Chapter 1: The Feed

I was born in the city of Boston in the year 1999. No — not 2069, not 2079. I mean 1999. The twenty-sixth of December, in those last days of the twentieth century, when the end of the old millennium was starting to feel real and everyone was half-convinced the computers would crash at midnight and take civilization with them. There was still a cold east wind that blew in off Boston Harbor in those days, a wind so bitterly, bone-achingly cold that no one living in North America's now-balmy climate would be able to imagine it. That wind shaped me. It was the first thing I knew about the world: that it was beautiful and that it would cut right through you.

My name is Juliana West, and these facts will seem strange, I know, given that I appear to be a woman in her late twenties. I'll ask your patience. The explanation will come, and it is not as impossible as it sounds. If you'll grant me that I know when I was born, and that it was the year 2027 when the events I'm about to describe took place, I'll go on.

I was rich. Not the kind of rich that people in my time called rich, not private-jet rich, not island rich, not three-commas rich. I was the other kind, the kind that pretended it wasn't rich at all: startup rich. The distinction mattered enormously to me then. I want you to understand this, because it explains almost everything about the person I was and the blindness I carried like a second skeleton inside me.

My parents were more than comfortable. My father was a pediatrician in Newton; my mother taught art history at Boston College. We lived in a white Colonial on a street of white Colonials. I went to a private school where they taught us leadership and community service in the same breath, so that we understood instinctively that the two were the same thing: that to lead was to serve, and to serve was to lead, and that the people who did both best went to Harvard. I went to Harvard.

From Harvard I went to Stanford, for my MBA. I'd arrived in Palo Alto with the sincere intention of someday running a nonprofit. I had a plan. I was going to work in education: close the opportunity gap, democratize access, all the phrases we used like incantations, as though saying the right words in the right rooms would rearrange the world. I lasted one semester before the entrepreneurship bug bit me, and I want to be honest about how it happened, because I think it says something about the sickness of my time.

What happened was that I met people who were doing things. Not studying things, not analyzing things, not writing grants to maybe, eventually, if the funding came through, do a small thing. They were building companies. They were twenty-four and they had employees. They had users, millions of users, and the users were growing, and the graphs went up and to the right, and when they talked about changing the world, they could point to the graphs as proof. Next to them, my nonprofit plan looked quaint. Small. I didn't want to be small.

Within a year I had pivoted. I was building Lumen, an AI-powered career networking platform that would, as I described it to investors with absolute sincerity, "democratize professional opportunity." We would use machine learning to match people with jobs, mentors, and connections regardless of their pedigree. The algorithms would see talent that human gatekeepers missed. We would be the great equalizer.

That Lumen was, in practice, yet another tool for the credentialed to find each other more efficiently, that it served Stanford MBAs and Ivy League engineers far better than it served anyone else, that the AI we trained reflected and amplified exactly the biases it was supposed to correct, these things I did not allow myself to see. I was too busy. The graphs went up and to the right.

By 2027, Lumen had fifteen million users and was growing fast enough to terrify our largest competitor, a company called Inconnect, which had dominated professional networking for two decades. I was twenty-seven years old. I had been on the cover of Wired. I had given a TED talk. I had an apartment in SoMa with fourteen-foot ceilings and a view of the Bay that made me feel, each morning, like I was surveying a kingdom. And what a kingdom. San Francisco in 2027 was a city of astonishing beauty and astonishing cruelty, often on the same block. The fog would come rolling over Twin Peaks like something out of a dream, the Golden Gate would catch the last light, the Victorian facades would glow pink and amber. And at your feet, on the sidewalk outside your building, a man would be curled on the concrete in a fentanyl nod, barely breathing. In the Mission, I'd walk past street-level windows and see rooms packed with bunk beds, eight or ten to a room, immigrants paying a thousand dollars a month each for a mattress in a city that had decided housing was an investment vehicle rather than a place to live. I stepped around all of it. 'We all' did. I was engaged to marry Edward Bartlett, whose family money was older and quieter than mine, and whose jawline and gentle humor made every room we entered together feel like a photograph. We were, by the standards of our world, a perfect match: same schools, same values, same conviction that the system was fundamentally sound and that the best people rose within it.

This was, you understand, the core belief of my class. I grew up in the particular American delusion that the future would be better than the past because it always had been, at least the slices of the world in which we passed our days, and I was hardly alone. I mean the class of people who had been educated at the right schools, who worked in the right industries, who lived in the right cities and ate at the right restaurants and donated to the right causes. We believed in meritocracy the way our grandparents had believed in God, not because we had examined the evidence, but because the belief was structural, load-bearing, and to question it would have been to feel the floor shift beneath us. I had earned my success. I worked eighty-hour weeks. I had built something from nothing. If the system rewarded me, it was because the system recognized merit, and if it failed to reward others, well, they needed more education, or more grit, or more access to platforms like mine that would connect them to opportunity.

I am aware that this will sound monstrous to you. I ask only that you believe it was entirely ordinary. Everyone I knew believed the same thing.

Let me try to give you some picture of how we lived then, and how the rich and the poor stood in relation to one another, because without this you will not understand the world I lost, or what it meant to lose it.

Imagine a great feed. I use the word deliberately. In my time we called them feeds, the streams of content that our devices delivered to us every waking moment, algorithmically sorted, perpetually refreshed. Every person alive was connected to the feed, and the feed shaped everything: what you saw, what you wanted, what you feared, how you spent your hours and your money and your attention. But here is the thing about the feed that I want you to understand, because it is the key to the whole arrangement:

A very few people owned the feed. And everyone else was fed through it.

If you owned the feed, if you were one of the handful of people who controlled the platforms, the algorithms, the data, you sat above the current. Your wealth accumulated not from any labor of your own, but from the fact that you had positioned yourself at a bottleneck through which all human commerce and connection was forced to pass.

Everyone else — and I mean nearly everyone, billions of people — existed inside the feed. They worked through it: finding jobs on platforms that took a cut of their wages, driving cars dispatched by algorithms that set their routes and their rates, delivering food ordered through apps that charged them fees for the privilege of laboring.

The workers who toiled inside the feed were called many things. "Gig workers." "Independent contractors." "Creators." The language was designed to obscure the relationship, to make exploitation sound like freedom. A woman driving fourteen hours a day for a ride-sharing app was called an "entrepreneur." The platform took its cut — thirty percent here, twenty percent there — and the workers bore all the risk: no insurance, no sick days, no pension, no guarantee of tomorrow's work. But the language of freedom clung to them like a perfume they hadn't chosen to wear.

And here is the part that I find, looking back, most difficult to explain: those of us who rode above the feed — who owned it, or who profited from it, or who simply benefited from the world it had created — we felt sorry for the people inside it. Genuinely. We posted our concern to our own platforms, which our own algorithms then served to other people like us, creating a warm loop of compassion that cost nothing and changed nothing.

We told ourselves the system was regrettable but necessary. That the alternative was worse. That the market was the most efficient allocator of resources — this phrase, "the most efficient allocator," was our catechism, our creed — and that any attempt to redistribute by force would only make everyone poorer.

What we did not tell ourselves, what we could not afford to tell ourselves, was that we were the ones the system was designed to benefit, and that our belief in its necessity was not wisdom but self-interest wearing the mask of reason.

I was, you see, not idle. That was the difference between my privilege and the privilege of earlier ages, and it mattered to me enormously. The idle rich of the nineteenth century had known, on some level, that they contributed nothing. It was possible for a man living on inherited wealth to feel a pang of conscience about it, because the parasitism was naked. But I worked. I worked harder than almost anyone I knew. I built a company. I created jobs. I told myself I was solving a problem. And because I worked so hard, I could not see that the system I worked within was itself the problem, that my eighty-hour weeks were spent, in the final accounting, making the feed a little more efficient at sorting people into winners and losers, and that I was building a tool that, for all its rhetoric of democratization, primarily helped people like me find other people like me.

This is what I mean when I say I was blind. Not stupid — I want to be clear about that, because the distinction matters to the person I was then and to the story I am trying to tell. I was intelligent. I had read the right books. I could discuss inequality at a dinner party with genuine feeling. But my intelligence operated within a frame, and the frame was invisible to me, and the frame said: The system works. You are proof that it works. The people it doesn't work for need what you have — access, education, connection — and you are building the tool that will give it to them. Within that frame, I was a hero. It never occurred to me to question the frame.

Edward's family, the Bartletts, were old Boston. His father had been a partner at a law firm for thirty years; his mother sat on the boards of the MFA and the BSO and a half-dozen other institutions whose acronyms served as a kind of social currency. Edward and I had been together since college. He'd studied history while I studied economics, and then, in a move that charmed me for its impracticality, he'd gotten a master's in urban planning. I'd pulled him out to San Francisco when I left for Stanford, and he'd stayed when I stayed. He worked at a housing nonprofit — actually worked at a nonprofit, the thing I had abandoned — and he was good at it, patient and methodical in a way that I, with my startup's frantic pace, found both admirable and slightly baffling. How could he stand the slowness? The incrementalism? He was working on a project to convert a vacant lot in the Bayview into affordable housing. It had taken three years of permitting meetings. I could have built a software feature in a weekend that would reach more people than his housing project ever would.

Or so I told myself.

Our wedding was set for October 2027. The venue was booked, the caterer selected, the invitations letterpressed on heavy cotton stock, because even in a digital age, some performances of status required the physical world. What delayed us was not, as in some older story, a house under construction, but something more characteristic of our time: I couldn't stop working long enough to plan it.

This was the year that AI, as we called the first instances of intelligent technology back then, had begun to frighten people. Not the people who built it — we were exhilarated, intoxicated, racing to see who could ship the most powerful models, and tools built upon those models, fastest. But everyone else was scared. And they had reason to be. In 2027, for the first time in history, technology could do things that only a few years earlier almost everyone on earth believed technology would never be able to do. It could write legal briefs and diagnose diseases and generate marketing copy and compose music and produce photorealistic images of anything you described. It could code. It could tutor children in any subject. It could hold a conversation so fluent and warm that lonely people — and there were so many lonely people — preferred talking to it over talking to other humans. All of this was brand new, still fantastical, and yet we seemed to grow accustomed to each new unimaginable capability in days or even hours. Every week delivered another shock: another industry disrupted, another category of human labor rendered unnecessary.

The economists debated what this meant. Some predicted catastrophe: mass unemployment, social collapse, the end of the middle class. Others predicted abundance: AI would make everything cheaper, freeing humans to pursue higher callings. The optimists and the pessimists argued with equal conviction, and the rest of the population watched with a growing dread that expressed itself in strange ways: hoarding, conspiracy theories, surges of nostalgia for an imagined past, political movements that promised to turn back the clock.

What nobody debated was whether the gains from AI should be shared. That question, the only question that actually mattered, was somehow not on the table. The assumption, so deep it was invisible, was that AI belonged to the people who started the companies where it was born. That the profits from automating human labor would flow, naturally and rightly, to the companies and investors who had funded the automation. That this was simply how progress worked: someone invents a thing, someone profits from the thing, and eventually, through some mechanism no one could quite specify, everyone benefits.

You, reading this in 2100, will recognize this logic. It is the same logic that every privileged class in history has used to justify its position. The feudal lords said the land was theirs by divine right. The industrialists said the factories were theirs by right of capital. The tech founders said the algorithms were theirs by right of innovation. And the masses — the workers, the users, the data sources, the gig laborers — were told to be patient, to retrain, to adapt, to learn to code, as though the problem was their skills and not the system.

I was, as I said, part of the class that profited. Lumen's AI matched professionals to opportunities. This was useful. It was also, in a way I couldn't see then, a machine for reinforcing the existing hierarchy: for making sure that the people who already had credentials and connections got more of both, while the people who lacked them remained invisible to the algorithm. But the graphs went up. The revenue grew. And I was too busy, too proud, too enclosed in my beautiful frame, to ask who the system was really for.

My particular grievance against the forces of disruption in the spring of 2027 was personal and petty, as such grievances usually are. My head of engineering had been poached by a competitor. Two of my best data scientists had left to start their own company. The entire market for AI talent had become so overheated that salaries had doubled in a year, and I was spending half my time on recruiting instead of building. Meanwhile, my wedding planning was behind schedule, my mother was calling daily with opinions about flowers, and Edward, patient Edward, had begun to look at me with an expression I couldn't quite read: not anger, not disappointment, something gentler and sadder that I didn't have time to investigate.

The broader disruption I treated as background noise. Gig workers were staging protests in several cities, something about being replaced by automated drivers and delivery drones, about algorithms that treated them as placeholders until the robots were ready, about the cameras in the trucks that monitored their every movement while the companies quietly tested the driverless versions. There were rumblings about regulation. The European Union had proposed something, and the American Congress was holding hearings that everyone knew would lead to nothing. On social media, which is to say on the feed, people argued about AI with the same desperate circularity that characterized all public discourse in my time: the same points made and remade, the same outrage performed and forgotten, the same sense that something was very wrong and nobody was in charge.

I noticed all of this the way you notice weather. It was happening. It affected my commute, my hiring pipeline, the general mood at industry events. But it did not touch the core of my life, which was Lumen, and Edward, and the wedding, and the particular, consuming, luminous selfishness of being twenty-seven and successful and in love in a city that, despite everything, still felt like it was mine.

I had no idea, of course, that I was about to lose all of it. Not gradually, not through failure or heartbreak or the ordinary erosions of time, but all at once, in a single evening, through an act of betrayal so complete that even now, writing from the safety of this new world, I feel the ghost of that night's terror move through me like the old east wind.

But I'm getting ahead of myself. Let me stay in 2027 a little longer. Let me stay in the version of the world I understood, the only version I thought was real, before I tell you how I learned otherwise.