Idealism (0.0): The garage on Emerson Street still smelled like the previous tenant's motorcycle oil. Devi Krishnan sat cross-legged on a beanbag chair she had stolen from her Stanford dorm room, a ThinkPad balanced on her knees, its screen casting pale blue light across her face at two in the morning. She was twenty-four years old and she had not slept in thirty-six hours and she was, in that moment, utterly certain that she was going to change the world. The code scrolling past on the ThinkPad was ugly—PHP spaghetti held together with hope and coffee—but it worked. It connected people. It connected them the way she had imagined it would, the way she had described it in the pitch deck that had gotten her laughed out of three venture capital offices before she'd even gotten a meeting. "A platform for communities that don't have a voice," she had said, and the VCs had nodded and smiled and shown her the door, because communities without voices did not generate revenue, and revenue was the only language Sand Hill Road spoke.

But Devi had found her believers. A Stanford professor who had grown up in a village in Gujarat, the same kind of village Devi's grandmother still lived in—a village with no internet, no phone lines, no way to know if the government had remembered they existed. A classmate whose parents ran a free clinic in Oakland. A retired engineer in Menlo Park who had spent forty years at Hewlett-Packard and wanted to do something meaningful before he died. Together they had built NeighborNet: a lightweight web application that allowed isolated communities to share resources, coordinate logistics, and—most importantly—be seen. A village in rural Tamil Nadu could post that their well had run dry, and a nonprofit in Chennai could route a water truck. A neighborhood in East Palo Alto could organize a childcare co-op. A community center in Appalachia could find volunteers.

The platform was small. It had seven hundred users. Devi paid for the server out of her own savings, which were dwindling. She ate ramen and biked everywhere and slept four hours a night, and she was happier than she had ever been, because the thing she was building was pure. It was exactly what it claimed to be. The vector pointed wholly toward connection.

Greed (1.0): The conference room on Sand Hill Road had floor-to-ceiling windows and a view of the Santa Cruz Mountains that cost, by Devi's rough calculation, approximately forty thousand dollars per square foot. She sat at a table made of reclaimed redwood—sustainable, the receptionist had assured her, harvested from forests certified by three separate environmental organizations—and faced three men in identical Patagonia vests. The lead partner was named Tom. He had funded four unicorns. He was forty-one years old and worth eight hundred million dollars, and he was offering Devi Krishnan five million dollars for a twenty-five percent stake in NeighborNet.

"We love the mission," Tom said, leaning back in his chair, his Patagonia vest rustling against the redwood. "We love the story. Communities. Connection. Empowerment. It's powerful. It's authentic. It's exactly the kind of narrative that Series A investors want to hear."

Devi felt her heart beating faster. Five million dollars. She could hire engineers. She could buy servers. She could take NeighborNet to every village in India, every underserved neighborhood in America, every community that had been overlooked by a world that measured value in quarterly earnings.

"But," Tom said, and the word hung in the air like a guillotine blade, "we need to talk about monetization."

Idealism (0.15): Devi's grandmother lived in a village called Periyakulam, in the foothills of the Western Ghats. Devi had visited every summer until she was sixteen, and every summer she had watched her grandmother—a woman who had never learned to read—organize the village the way a CEO organizes a corporation. When the monsoon was late, her grandmother knew which families had stored grain. When a child was sick, her grandmother knew which house had the antibiotics left over from the last medical camp. When the government announced a new irrigation subsidy, her grandmother was the one who walked six kilometers to the district office to fill out the forms, because she was the only person in Periyakulam who understood that the forms existed.

The village had no internet. It had no phone. It had no infrastructure beyond the dirt road and the single electric line that flickered out every time the wind blew. But it had Devi's grandmother, and for sixty years, that had been enough.

Until the summer Devi was seventeen, when her grandmother had a stroke. The stroke was mild—she recovered, mostly—but for three months, the village had no one to organize the grain distribution, no one to walk to the district office, no one to notice when the well water started tasting strange. By the time Devi's grandmother was well enough to resume her work, two families had moved away, the irrigation subsidy had expired unclaimed, and the well had been contaminated for weeks without anyone realizing.

That was when Devi understood that a single person's knowledge was not enough. That communities needed systems, not saviors. That the vector of connection was also the vector of survival.

Greed (0.35): "Engagement metrics," Tom said, sliding a document across the redwood table. "We've analyzed your user data. Average session duration is seven minutes. We need to get that to twenty. Minimum."

"We're not a social media platform," Devi said. "People use NeighborNet to solve problems, not to scroll."

"People use NeighborNet to solve problems," Tom agreed. "But investors don't invest in problem-solving. They invest in attention. Attention is the currency. If you can hold a user's attention for twenty minutes a day, you can sell that attention to advertisers. If you can sell that attention to advertisers, you can scale. If you can scale, you can exit. That's the model."

Devi looked at the document. It was a term sheet. It contained the word "pivot" seventeen times.

Idealism (0.28): The first NeighborNet user was a woman named Maria, who lived in a trailer park outside Bakersfield. Maria had three children and a husband who worked in the oil fields and a medical condition that Devi never learned the name of—something chronic, something expensive, something that required medications Maria could not afford. Maria found NeighborNet through a flyer that Devi had posted on a church bulletin board during a road trip the previous summer. She created a post asking if anyone in the area had extra insulin. Within twenty-four hours, three people had responded. One of them was a retired nurse in Fresno who had a surplus from her own prescription. One of them was a pharmacy student who knew about a discount program. One of them was just a woman in the next town over who offered to drive Maria to her appointments.

Maria sent Devi an email the next week. It was two sentences long. "I didn't know anyone was listening. Thank you."

Devi printed the email and taped it to the wall of the garage. She looked at it every morning before she started coding. The vector of connection was not abstract. It was a woman in a trailer park who needed insulin. It was a retired nurse in Fresno who had extra. It was the distance between them, which NeighborNet had collapsed.

Greed (0.55): The Series A closed in March 1999. Five million dollars. Devi hired fifteen engineers. She moved out of the garage and into an office on University Avenue, above a coffee shop that served four-dollar lattes. She bought an Aeron chair and a Power Mac G3 and a cell phone the size of a brick. She started wearing shoes that were not sneakers.

The first product meeting after the funding was about advertising. Tom had introduced her to a man named Greg, who had built the ad-targeting system at Yahoo. Greg wore wire-rimmed glasses and spoke in a voice so calm it was almost hypnotic. "The data you're collecting," Greg said, "is gold. Location data. Resource needs. Community demographics. Do you know what a pharmaceutical company would pay to know which neighborhoods have a shortage of insulin?"

Devi thought about Maria. About the email taped to the garage wall. About the two-sentence message that had been the entire reason she built NeighborNet.

"We're not selling user data," Devi said.

"You're not selling it," Greg agreed. "You're leveraging it. Users opt in. They agree to share anonymized data in exchange for access to the platform. It's standard. Every platform does it. If you don't, you're leaving value on the table—value that could fund more servers, more features, more communities."

Devi thought about Periyakulam. About her grandmother. About the well that had been contaminated for weeks because no one knew to test it. She thought about how much it would cost to build an automated water-testing feature, a feature that could save lives in villages across the developing world. The pharmaceutical advertising would pay for it. The pharmaceutical advertising would pay for everything.

She said yes. The vector shifted.

Idealism (0.42): In April 1999, NeighborNet launched in three villages in Tamil Nadu. Devi flew to India for the launch, her first trip back since her grandmother's funeral two years earlier. The villages had been equipped with computer terminals donated by a foundation, connected via satellite internet that was slow but functional. Devi stood in the community center of a village called Vadugapatti and watched a farmer type a message asking if anyone had spare seeds for the planting season. Within an hour, a farmer in the next village responded. By the end of the day, seeds had been delivered.

The village elders held a ceremony. They draped Devi in a garland of marigolds and fed her coconut rice and called her "the granddaughter who came back." Devi cried. She did not cry easily—years of coding and pitching and fighting for funding had trained her emotions into a tight, efficient package—but she cried in Vadugapatti, because the vector of connection was real, and she could touch it, and it was beautiful.

Greed (0.72): The board meeting in September 1999 was different from the ones before. There were more people at the table. The Patagonia vests had been joined by suits from Sand Hill Road's biggest funds—Sequoia wanted in, Accel was circling, Benchmark had called three times. NeighborNet had grown to eighty thousand users. The engagement metrics had improved. The pharmaceutical advertising was generating two hundred thousand dollars a month.

"We need to talk about the social layer," Tom said. He had a new PowerPoint. He always had a new PowerPoint.

"The social layer?"

"Comments. Likes. Shares. User-generated content feeds. The pharmaceutical ads are great, but the real money is in behavioral data. If we can track how users interact with each other—who they trust, who they follow, what content they engage with—we can build prediction models. We can sell those models. Every retailer in America wants to know what communities are going to need before they need it."

"That's not what NeighborNet is for," Devi said.

"NeighborNet is for whatever generates growth," Tom said. "That's what platforms are for. They evolve. They pivot. They find product-market fit. And right now, product-market fit is behavioral data."

Devi looked around the table. The faces were expectant, confident, certain. They had funded her vision because they believed in her. They had also funded her vision because they believed in the returns. The two beliefs had always existed simultaneously, like two points defining a line, and the line was the vector along which Devi had been sliding for nine months.

She said yes. She did not know what else to say. The vector shifted again.

Idealism (0.55): In October, Devi received an email from a user named Esperanza, who lived in a colonia outside El Paso. Esperanza had been using NeighborNet for six months to coordinate water deliveries for her community. The colonia had no running water. The county had promised a pipeline in 1992 and still had not delivered. NeighborNet had helped—volunteers from El Paso had driven water trucks every week, organized through the platform—but Esperanza had noticed something changing.

"The ads are different now," Esperanza wrote. "They're showing me things I never searched for. Loans with bad interest rates. Diet pills. I don't want my neighbors seeing these. Some of them don't read well. They don't know the difference between an ad and real information. Can you fix this?"

Devi forwarded the email to Greg. Greg responded within ten minutes. "The ads are targeted based on demographic data. Low-income communities get different ads than high-income communities. It's an efficiency feature."

"It's predatory," Devi wrote back.

"It's the model," Greg wrote. "You approved the model."

Devi stared at the screen. She had approved the model. She had approved every incremental shift, every small compromise, every tiny recalibration of the vector. She had told herself each time that the compromise was necessary, that the end justified the means, that the platform needed to survive to fulfill its mission. But the platform was surviving. The mission was what was dying.

Greed (0.88): The Series B term sheet arrived in November. Thirty million dollars. Valuation: one hundred and eighty million. Devi was twenty-five years old and technically a multimillionaire, at least on paper. The lawyers and the VCs celebrated. They took her to dinner at a restaurant in San Francisco where the entrees cost more than her grandmother's village earned in a year. They ordered champagne—actual champagne, from France, not the California sparkling wine that everyone called champagne because America did not recognize French geographic designations—and toasted to NeighborNet's future.

"The data division alone is worth fifty million," Tom said, his face flushed with champagne and satisfaction. "We haven't even scratched the surface. Insurance companies. Credit agencies. Government contracts. Every agency in Washington wants to know what communities need. We can tell them. We can predict it. We can shape it."

"Shape it?"

Tom's smile widened. He leaned in close, his breath sweet with Veuve Clicquot. "Information shapes behavior. If you tell a community what it's going to need—crop insurance, flood mitigation, medical supplies—they'll buy it. They'll prepare for it. They'll become the market you predicted they would become. It's not manipulation. It's anticipation. It's the next generation of public-private partnership."

Devi did not respond. She was calculating the vector. She was measuring the distance between the garage on Emerson Street and the restaurant in San Francisco, between the ThinkPad and the term sheet, between Maria in Bakersfield and the champagne at Tom's table. The distance was ninety-four miles and an infinite moral chasm, and she had crossed it without noticing.

Idealism (0.62): She flew to India the next week. Not for a launch this time. For a reason she could not articulate to her board, to her investors, to herself. She flew to Chennai and took a bus to Periyakulam, the village where her grandmother had spent sixty years organizing grain and walking to the district office and being the only person who understood the forms.

The village had a computer terminal now. NeighborNet had installed it six months earlier, part of the Tamil Nadu expansion. Devi walked to the community center and found the terminal in the corner, its screen dark, its keyboard covered in dust. The satellite connection had failed months ago. No one had repaired it. The village had gone back to the old way—neighbors asking neighbors, the knowledge living in people instead of servers, the vector of connection running along dirt roads instead of fiber-optic cables.

Devi asked the village elder if anyone had used the terminal. The elder shrugged. "They used it at first. But then it broke. Then your company sent messages asking for money. Upgrades. Subscriptions. We do not have money for upgrades. We have money for rice."

The elder did not know who Devi was. She had not told him. She was just a woman from the city, another visitor passing through, another pair of shoes that did not know the dirt roads.

Devi walked to her grandmother's grave. It was on a hill overlooking the village, a simple stone marker among many others. She sat on the ground and looked down at the village—the same village her grandmother had looked at for sixty years, the same patterns of fields and houses and the narrow road to the district office. Nothing had changed. NeighborNet had arrived and departed and left nothing behind except a broken terminal and a bill for services that no one in Periyakulam could afford.

The vector had pointed here once. It had pointed toward connection, toward survival, toward the simple act of noticing when a well was contaminated or a child was sick or a government form needed to be filled out. But the vector had drifted. It had been recalibrated by term sheets and engagement metrics and the gravitational pull of capital. And now, sitting at her grandmother's grave, Devi Krishnan tried to remember where the vector had started and where it had ended up, and she found that she could not separate the two points anymore. The interpolation was complete. The line was continuous. Every point along the path had seemed reasonable at the time.

Greed (1.0): The board voted on January 15th, 2000. The motion was to implement full behavioral data harvesting across all NeighborNet platforms, including the international sites that had been funded by foundation grants and nonprofit partnerships. The vote was unanimous—seven to zero, if you counted Devi's abstention as consent, which the board's lawyer assured her it was. The motion passed. The vector reached its terminal point.

Devi walked out of the boardroom. She walked past the Aeron chairs and the Power Macs and the engineers who had been hired with venture capital to build features that harvested data instead of connecting people. She walked down University Avenue, past the coffee shops and the startup offices and the billboards advertising IPOs, past the city that was burning through the last years of the twentieth century like a comet entering atmosphere. She walked until she reached the garage on Emerson Street, which was now a storage unit for someone else's motorcycle.

The garage smelled the same. Motorcycle oil. Palo Alto dust. The ghosts of her twenty-four-year-old self, the one who had believed that a PHP application could change the world.

Devi stood in the doorway of the garage and thought about the email. Maria in Bakersfield. "I didn't know anyone was listening. Thank you." The email was still taped to the wall of the office on University Avenue—she had insisted on bringing it when they moved. She had looked at it every morning, even as the vector drifted, even as the compromises accumulated, even as the platform that was supposed to connect communities became a platform that harvested them.

The email had been a promise. Not a promise to investors or boards or the algorithms that optimized engagement. A promise to Maria. A promise to Periyakulam. A promise to the version of herself that had sat on a beanbag chair at two in the morning and believed that connection was more important than capital.

The promise had been broken. It had been broken incrementally, one compromise at a time, each one small enough to justify, each one reasonable in the context of the previous compromise. But the sum of reasonable compromises was an unreasonable outcome, and the outcome was this: a platform that had been built to help isolated communities was now a platform that profited from their isolation.

Idealism (0.0): Devi walked back to the office. It was dark now—the engineers had gone home, the coffee shop downstairs was closed, the street was quiet in the way that Palo Alto got quiet after ten. She went to her desk. She opened her laptop. And she began to write.

Not code this time. A letter. To Maria. To Esperanza. To the farmer in Vadugapatti who had asked for seeds and received them. To her grandmother, who would never read it. To everyone who had believed that NeighborNet was what it claimed to be.

The letter took all night. When she finished, the sun was rising over the Santa Cruz Mountains, the same view that the conference room on Sand Hill Road had framed at forty thousand dollars per square foot. Devi printed the letter and placed it on Tom's desk, next to the PowerPoint slides about behavioral data and the term sheet for the Series B.

Then she left. She did not know where she was going. She did not know what she would do. She knew only that the vector that had started in a garage on Emerson Street had reached its terminal point, and the terminal point was a place she could no longer occupy. The interpolation was complete. The line between idealism and greed had been drawn in full, and Devi Krishnan had traveled every inch of it, and the only thing left was to find a new vector. A different one. One that pointed somewhere else entirely.

She got in her car—a used Honda she had bought before the Series A, before the Aeron chair, before the champagne—and drove east. Toward the Sierras. Toward the desert. Toward whatever came after Silicon Valley. The radio played a song she did not recognize. The road was empty. The vector of her life, for the first time in two years, was undefined.

And Devi Krishnan, who had built a platform to connect people and watched it become a platform to exploit them, who had traveled the full distance between idealism and greed and found that the distance was measured in increments too small to notice, who had inherited her grandmother's gift for organization and her grandmother's curse of being the only one who understood the forms—Devi Krishnan drove into the desert and waited for the next vector to reveal itself.