When Y Combinator flies people to San Francisco for Startup School, it gives each person $500 USD in flight credit but does not cover accommodation. A lot of people I knew were able to stay with friends or family in the city, or sleep on the floor of someone’s hacker house. I ended up booking a three-bed hostel room with two other girls I met through a WhatsApp group. It was the cheapest place we could find for a single weekend within reasonable distance of the venue, and we were all happy with the amount we saved.

I did not know this until we arrived, but our hostel was in the one neighbourhood the Startup School website specifically advised us not to stay in: the Tenderloin.

Later, in the back of an Uber after the first day, one of my roommates—who had flown from Australia just for the event—asked Siri how the neighbourhood got its name. Tenderloin feels unnatural to say aloud when referring to a section of a city, weirdly intimate in the mouth, bringing to mind butchers and the smell of raw meat. Siri spat out a story about police officers making so much money in bribes while patrolling the area that they could afford to eat more expensive cuts of meat, such as tenderloin.

The hostel was pretty much exactly what we paid for. The toilets were shared by the floor, while the room had its own sink, so you could brush your teeth in the dark while your roommates slept and hope they had remembered earplugs. There was barely enough room between the three beds to navigate around our shoes and suitcases. One window faced a brick wall. The other overlooked a narrow alleyway where, if you looked down, you could see people sleeping on cardboard and newspaper, or standing on two feet but bent at the waist, unmoving.

The Tenderloin has something of a reputation in San Francisco, as evidenced by the widened eyes and aghast expressions of everyone we met whenever our accommodation came up. I cannot say the reputation is undeserved. Discarded needles littered parts of the sidewalk, and sirens sounded throughout the day and night. People who appeared to be passing through speed-walked with their heads held up, careful not to meet the eyes of anyone sitting or lying along the sidewalk.

But even on those streets—or rather, above them—the strange, tech-centred heart of San Francisco managed to creep in.

The advertisements were the first thing that made the city feel unreal. In a normal airport, you might expect to see ads for airlines, hotels, soft drinks, credit cards, or mobile data. At SFO, almost every ad seemed to be for a technology company.

They were usually minimalist, with some vague tagline and a proper noun underneath that you had to assume was the company’s name. The ad would offer almost no indication of what the company actually did. The crypticness was presumably meant to draw you in and direct you to a website, where you would probably find an About page featuring two plucky Stanford dropouts describing themselves as “cracked builders” and “lifelong learners.”

At least, that was the idea for people in tech who were curious enough to follow the URL. To most people passing through the airport, the ads must have been completely meaningless. They sprawled out of the airport and into the city.

“Good agents leave traces.”

“Open Source AI Engineering Platform.”

“Go on call ship.”

If you raised your eyes above the sidewalks, there were larger and more expensive billboards climbing toward the sky. A small plane circled overhead with a banner streaming behind it:

LIFT SHIP REPEAT — TERRA API.

The version of San Francisco I encountered also seemed to speak in some kind of code language that you had to adopt immediately to be recognized as part of the in-group. At an event like Startup School, with six thousand people in tech—or people who desperately wanted to be in tech—filling a stadium, I was inundated with it.

People did not greet you by asking, “What’s your name?”

They asked, “What are you building?”

My default answer was “Nothing at the moment,” which was usually fine, because the question was often less a request for information than an opening through which the other person could explain what they were building. Every conversation seemed to have some purpose: networking, recruiting, pitching, or advertising by word of mouth. The same words appeared everywhere. Cracked. Maxxing. Growth. Slop. Ship. Agent. Larp. Ordinary words compressed into a dialect of constant production.

On the first morning, my roommates and I arrived at 7:30 to get through the doors before too much of a line formed. We placed our bags on a conveyor belt, walked through metal detectors, and were given lanyards and tote bags full of sponsor merchandise, including Touchland hand sanitizer, a YC-branded power bank, and a pack of “Silicon Valley trading cards.”

Instead of Pokémon or athletes, the cards featured technology executives, each ranked by rarity.

Mati Staniszewski: Rare.

Steve Jobs: Ultra Rare.

A couple of high school students ran over while I was eating breakfast and asked which cards I had pulled, looking for anything worth trading. I gave away my Steve Jobs in exchange for an Alexandr Wang, classified as Core.

I made a half-hearted attempt at networking, but somehow almost everyone I met was younger than me. Many were rising high-school juniors or seniors. A handful had just graduated and would proudly tell me which university they were attending in the fall. I’m sure they were good schools, but I only know a limited number of American postsecondary institutions, so the distinction between one Texas university and another was not particularly meaningful to me.

Most of these kids did not have much concrete experience in technology yet, but their eagerness was completely sincere. They pitched fledgling startup ideas to one another, exchanged LinkedIn profiles, and pored over little plastic cards printed with the faces of CEOs.

Eventually, we filed into the main stadium for a series of opening talks. Music boomed through the venue while people continued attempting to network over it. Across the crowd, attendees held laptops and iPads above their heads displaying enormous QR codes. Scanning one might take you to a LinkedIn profile, a personal website, or the landing page for a company that did not fully exist yet.

I sat two seats away from a boy who rarely went five seconds without talking about himself. I asked where he went to college, and he said he did not believe in college, so he had never attended. What he was doing instead remained less clear, although he certainly managed to talk about it. Then Jensen Huang took the stage. Near the end of his thirty-minute slot, he advised the audience to “stay in school.” The entire stadium erupted in applause. The boy applauded too.

After lunch, the stadium was split into two simultaneous sessions. In the style of a silent disco, we were given headphones and could listen to either talk depending on which channel we selected. I remember little of the first. The second, however, I found genuinely invigorating.

The speaker was Blake Scholl, the founder and CEO of Boom Supersonic, a company working to bring back commercial supersonic flight. Supersonic passenger flight over the United States has been restricted for more than fifty years because of the sonic boom. Boom’s experimental aircraft had recently broken the sound barrier without producing a boom audible on the ground, using an atmospheric effect known as Mach cutoff. Scholl presented the demonstration as a step toward overturning the blanket speed limit and regulating supersonic flight by how much noise actually reached the ground.

He spoke about humanity’s urge to go faster, and about challenging limits that everyone else had accepted as permanent. About pursuing good ideas even when the law had not caught up, even when the engineering seemed impossible, and especially when other people called you crazy.

I thought, finally, something cool. Something that actually lived up to the hype.

All day, people had talked about building and innovation, but the words often seemed detached from anything tangible. How did the hundredth company working on B2B SaaS make life any better for the average person? Supersonic flight could. My roommate had flown all the way from Australia for a two-day event. What would it mean if she could make that journey in eight hours instead of fourteen? What if the world could feel that much smaller? For once, “building the future” seemed to mean building an actual machine that could move through the air.

I did not know whether it would be economically feasible, but for the first time that day, I understood the excitement. It almost seemed too good to be true. Back at the hostel that night, I searched for more information about Boom. Something this transformative, I thought, should have been much bigger news.

The details were more complicated than the presentation had made them sound. Under the existing rules, Overture would cruise at Mach 0.94 over land—only around twenty percent faster than an ordinary passenger jet—and reach Mach 1.7 only over water. Faster overland travel depended on regulations that had not yet changed.

The aircraft would still save several hours between Australia and California. But it would carry fewer than eighty people in an all-premium cabin, with fares Boom compared to business class. What had briefly sounded like liberation from distance began to look more like another way for the people who moved most easily through the world to move through it faster.

After the first day ended, afterparties erupted across the city, hosted by various companies. My roommates and I had failed to make it off most of the RSVP waitlists, so we spent the evening travelling back and forth on delayed trains, seeing which parties we could enter without registrations and trying to get some free dinner. I am not much of a partier, but these parties were strange even by my standards. They were mostly held in brightly lit offices, where people stood stiffly in small circles and spoke in the same alien vocabulary they had been using all day.

At one party, a DJ played to an empty dance floor. I went into the kitchen for a glass of water and overheard two teenagers speaking nearby. One said he would be attending the University of Waterloo in the fall. The other immediately asked for his LinkedIn and began gushing about how “cracked” he must be, explaining that Waterloo was like the “Georgia Tech of Canada.”

I hung around long enough to hear the incoming student begin pitching his startup. It involved putting a group of AI agents in a room and feeding them context. I could not determine much more than that, but the other boy nodded with great seriousness.

The second day began later, presumably to accommodate the afterparties that had stretched into the previous night. My roommates and I took a little too long getting coffee and arrived to find not so much a line as a crowd wrapped around the block. We missed the first talk, but made it through security shortly before Alexandr Wang came onstage.

The boy sitting to my left was still in high school. We exchanged the standard introductions and eventually asked what he was building. He answered for several minutes without ever quite saying. He was meeting great people. It was the best time in history to start a company. The opportunities in AI were unprecedented. His answer closely resembled what Garry Tan had told the entire stadium the day before.

Throughout the talk, the boy kept his laptop open. He moved between editing his résumé and filling out an application for the next YC batch. I tried not to read over his shoulder and began noticing the other screens around the stadium instead. Even with some of the most powerful figures in technology directly in front of them, a large number of people were prompting Claude, scrolling through LinkedIn, answering messages, or updating websites.

When I was in high school, I did not have LinkedIn. I had never heard of Y Combinator. At breakfast that morning, I had met a seventeen-year-old from San Francisco who had booked himself into our Tenderloin hostel solely to be closer to Startup School. Around him were thousands of other young people who already knew the routine: always shipping, always networking, always “grinding.”

The last speaker was the one everyone had been waiting for: Sam Altman. By then, we had been sitting through talks for most of the day, but the crowd was suddenly attentive again. He spent most of his slot talking about AI and where he thought it was headed. Near the end, he turned to one of the annoyances of running YC: the people on Twitter who mocked it.

Running a startup, he said, was difficult. Creating something of actual value in the world was difficult. Making a sarcastic comment online, collecting ten thousand likes, and feeling as though you had accomplished something was easy. But it would, he warned, “poison your soul.”

Taking shots at people attempting difficult things was not merely annoying. It was “morally bankrupt.” A kid might post excitedly about a bad startup idea, and someone else could humiliate him with a clever reply. Over the years, Altman said, many trolls had probably felt as though they had gotten the better of him.

“Over the decade,” he concluded, “none of them did.”

Beside me, the boy continued filling out his YC application.

Altman looked out across the stadium.

“I don’t doubt,” he said, “there are future trillionaires sitting among you.”

Everyone stood and applauded as he walked offstage. Music blasted through the stadium, and a cameraman walked backward in front of him so we could watch his exit on the screens. A group of boys in the stands above the doorway were practically falling out of their seats trying to wave him down. Altman looked up, and just before he went through the door, one of them dropped something straight into his hands.

I left the second night’s parties early to catch my flight. On the BART back to the airport, San Francisco looked prettier to me, the skyline glowing in the distance while the shadows concealed some of what daylight had made difficult to avoid.

The city darkened, but the advertisements did not. At street level, backlit double panels interrupted the sidewalk, unattached to a shelter or kiosk.

“The last line of defense between your agent and production,” one read. “San Francisco has lost a lot of its soul,” read the other.

Between them, in their combined glow, a man lay beneath a blanket. Only his bare feet were visible.