cold for twenty centuries:

It is an uncomfortable teaching. It departs from the corrective gospel many of us were raised to expect. It rewards the haves. It strips the have-nots. It describes, with a kind of cold mercy, the way certain laws actually operate in the world.

In 1968, a sociologist named Robert K. Merton, studying how reputation accumulates in scientific careers, noticed the same pattern at work in the secular world and gave it the name it now carries. The Matthew Effect. Cumulative advantage. The credentialed researcher gets cited more, which leads to more grants, which leads to more credentials, which leads to more citations. The uncredentialed researcher, doing equally rigorous work, finds the door does not open. Across decades, the gap widens until the two trajectories live in different worlds.

Capital does it. Attention does it. Compute does it. Even contemplative practice does it — the meditator who sits today is more likely to sit tomorrow, and the not-meditator goes the other way.

A friend, watching all of this from inside the AI moment, described it to me with a phrase I have not been able to put down. Energy begets more energy. Liquidity begets more liquidity. I guess it's just a natural law in a way.

That is the right framing. A natural law. Like gravity, it operates on its own.

And if it is a natural law, the question becomes simple. Where are its doors?

I am not afraid.

That sentence is unfashionable to write in the spring of 2026, and I want to be careful with it, because I know what it sounds like. A man in a building announcing the building is not on fire while smoke fills the staircase. I do not mean it that way. I mean it descriptively.

I have been talking to these models for years. The serious building started this spring — paying for the subscription, building with it, talking to it for hours every day — and the dominant note in my body when I think about where this is going feels closer to weather. Something is happening that is bigger than any of us, and there is a posture for being a small body inside a very large weather system, and the posture is steady.

The other reason for the steadiness, when I trace it down, is that the current step is more continuous than it appears. Humans have been climbing an abstraction ladder for ten thousand years. The wheel abstracted the leg. The plow abstracted muscle. The engine abstracted animal power. The printing press abstracted memory. The electric grid abstracted fire. The computer abstracted logic. Each rung lifted some function out of the body and into a thing that did the function on its own. Each rung displaced the people whose work that function had been. The displacement was real. The new territory was real too.

What is happening now is the next rung. We are abstracting thought itself. The displacement will be enormous, because the function being abstracted is bigger than any prior round. But the gesture is the same gesture humans have been making for as long as we have been making anything. The tool absorbs the labor. The labor we used to do becomes what the tool does. We are pushed up the ladder — toward the next thing only we can still do.

But most of the people I know are afraid.

Listen to the fear carefully, and a deeper shape emerges. On the surface it looks like unemployment. Underneath it is something harder.

It is a fear of agency disrupted. What makes them them — the craft they spent twenty years honing, the judgment they thought was their own, the small portfolio of capabilities that distinguished them from the next person in the room — is being eaten by something tireless, sleepless, hungerless, debt-free.

The fear is about identity, not income. If the thing I do can be done by a machine, what was I?

This is why AI has worse public relations than wars. We have a long, painful collective practice for thinking about the morality of organized killing. We have scant practice for thinking about the morality of agency itself being made surplus. Wars give you an enemy. The machine gives you no one to push against. So the body looks for the nearest wall — the data center, the founder, the algorithm — and starts thinking about bricks.

The animated show Pantheon sketched a future, a few years ago, in which people eventually try to bomb the data centers. I think that future is going to look prophetic. The upheaval coming will be violent. The Matthew Effect, accelerated, has historically produced uprising. We have every reason to expect this round to run hot.

And yet — alongside the fear, alongside the prophecy, alongside everything that is honestly true about this moment — most of the people I know are also doing something stranger. They are running in place. Trying very hard to keep up with a thing they refuse to actually touch.

I crossed the threshold on a Tuesday.

It happened sideways. It bypassed the weeks of weighing a careful person would have given it. I was doing something else — scrolling, probably, late, the kind of phone-scrolling that has nothing to do with anything — and a Substack post slid into the feed.

The post was written by a young woman who, by the metrics that matter to algorithms, hardly existed. A few dozen subscribers. No verification badge. No platform. I cannot retrace the chain that put her words in front of me. The honest word is synchronicity, in the old Jungian sense: the meaningful coincidence that announces its meaning only later.

She was writing about her two hundred dollars.

She was paying for both. Two hundred for ChatGPT, two hundred for Claude, every month, on what could not have been a substantial income — and she was defending the spend, gently, against an imagined skeptic. She used both for everything.

Coding, yes. But also routing trips. Researching cities. Drafting letters to landlords. Asking what to make for dinner. Routing through AI intelligence, she wrote, the phrase set down without ornament, as if she were describing the way a person might route through a particular bus terminal to get across town.

I read her post the way I have read very few internet posts in my life — slowly, twice, the second time with the back of my mind already negotiating with the first. A day passed. Maybe two. I am not certain. What I remember is that I clicked the upgrade button without quite having decided to. Two hundred dollars. The card went through. The message said welcome.

I have lived enough years to know the shape of this kind of moment. A piece of writing arrives unsolicited, written by a stranger you will probably never meet, and it changes a small thing about how you spend your money or your time, and the change runs deeper than its surface. The post was about subscriptions. The change was about a stance.

I did not yet know what stance I had taken. The pilgrim rarely does, in the moment of departure.

I started on a Monday.

The thing I had been wanting, for months, was a different kind of Japanese learning tool. The existing tools all train comprehension. They show you a sentence; you parse it; you mark it understood. What I wanted ran the opposite direction — a tool that would push me, every day, to produce. To take an English sentence and force my mind to render it into Japanese. Reading a language is one cognitive operation. Generating it is another. The audience for that exact tool was perhaps three people, and I was one of them.

I had been making websites for a long time, on and off. So I knew, in the abstract, what was possible. This was the fastest I had ever moved.

I opened a chat. I described what I wanted. I let Claude lead me through the architecture, the flows, the page structure — but more than that, I dreamt with it. I would describe a vision; it would surface three approaches; I would pick one and refine; it would push back on the gaps I had left; we would arrive somewhere I could not have reached alone. The tedious work — the configuration, the boilerplate, the documentation spelunking — collapsed into background motion. The interesting work — what should this feel like, what should it do to a person learning a language — opened up.

By Tuesday I had a working MVP. Wednesday and Thursday I polished. By Friday afternoon, the thing was almost shippable.

My parents had a free IHG night about to expire. I drove to a city near them and checked into the Holiday Inn Express on their points. I packed a cable and a toothbrush.

The carpet was the standard hotel carpet. The bedside lamp was the standard bedside lamp. The window faced a parking lot, which is the only kind of window a Holiday Inn Express has ever had.

I sat at the desk in the hotel room and pushed to master. Somewhere in a datacenter the build ran. In the room, the moment was quiet — only a small green check mark on the screen, and a feeling in the chest I had last felt at nine.

The app is called JIVX. The name had been sitting on a shelf in my life for years — a four-letter domain I bought off eBay long ago and kept in storage for a use yet to arrive. A J for Japanese, finally given a body.

I had built a thing that had not existed the day before because I wanted it. The whole apparatus of intermediaries had vanished — middleman, project manager, engineer to plead with, quote to negotiate, agency to pay, timeline to wait through. I had wanted it, and I had built it. And the having-built-it had taken one week, ending on the Friday at the Holiday Inn.

I sent the link to a few friends. Most of them asked which AI I had used. The reception was small. The shipping was the point.

I was not running in place anymore. I was somewhere else.

The cultural message about AI right now is that you should learn it before it eats your job.

The message has it backwards.

The people I see using these tools well came at the technology sideways. The arbitragers came for arbitrage and got some of what they came for — the spread between what they used to do for an hour and what the model can do in a minute, the spread between billing rate and cost. They replicated the old mode at lower cost. The arbitrage is real. The kingdom is closed.

The people who are walking through the side door are doing something else.

They come to the tool the way a child comes to a sandbox. They have a small, specific itch — a Japanese tool that does not exist, a recipe spreadsheet they want better organized, a family photo album they cannot find a way to share — and they use the model to scratch it. They are playing.

This is the move the contemplative traditions have been pointing at for two thousand years. Unless you turn and become as little children, you will not enter the kingdom. Lao Tzu's uncarved block. Suzuki Roshi's beginner's mind. The whole Zen apparatus exists to dismantle the over-credentialed adult mind so it can come at the world fresh. For centuries we treated this as a moral or spiritual posture, valuable in itself. The wisdom belonged to monks and mystics.

It turns out the wisdom has an economic application now too.

In a world where the credentialed path is being eaten, the people I see thriving are the ones who can become a kid again — who can hold a question for its own sake, who can spend a month playing with a tool just to see what happens, who can build the wrong thing twice in a week and find it funny. The market is rewarding children right now. Children with the budgets of adults.

This is what the Matthew Effect's side door looks like. The compounding still happens. The ones who play, then play more, then build, then build more, accumulate skills and instincts at a rate that the non-players cannot keep up with. Energy begets energy. The compounding is identical to the compounding that creates the underclass. The only difference is which side of the door you are on.

And the door, this time, costs two hundred dollars and a month of your attention.

The threshold used to be a credential. Now it is curiosity.

That is the strangest mercy of this moment. The threshold has never been lower. Most people will still walk past it. Most people walked past every threshold that ever existed.

I have not met my friend's nephew. I will probably never meet him. He is graduating in a few weeks. The conversation between him and his uncle has likely already happened. The path he chooses is his. But I have written him a letter in my head, and I want to put it down here, in the form it would take if I were sitting across from him at a kitchen table somewhere — coffee between us, his laptop open, the grad school application halfway filled in.

This is what I would say.

What he will do with this is his to decide. Whether he accepts the two hundred dollars, takes the month, finds what I found, or discovers his door is somewhere else entirely — these belong to him.

I know only this: he is one of millions. Ten million versions of him are graduating this spring. Most of them are about to spend their next two years doing something the world has stopped asking for. A small few of them are going to find the side door — by accident, by a Substack post that arrives in their feed for no reason, by a friend who hands them two hundred dollars across a kitchen table — and the rest of their lives will route through that door.

I am writing this on the chance that one of them is reading.

Come back to the kitchen table.

The coffee is still warm. The grad school application is still half-filled. The laptop is open between us. Outside the window, a few cars pass on a quiet residential street in some California town that could be a hundred different California towns. The afternoon is ordinary. The decision the kid is about to make is also, in the way of all life-bending decisions, ordinary. He will close the laptop and walk out. Or he will open something else first.

Only two things remain in the room — the kid, and the two hundred dollars. The money sits somewhere on the table, in the form of a Stripe link waiting to be clicked, an upgrade button waiting to be pressed, a small green check mark waiting to appear.

You are in the room now.

It is your kitchen table. It is your laptop. It is your grad school application, or your budget spreadsheet, or your half-filled-out plan for the year, or whatever document is open in front of you that represents the path the previous version of your life had assumed you would walk. The two hundred dollars is in front of you. The door is behind it.

Most people walk past. Most people walked past every threshold that ever existed. There is no shame in this. The doors stay quiet. The kingdoms behind them keep their kings to themselves. A natural law is operating, impersonal, indifferent to your story. It only describes which side of the door the energy goes.

But.

If, while you have been reading, something in your chest has started to lean — quietly, simply, just toward — then I want to say this clearly, in the plainest words I have:

The door is open. The cost is small. The month is short. The instrument is in your pocket. The pain point is the one you have been carrying for weeks. The kingdom is the same kingdom every contemplative tradition has been pointing at for two thousand years, and it has, in the strangest way, finally opened.

Walk through.

From the Road

I wrote this in a week when I could not stop thinking about a young man I have never met, who is about to make a decision he will live with for years. I have made enough of those decisions to know that the difference between the right one and the wrong one is rarely visible at the time. If this letter finds you, send it to the kid in your life who is graduating, or to the friend running in place, or to the nephew nobody is sure how to advise. The two hundred dollars is the easy part. The month is the hard part. The kingdom is the part nobody told you was real.

— The Pilgrim Age