On the third day, half the people on the Startup Bus got motion sick. We hadn’t slept for two or three nights, the roads through the Smoky Mountains were perilously curved, the tour bus was traveling at top speed, and we had all been staring at our laptop screens for far too long.

Someone on my team bumped the table where we sat and it collapsed on our laps for the third or maybe 10th time that day. Alicia Hurst, my team’s designer, grabbed her computer before it fell, but her giant water bottle hit the floor. Again. Emma Pinkerton, our business strategist, held up the table while I scrambled to find the bolt that would half-secure it to the wall. Again. I dug around under the tangle of backpacks, purses, computer bags, energy-bar wrappers, extension cords, and tortilla-chip crumbs until I found the bolt.

Order was restored briefly and then I heard Jennifer Shaw, one of the Startup Bus conductors—as they call themselves—grab the mic. “Hey, hey, hey, New York!” she said, for the 100th time. Shaw and Edwin Rogers were leading the New York contingent of the Startup Bus, a harebrained annual competition that is part hackathon, part road trip. I was one of 24 “buspreneurs” (I know) who had signed up to spend three days on the bus pretending to start a tech company. We were on our way to Nashville, Tennessee, where we would meet up with the four other buses coming from San Francisco, Chicago, Mexico City, and Tampa. All of the teams would compete to determine who had built the best technology company while on the bus.

When Shaw greeted us on the mic, we were supposed to cheer. On the mountain roads, however, she got only weak responses. “That was some weak fucking sauce!” Shaw said. She is 36, relentlessly cheerful, with long red hair and a gap between her front teeth. “Let me hear you! Hello, New York!” The responses were a bit louder this time, and she seemed satisfied. She paused. A brief look of confusion crossed her face, as if she’d forgotten what she stepped up to the mic to do. She had only gotten about two hours of sleep the night before. Rogers, her colleague, took the mic.

“We’re going to start pitching again soon,” he promised. “We’ve gone easy on you so far. But the qualifiers are tomorrow, and the judges are not going to be easy on you. They are all entrepreneurs, investors, people who have ridden the bus before. They know how hard it is. You’re going to have to show them that your idea has traction. They want to see users, revenue, a product that is going to make a billion dollars.” He was getting worked up. His version of coaching involved berating us, as if it weren’t already difficult enough being on a crowded bus where the wifi only worked sometimes and the broken electrical system meant sharing three plugs among 50+ devices. I jammed in my orange foam earplugs and turned back to my laptop, where I was working on a pitch deck for my team’s pizza-calculation app—more on what that means in a minute—called Pizzafy. Our website address, purchased on the first day of the trip, was pizzafy.me.

Startup Bus is arguably the looniest of the many hackathons that take place every weekend across the country. A hackathon is a marathon computer-programming competition that, among computer programmers, is slightly less popular than video games, Ultimate Frisbee, or Game of Thrones. It lasts 24 hours to five days, and usually there is a lot of Red Bull and very little sleep.

Startup Bus is part of a special subset, destination hackathons, which require attendees to travel to some remote location for the duration. (A spinoff run by a former buspreneur, Starter Island, requires attendees to spend five days coding on a yacht in the Bahamas.) According to the Startup Bus founder Elias Bizannes, some 1,300 people have ridden on one of the buses so far and been “initiated” into the Startup Bus community. The first bus went from San Francisco to Austin five years ago, landing its entrepreneurs at the South by Southwest festival. This year, the buses were meeting up in Nashville at a tech conference called 36|38. June is a better time for a nationwide road trip than March: Last year, the Kansas City bus got iced in on the highway for 12 hours on its way to Austin.

People who haven’t participated in hackathons talk about them as hotbeds of innovation, as the kinds of places where great thinkers come together and dream up exciting new ideas. Hackers don’t see it that way. They share an open secret: Nothing useful is ever created at a hackathon. There’s even a term for the useless software that people make: vaporware. The idea is that it’s created, and then it evaporates because nobody works on the project after the hackathon (despite everyone’s best intentions).

In reality, a hackathon is a sporting and social event. It’s like a regatta for nerds. Hackathons also serve as outrageously complex recruiting events. Venture capitalists and head-hunters for top tech firms haunt hackathons in order to spot and poach talent. On the surface, however, nobody talks about hackathon software as ephemeral. People pretend that they are really starting businesses, that they are creating software that will have an impact, that they are doing something that has the potential to change lives. The fantasy of creating the next Google is seductive—so seductive that people sign up to spend days with strangers, forgoing sleep, in order to play at being tech entrepreneurs.

The Startup Bus began as a drunken fantasy. Bizannes, its founder and CEO, had moved to San Francisco in 2010 from Australia, where he had worked as an accountant. He was attracted to startup culture, but he was down to his last couple hundred dollars in his bank account, and if he didn’t launch something soon he was going to have to leave California. Drinking with friends one night, he had a brainstorm: What if he launched his own version of Startup Weekend, a popular hackathon? But made everyone ride on a bus? He called Steve Repetti, an investor he knew in Florida, and woke him up. Repetti agreed to invest $5,000 dollars on the condition that he could have a seat on the bus. The project launched a few months later. Bizannes, who went on to work in venture capital at Charles River Ventures, now runs the Startup Bus every year and also runs Startup House, a residential incubator for hackers like the one lampooned on HBO’s Silicon Valley. He’s also infamous for serving as a judge at the 2013 TechCrunch Disrupt hackathon where two participants proposed an app, TitStare, for staring at women’s breasts.

My hackathon team, however, was focused on what I hoped was non-controversial: pizza. I was on the bus to write about the experience. But because I am competitive, I also wanted to win. And I had a plan. It was born of disappointment. At my first hackathon, three years ago, I pitched an idea for software that I really wanted for myself. It was community-garden finder that would let you enter your location and would list every community garden nearby along with contact info and the estimated length of the waitlist for a plot. Nobody wanted to join my team. I learned from that experience that the ideal hackathon project is achievable in the time available, is based on something generically appealing to most people in the room, and has just a hint of whatever is the hot technological topic of the day. Two years ago, data science was really hot. Right now, people are into hardware: autonomous vehicles, fabrication, wearable technology. I suspect that artificial intelligence is the next trend. For this hackathon, I was prepared with a surefire idea. My husband had come up with it as a joke, but the more I thought about it, the more it seemed perfect.

Reuters / Fernando Donasci

When the New York bus left Manhattan at 6:30 a.m. on the first day, an hour and a half after our scheduled departure time, Shaw and Rogers made everyone on the bus stand up and pitch a business idea. Some were better than others. Dre Smith, a software developer, got up and said, “My idea is simple. I want to build a virtual-reality dance party.” People liked that. Another developer proposed an app to help people schedule conference rooms more efficiently. That already exists, I thought to myself. Predictably, there were a couple of ideas for apps that would help Millennials meet each other. (Every hackathon includes an idea for making an app that replicates the experience of online social networking in real life.)

A few people on the bus had some variation of, “It’s like ____ for _____,” as in: “My idea is to build an app that’s like Airbnb for boats,” which was a pitch by a second red-haired woman named Jen. This seemed profoundly unrealistic to build in three days on a bus, but I’ve wanted to learn to sail, so I imagined it might be fun to hang out with buspreneurs who were into boats. I made a note to work on her app if I couldn’t form a team around my own idea.

It was my turn. I stepped up to the mic. “I have an idea for an app that calculates exactly the amount of pizza you need for a party,” I said. People’s heads perked up. “I used to throw a monthly pizza party with a group of friends, and every time we got hung up on how much pizza to order. We called it doing pizza math, and we always messed it up. I want to create an app that calculates the amount of pizza needed for an event based on who is coming, their age, their gender, and their preferred toppings.” People clapped. I was more relieved than I expected. It was possible that this insane plan could work.

Our team’s pizza technology came together thanks to Eddie Zaneski, the other hacker in our group. Hackathons tend to be sausage fests, so it was unusual for our team to be predominantly women. Eddie is 25 years old, 6’ 7” tall, and dresses exclusively in free t-shirts from tech events. “I haven’t bought clothes in years,” he told me as we sat across from each other at the rickety table. “I have a bigger wardrobe than my girlfriend.” Eddie is a developer evangelist at a tech company called SendGrid, which means his job is to go around the country attending hackathons, throwing pizza parties, and handing out t-shirts to developers to convince them to use SendGrid. SendGrid is the technology that many tech companies, including Uber and Airbnb, use to send out auto-generated emails like receipts and marketing messages. Eddie was worried that he had brought too many t-shirts. Our bus had only 28 people. Three giant boxes, which stacked reach four feet high, were in the belly of the bus.

Eddie decided he had more important things to worry about, like getting our pizza-calculator app to work before the hackathon qualifiers in Nashville, so he put on his blue headphones and turned back to his laptop. The apple on its lid glowed through a layer of stickers from other tech events and tech companies: 18F, Iron.io, Penn Apps, Github, and HackRU, Eddie’s favorite hackathon at his alma mater, Rutgers University. I had picked Eddie for my hackathon team because of his stickers. Hackers parse each other’s laptop stickers like fashion mavens parse clothing labels. Eddie’s sticker from 18F, the government open-data team, suggested that like me he was into civic hacking and using technology for social good.

Our app was built using Node.js, a microframework called Express.js, a MongoDB object-relational mapper called Mongoose, and authentication middleware called Passport. We deployed it on Heroku and used Bootstrap for the front end. These are all free software tools that developers use to make other software. Building an Internet app right now is a lot like building a custom Lego house. The building blocks, or bits of code, are all out there on the Internet. The biggest repository is GitHub, a code-sharing site. We decided what we wanted our app to do, grabbed the pre-built pieces of code that would serve as the structural foundation for the “house,” and then started building walls and decorating. Most contemporary software development is a craft, like building houses or furniture. Hackathons are a good way to practice new techniques with other, (slightly) more experienced people in the room. This is another secret in the hacker community. Written instructions and online videos are only useful up to a certain point; to get really good, or to make something really fast, you have to be in the same room with people and you have to talk to them face to face.

In order to get “traction” for our app, we had to get real people to sign up for it and also find some kind of marketplace validation. I called up Domino’s Pizza, which owns a 9-percent slice of the $40 billion pizza market in the United States. Tim McIntyre, a vice president of communication, was kind enough to take my call. I explained the app: group pizza ordering, algorithm to determine toppings, etc. “That sounds like a good idea,” he said, sounding surprised. “Apps like that—there’s a big appetite for them!” Along with 55 years of experience, Domino’s has online ordering, and even a feature where you can tweet a pizza emoji at them and get your favorite pizza delivered to your door. However, they did not have a group pizza-calculating app. I decided this meant we had discovered an underserved market niche, and put McIntyre’s quote into the PowerPoint slide deck I was developing for the final presentation.

By the time the bus arrived at our hotel in Nashville, my team was wrecked. Too much junk food, not enough sleep. But our code worked, and we had a presentation written, and we were ready to see what would happen. On the morning of the qualifiers, all the teams piled onto the grimy New York bus to head to the competition location, Studio 615, a warehouse event space in north Nashville. It was a capitalist high church fantasy scene: a bright white box with high ceilings, makeshift stage, and club music pumping at top volume. We all surged in together, piling our laptops onto the long folding tables covered in black plastic. Some people danced. The space was set up like a fashion show, except it was 9:30 a.m., and in the corner was a spread of cinnamon pecan rolls and sweet tea. Eddie’s t-shirts were piled chin-high on a table, along with free t-shirts from two other tech companies and several boxes of stickers.

The first round took place in the green room, a small retreat off the main warehouse space that was covered in charcoal gray geometric wallpaper. On the wall hung a floor-to-ceiling painting of a naked woman lying in a desert at sunset, holding a can of Redi-Whip. The judges crammed on a couch to watch each team pitch: Bizannes, Repetti, and the two Startup Bus national directors, Ricky Robinett and Cole Worley. The national directors were the only ones getting paid. Everyone else was a volunteer, including the conductors. “I basically gave up two months of my life to organize this,” Shaw had told me during a lunch stop at a Pizza Hut in Punxatawney, Pennsylvania. She was sitting with Mike Caprio, another Startup Bus alum who was riding the bus in order to mentor people in code and business strategy.

Shaw offered to pay for lunch. “Thanks for getting this. I am broke as a joke,” Caprio said. I was surprised: both Shaw and Caprio had been introduced to me as entrepreneurs who had started two companies. Another secret of the tech community: sometimes “entrepreneur” means “runs a successful company” and sometimes it means “more ideas than money.” People inside tech don’t talk about money the same way as people in other industries. Hackathoners chat about tech-company valuations like regular people talk about sports statistics. Instacart is the Startup Bus success story—its founders met through the bus, and eventually started a company together. Instacart is now worth $2 billion, as at least a dozen people told me in Nashville. It’s rumored that some programmers make a living by going from hackathon to hackathon and winning. Personally, I have yet to do anything more than break even on prize money. My fellow buspreneurs and I paid $300 each to ride the bus, plus we were paying for our own food and for five nights of quadrupled-up hotel rooms. Chasing a billion-dollar dream is not cheap.

I had been warned that the judges would ask about monetization, or how the proposed company would make money. The first team, Shar.ed, shuffled into the green room, plugged in a laptop, and prepared to pitch. I had been next to them on a bus for three days, but wasn’t sure what their project was. Shar.ed pitched an idea for crowdsourced on-demand for-profit education, where people could vote for classes they wanted to take and instructors would prepare exactly those classes. They had started an Indiegogo campaign to partially fund the project and had already collected a few hundred dollars. Next up was Screet, a service that proposed to deliver on-demand products to couples in the throes of passion. Aimed at those who want to be safe but don’t want to make a trip to the drugstore, Screet was a smartphone app that would summon a Lyft or Uber driver who would unobtrusively drop off condoms, dental dams, or latex gloves that he or she kept stored in the car trunk in plain, SKU-labeled boxes. This service was going to be especially useful for LGBTQIA people, Screet claimed, because dental dams are hard to find in stores. After two pitches, I wandered out of the green room and rejoined my team to watch the rest of the pitches on the simulcast. Pizzafy was next to last.

Reuters / Steve Marcus

I was nervous. I pitched. We made it to the semifinals! So did Screet, along with a Chicago team that had made a toy that was controlled by an iPad app and helped kids and parents play imaginary dinosaur games together, and some other teams. We ate box lunches. The music played. We pitched again, on the small stage in the main space this time. The pitches were livestreamed. At least a dozen people from other Startup Buses tuned in. SPACES, a New York team that was working on a virtual-reality app, got up on stage and thanked the judges. “We are grateful for the opportunity, but our presentation contains proprietary material, and we are going to decline to pitch,” said John Clinkenbeard, the team CEO. The room erupted. Edwin Rogers started whooping, “New York bus! New York bus! New York bus!” The team, which included Dre Smith of the virtual dance party, had secured $25,000 in funding from an outside investor. The team came down from the stage, went through the crowd shaking hands and accepting congratulatory hugs and enjoying the ruckus. The national directors, standing to the side in their headsets, looked angry. It was a hard act to follow.

Pizzafy advanced again, along with Screet, plus an education company from the Mexico City bus, and a Chicago bus project that sends a text message when you take a pill. Everybody else went out to party in Nashville that night. Emma, Eddie, Alicia, and I went back to the hotel. People from other buses came in from drinking, and sat down to help or chat. People talked about their lives outside the bus. It started to feel like I imagine a barn-raising feels: lots of people from the community showing up and helping to make something that will benefit only a few people, because everybody eventually needs a barn. These hustlers, hackers, and hipsters on the bus would eventually need to hire people or hire companies or get an breathtakingly specific technical question answered in the real world, and they were laying the foundation for this by helping to build our pizza-party app. This is the other secret to hacker culture: Sometimes you do a lot of insane technical work that has no apparent purpose. You do it because it’s a rush. Like marathons.

We worked all night, and all of the next day. We designed an audience-participation stunt, redid our slide deck, honed the pitch until I had every pause and every pizza pun memorized. Finally, in the late afternoon it was time for the final pitch. One of the Instacart guys was a judge for the final round.

The runner-up was PillyPod, “a device that alerts you when loved ones don’t take their meds.”  They had started the week with the URL pillypad.co, but discovered that pillypad.com is a porn site, so they changed the vowel in their name.

The winner: my team!

The judge announced our name, the club lights started going mad, and the DJ blared Katy Perry’s “Dark Horse.” The four of us made our way up to the stage. Shaw hugged me, Rogers hugged me, people I didn’t know hugged me. Rogers cried. I stood on stage with my team and for a few minutes. I felt amazing.

After my week on the Startup Bus, I can tell you what it feels like to win a hackathon. It feels like gorging yourself at a pie-eating contest and discovering that the prize is … more pie. Yes, I’d be happy to sell my new pizza-technology company for a bunch of money. But I’m not holding my breath. The final secret of hacker culture is that Google is a black swan, a lightning strike, an outlier, a Goliath. Most hackers need to keep their day jobs.