Chapter 9: The Rebuild

November.

David’s checking account: $2,950.

The number should terrify him. Five weeks ago it was $3,847. Now it’s barely enough to cover one month’s rent with $1,350 left over.

But he doesn’t check the banking app as often anymore. Not because the number stopped mattering. Because checking it won’t change anything.

The fourth prototype is taking shape.

Not the fourth iteration. The fourth reimagining.

Alex is sitting across from him, staring at the wireframe on David’s screen. It’s been three minutes since either of them spoke.

“It’s completely different,” Alex says finally.

“Yeah.”

“We’d have to rebuild everything.”

“Yeah.”

“From scratch.”

David nods. The wireframe shows something that looks nothing like what they’ve built before. No list of content. No algorithmic recommendations. No lesson sequences.

Just a blank space. A search box. And underneath: “What are you curious about?”

When you type something—anything—the screen transforms into a living knowledge graph. Your question becomes a node. Related concepts branch out from it like roots spreading through soil. You can see the connections. See how your curiosity touches physics, history, art, science. See that your random question isn’t random at all—it’s a doorway.

You choose where to go. The system doesn’t decide for you. It just shows you what’s possible.

“This is what you wrote at 3 AM?” Alex asks.

“Yeah.”

“Fifteen pages of notes?”

“Yeah.”

Alex leans back. Rubs his face. “David. We have maybe six weeks of money left. Probably less. You want to throw away three months of work and start over?”

“The three months of work was wrong.”

“It wasn’t wrong. It was getting better. The feedback from the last test was positive—”

“It was ‘better.’ Not good. Not essential. Better.” David turns the screen toward Alex. “This is different. This makes curiosity visible. Makes it real. Shows kids that what they care about matters.”

“How do we build the knowledge graph? That’s hundreds of hours of content curation. Thousands of connections.”

“We start small. Twenty topics. Build out from there.”

“We already tried manual curation—”

“Not like this.” David opens another document. “Before, we were curating paths. Start to finish. A to Z. This is different. We’re mapping connections. Any question can lead anywhere. The kids follow their own curiosity.”

Alex is quiet for a long moment.

“The girl from the first user test,” he says. “The one who typed something random.”

“Yeah.”

“This would show her it matters.”

“Yeah.”

More silence. The apartment is cold. David hasn’t turned on the heat. Forty-seven cents per hour, the calculator in his head says. Every hour the heat is on costs forty-seven cents.

“If we do this,” Alex says slowly, “we’re all in. No backup plan. No safety net. If it doesn’t work—”

“I know.”

“Do you? Because I’ve been thinking about backup plans. I’ve been looking at jobs. Not applying. Just looking. And I need to know if you’re—” He stops. Starts again. “I need to know if you’re sure.”

David thinks about the 3 AM moment five weeks ago. The cursor hovering. The escape route open. The moment he closed the browser—all of it.

“I’m not sure it’ll work,” he says. “But I’m sure we have to try.”

Alex nods slowly. “Okay.”

“Okay?”

“Okay. Let’s rebuild.”

They spend the next week mapping.

Not coding. Not designing. Just mapping.

Twenty curiosity topics. Starting points that could branch anywhere:

• How do airplanes fly?
• Why do people fight wars?
• What makes music sound good?
• How does the internet work?
• Why do we dream?

For each topic, they map connections. Not linear paths. Webs. Networks. A kid curious about airplanes might branch into physics, or engineering, or World War II history, or environmental impact of aviation. Every connection mapped. Every node linked.

The work is meticulous. Exhausting. Sometimes they spend three hours debating whether two concepts should be connected, and if so, how strongly.

David’s checking account: $2,650.

Rent paid. Rice and eggs. The heat stays off. He wears two hoodies.

“This connection doesn’t feel right,” Alex says on Thursday afternoon. They’re mapping “Why do we dream?” and trying to connect it to neuroscience, psychology, and creativity.

“Which one?”

“Dreams to artistic inspiration. It’s too vague. Too fluffy.”

David looks at the diagram. “What if we connect through REM sleep? Show how brain activity during dreaming is similar to creative problem-solving?”

“That’s better. More concrete.”

They redraw the connection. Add a node for REM sleep. Link it to memory formation, creativity, problem-solving.

The web grows more complex. More beautiful.

At 11 PM, they step back. The whiteboard is covered in nodes and connections. Twenty topics, maybe three hundred nodes, thousands of links.

“This is insane,” Alex says. Not angry. Just stating fact.

“Yeah.”

“We can’t actually build this in six weeks.”

“No.”

“So what do we do?”

David stares at the whiteboard. All those connections. All those doorways.

“We build enough to prove it works,” he says. “Pick five topics. Map them fully. Build the interface. Show that when a kid types ‘How do airplanes fly,’ they see a world open up.”

“Five topics.”

“Yeah.”

“That’s still a hundred hours minimum.”

David checks his phone. It’s 11:17 PM. “Then we better start.”

Two weeks blur together.

David works sixteen-hour days. Sometimes eighteen. The code sprawls across his screen—graph algorithms, connection weights, visualization layers.

Alex handles the content. Finding the best videos, articles, simulations for each node. Making sure the connections feel organic, not forced.

They don’t talk much. Just work in parallel. Coffee going cold. Meals eaten at desks. The apartment dark except for two laptop screens.

David’s mother calls twice. He answers both times. Tells her he’s okay. That he’s close to something. She doesn’t ask for details. Just: “Are you eating? Are you sleeping?”

“Yes,” he lies.

“David—”

“I’m okay, Mom. Really.”

She doesn’t believe him. But she lets it go.

The balance: $2,180.

Mid-November. Four weeks until rent is due again.

The interface starts working. When you type a question, the graph blooms on screen. Nodes appear. Connections draw themselves. You can click any node, see what it links to, follow the thread wherever it leads.

It’s not perfect. The animations are choppy. Sometimes the connections overlap messily. But it works.

“Holy shit,” Alex says when David shows him the first working demo. “That’s actually cool.”

“Yeah?”

“Yeah. Like—I want to click around. See where things go.”

They spend an hour just playing with it. Typing random questions. Watching the graphs bloom. Following threads from topic to topic.

“We still need more content,” Alex says. “Some of these nodes are thin.”

“I know. But the concept works. You feel it, right?”

Alex nods slowly. “Yeah. I feel it.”

They recruit testers for the day before Thanksgiving.

Not twelve people this time. Just five. High school students. Ages fifteen to seventeen.

David can’t afford to rent the coffee shop room again. They do it at the apartment. Set up one desk, two chairs, laptops ready.

The first tester arrives at 2 PM. Sixteen years old. Junior in high school. Wants to study biology.

David walks her through it. “Type anything you’re curious about.”

She types: “Why do some people get cancer?”

The graph blooms. Cancer appears as the central node. Branches spread: genetic mutations, environmental factors, cell division, immune system, treatment methods, prevention research.

The girl leans forward.

“Whoa.”

She clicks on genetic mutations. The graph reorganizes. Shows connections to DNA, heredity, CRISPR, evolution.

She spends thirty minutes clicking. Following threads. Watching the graph shift and grow.

When they ask for feedback, she’s quiet for a moment.

“I’ve never seen learning look like this,” she says. “Like—I can see how everything connects. It’s not just memorizing facts. It’s understanding how things fit together.”

David’s hands are shaking under the table.

“Would you use this?” Alex asks.

“Yeah. Definitely.”

After she leaves, David and Alex sit in silence.

“That was different,” Alex says quietly.

“Yeah.”

“She actually got excited.”

“Yeah.”

The second tester: “How does electricity work?”

Spends twenty-five minutes following connections from circuits to physics to power grids to renewable energy.

Feedback: “This makes science feel less scary. Like I can start anywhere and just explore.”

The third tester: “Why do people speak different languages?”

Goes from languages to migration to culture to writing systems to brain development.

Feedback: “I feel like I’m discovering things instead of being taught them.”

The fourth tester arrives at 6 PM. Seventeen years old. Senior. Looks skeptical from the start.

David walks him through the same intro. “Type anything you’re curious about.”

The boy types: “How do video games work?”

The graph starts to bloom. Game design, programming, graphics, player psychology—

Then freezes.

David’s stomach drops.

The nodes are rendering on top of each other. A jumbled mess. The connections are overlapping so badly the screen looks like someone spilled spaghetti on a blueprint.

“Uh,” the boy says. “Is it supposed to look like that?”

“No,” David says. His fingers are already moving, trying to refresh. “Give me one second—”

The refresh makes it worse. Now the graph is flickering. Nodes appearing and disappearing. The browser throws a warning: “This page is slowing down your computer.”

“Sorry,” David says, his heart pounding. “Let me restart this—”

But the restart doesn’t help. The graph loads—tries to load—but the layout algorithm can’t handle the complexity. Video games connects to too many things. Programming, art, music, psychology, business, narrative design. The algorithm keeps recalculating, keeps trying to find space for all the nodes, keeps failing.

The screen is a mess of overlapping circles and tangled lines.

“This is confusing,” the boy says. Not mean. Just factual.

David force-quits the browser. Reopens it. “Okay, let’s try a simpler topic. What else are you curious about?”

The boy shrugs. “I mean… that was the thing I actually wanted to learn about.”

“I know, I’m sorry. The system has trouble with some topics. Let me—”

“It’s fine.” The boy is already pulling out his phone. Checking the time. “How much longer is this?”

David and Alex exchange a glance.

“We were hoping thirty minutes,” Alex says. “But if the tech isn’t working—”

“Yeah, I should probably go anyway. I have homework.”

They thank him. He leaves. Polite enough. But David saw his face. The disappointment shifting to dismissal. The look that says: This is just another broken app.

After the door closes, David sits very still.

“That was bad,” Alex says.

“Yeah.”

“The graph can’t handle complex topics.”

“I know.”

“We tested this. How did we not catch that?”

“We tested with the topics we built. Video games has too many connections. The layout algorithm couldn’t—” David stops. Runs his hands through his hair. “Fuck.”

Alex is quiet for a moment. “The fifth tester is supposed to be here in twenty minutes.”

“I know.”

“Do we cancel?”

“No.” David opens his laptop. Pulls up the code. “I can fix this. Add a complexity cap. Limit how many connections render at once.”

“In twenty minutes?”

“I can try.”

David’s hands fly across the keyboard. The layout algorithm, line 347. Add a maximum node limit. Render top fifteen connections only. Priority weighted by relevance.

Fifteen minutes. The code compiles. He tests it.

Types: “How do video games work?”

The graph blooms. This time—controlled. Fifteen nodes. Clean. Navigable.

But shallow. He knows it’s shallow. All the depth is hidden. A user would have to click three times to reach things that should be immediate.

The trade-off: usable but limited. Or ambitious but broken.

At 6:25, the fifth tester arrives. Fifteen years old. Quiet. Nervous.

David walks her through it. Holds his breath.

She types: “Why do we dream?”

The graph blooms. Dreams—psychology, REM sleep, memory, creativity. Eight connections. Manageable.

She clicks around for twenty minutes. Follows threads. Seems engaged.

Feedback: “This is cool. I’d use it.”

After she leaves, David and Alex sit in silence.

Three good tests. One disaster. One okay.

“Three out of five is good,” Alex says. Not quite believing it.

“One out of five was a complete failure,” David says. “The fourth kid left thinking we’re amateurs.”

“So we fix the algorithm.”

“How? If we limit connections, it’s shallow. If we don’t limit them, it breaks.”

“We solve it the same way we solve everything else. We figure it out.”

David wants to believe that. Wants to feel the confidence Alex is projecting.

But he can’t stop seeing that boy’s face. The confusion. The disappointment. The dismissal.

This is just another broken app.

Maybe it is.

Maybe fixing the bugs just reveals bigger problems. Deeper flaws. Fundamental limits to what they’re trying to build.

David closes his laptop.

“I need some air,” he says.

He walks outside. The November evening is cold. Clear. He stands on the sidewalk. Breathes.

Three months ago, they had nothing. Now they have something. But something that breaks. Something that disappoints.

Is that progress?

Or just more expensive failure?

David doesn’t know.

He stays outside for ten minutes. Until his hands stop shaking. Until the panic fades to just—exhaustion.

When he comes back inside, Alex is at the computer. Already working on the algorithm.

“I think I can optimize the rendering,” Alex says without looking up. “Prioritize central nodes. Lazy-load peripheral connections.”

“That’ll take days to build.”

“Then we better start.”

David sits down at his desk. Opens his laptop.

Three out of five tests went well. One was a disaster. One was marginal.

Not the validation they needed. But more than they had.

They keep building.

By the fifth tester, David has stopped taking notes. Just watching. Watching most people lean forward. Watching some people get confused. Watching the product work beautifully sometimes and fail spectacularly others.

All five tests are done. Three loved it. One hated the broken experience. One thought it was “okay.”

Two ask when they can use it for real. One asked if they were hiring testers. One couldn’t wait to leave. One seemed indifferent.

Not perfect. Not a clean win.

But something.

After the last one leaves, David closes his laptop. The apartment is dark. Almost 8 PM.

“We need to talk about money,” Alex says.

“I know.”

“You’re running out.”

“I know.”

“I have some savings. Not much, but—”

“No.” David shakes his head. “You’ve already sacrificed enough.”

“This is my company too—”

“Alex. I appreciate it. But no.”

Silence.

“What’s your plan then?” Alex asks. “You have what, four weeks left?”

“Three, probably.”

“And then?”

David looks at the laptop. Thinks about the five testers. The way they leaned forward. The way they said it felt different.

“Then we launch,” he says.

“Launch? David, we barely have five topics built. The content is thin. The interface is buggy—”

“I know. But we have enough to show it works. Enough to get beta users. If it’s real—if this is actually different—people will use it.”

“And if they don’t?”

Then I’m done, David thinks. Out of money. Out of time. Back to corporate or sleeping on someone’s couch or moving back with his parents with his tail between his legs.

But he doesn’t say that.

“Then at least we’ll know we tried the right thing,” he says instead.

They spend the week after Thanksgiving preparing to launch.

Not a real launch. No press. No marketing. Just posting on Reddit, education forums, reaching out to teachers they know. “We built something different. Want to try it?”

The balance: $1,850.

December 1st.

They have a landing page. A signup form. The prototype live but fragile. Twenty initial beta users lined up—mostly people Alex knows.

David sits at his desk. Finger hovering over the “Publish” button on the landing page.

“You ready?” Alex asks.

No. He’s not ready. The product is incomplete. Buggy. Might crash. Might fail spectacularly.

But he thinks about the girl who said “I’ve never seen learning look like this.”

Thinks about the fourteen-year-old from the first user test who typed something random because she didn’t think it mattered.

Thinks about sixteen-year-old David in that library, hearing Jobs’s voice, carrying a yearning he couldn’t name for ten years.

He clicks Publish.

The landing page goes live.

The first sign-ups start trickling in. Five by end of day. Twelve by end of week. Twenty-five by mid-December.

David watches the metrics obsessively. How long do people stay? What do they click? Do they come back?

The answers: Longer than before. Everything. Yes.

Not everyone. Some people sign up and never log in. Some use it once and don’t return.

But some—a small but growing number—keep coming back. Every day. Spending twenty, thirty, forty minutes following their curiosity through the knowledge graph.

One of them is a fifteen-year-old named Marcus.

David notices him in the analytics. User ID 0047. Logs in almost every day. Average session: thirty-two minutes.

Then, on December 18th, an email:

Subject: “This is amazing”

“I’ve been using your app for two weeks. I’m in 10th grade and I hate school. Everyone tells me I need to study harder, get better grades, beat the other kids. But I don’t care about being better than anyone. I just want to learn stuff that’s interesting. Your app is the first thing that lets me do that. I started with ‘How do video games work’ and ended up learning about AI and computer graphics and now I’m trying to build my own simple game. Thank you for making this. - Marcus”

David reads it three times.

Closes his laptop.

Walks to the window. The city is dark. December evening. Cold. Lights scattered across buildings.

His checking account has $1,620.

Two weeks until rent is due.

The product has forty-three users. Thirty-two active in the last week.

Not enough for revenue. Not enough for investors. Not enough for anything except proof of concept.

But Marcus—a kid who hates school, who doesn’t want to compete, who just wants to learn—spent two weeks following his curiosity and ended up trying to build a game.

That’s not nothing.

David’s phone buzzes. His mother.

He answers.

“Hi, Mom.”

“David. How are you?”

“I’m okay.”

“Really? You sound tired.”

He is tired. Exhausted. Running on fumes. But also—

“We launched,” he says. “The new version. It’s live.”

“And?”

“And it’s working. Not making money yet. But people are using it. Actually using it. One kid emailed me today. Said it’s the first thing that lets him learn what he’s interested in.”

Silence on the other end.

“That’s good,” she says finally. Her voice is careful. Not excited. Not dismissive. Just—careful. “I’m glad.”

“I don’t know if it’ll succeed, Mom. I really don’t. But I think we built something real this time.”

“How much money do you have left?”

David almost laughs. Of course she asks. Of course the practical question cuts through everything else.

“Not much,” he admits.

“Weeks? Months?”

“Weeks.”

He hears her take a breath. Hold it. “And if you run out?”

“I don’t know. But I’m not giving up yet.”

Another pause.

“Okay,” she says. “Just—promise me you’ll eat. And sleep. Even if you’re working all the time.”

“I promise.”

They hang up.

David sits in the dark apartment. Two desks. Two chairs. A whiteboard covered in graph diagrams. A prototype that forty-three people are using.

$1,620 in the bank.

Maybe three weeks left.

He opens his laptop. Looks at the user metrics. Marcus logged in again. Forty-five-minute session.

Then another email arrives. From a user he doesn’t recognize.

Subject: “Question about your app”

“Hi, I’m a middle school teacher in Oregon. One of my students showed me your platform. I have a student who’s been struggling all year—tests poorly, homework incomplete, seems checked out. Yesterday I saw him using your app during free period. He was following connections from ‘Why do rockets work’ to physics to space exploration to climate science. He spent an hour. Completely engaged. I don’t know what you’re building, but it’s reaching kids who normal education isn’t reaching. Can I share this with other teachers? - Sarah Chen”

David reads it twice.

Looks at the balance again. $1,620.

Looks at the metrics. Forty-three users. Growing slowly.

The math doesn’t work. Not yet. Maybe not ever.

But somewhere in Oregon, a kid who tests poorly and doesn’t do homework spent an hour learning because he wanted to.

And a teacher wants to share it.

David opens a new document. Types: “December 18 - Teacher outreach plan.”

His fingers move across the keyboard.

Outside, December continues. The city lights shimmer. The cold presses against windows.

Inside, David keeps building.

Not because the numbers add up. Not because success is guaranteed.

Because Marcus exists. Because Sarah Chen exists. Because somewhere, kids who hate school are finding that learning doesn’t have to be a cage.

The balance keeps dropping. The runway keeps shrinking.

But the product is alive.

And that’s enough. For now.

Christmas week.

David doesn’t go home. Can’t afford the plane ticket. Tells his mother he’s too busy. She doesn’t argue.

Alex flies home to see his family. David spends Christmas alone in the apartment. Rice and eggs. No decorations. Just his laptop and the metrics dashboard.

Fifty-eight users now. Forty-six active.

Three teachers have signed up. Two have emailed asking about classroom use.

The product is still fragile. Still incomplete. But it’s growing.

David’s phone buzzes. A message from Alex:

“Merry Christmas. Told my family about the launch. They think we’re crazy. Probably right. But also—my little sister tried the app. Spent two hours on it. She’s 13. She said it’s better than school. That’s something.”

David types back: “Merry Christmas. Tell your sister thank you.”

He sits at his desk. The apartment is quiet. Outside, families are probably having dinner. Opening presents. Being normal.

He thinks about last Christmas. Still at TechCorp. Salary. Benefits. His parents proud of him. The simple life he could have kept living.

Then thinks about this Christmas. Almost broke. Alone in a cold apartment. Building something that fifty-eight people use.

He doesn’t regret it.

The old life was a slow death. This—even the fear, even the uncertainty—is alive.

David opens the product. Types into the search: “Why do we celebrate Christmas?”

The graph blooms. Christianity, winter solstice, cultural traditions, commercialization, family, meaning.

He clicks around. Follows connections. Ends up reading about winter solstice celebrations across cultures.

The product works. Not perfectly. But genuinely.

His phone buzzes again. Not Alex this time. An automated alert.

Bank account balance: $1,520.

Rent due in thirteen days: $1,600.

David closes the alert.

Opens his code editor.

Starts fixing bugs. The connection rendering is still glitchy. Some nodes overlap. Some animations stutter.

He works until 3 AM. Not because he has to. Because he wants to.

The craft itself. The elegance of making something work beautifully. The same feeling from that night in June when he caught himself smiling while debugging.

Life has become amazingly enjoyable. Not despite the difficulty. Because of it.

He saves his work. Closes the laptop.

Tomorrow the sun will rise. The users will keep using. The balance will keep dropping.

But tonight—Christmas night, alone, almost broke—David feels something he hasn’t felt in months.

Not confidence. Not certainty.

Peace.

The kind that comes from knowing you’re exactly where you’re supposed to be, even when nothing makes sense.

He falls asleep at his desk. The apartment cold. The city quiet. The future uncertain.

But the work—the real work—finally aligned with the vision he’s been carrying since he was sixteen years old.

That’s worth more than a warm apartment.

More than certainty.

More than safety.

David sleeps.

And dreams of knowledge graphs blooming across screens. Of kids following their curiosity to places the education system never let them go.

Of freedom spreading like roots through soil.

December 31st.

The year ends not with celebration but with spreadsheets.

David sits at his desk. Alex is back from visiting family. They’re doing end-of-year review.

Users: 67 Active in last 7 days: 51 Teachers using it: 5 Revenue: $0 Checking account: $1,510 Days of runway: 8-9

Alex closes his laptop. “We need to talk about January.”

“I know.”

“You can’t pay rent.”

“I know.”

“I can cover it. One month. But that’s all I can do.”

David looks at him. “Alex—”

“I’m serious. One month. February you’re on your own. If we don’t have revenue or funding by then, I’m getting a job. Not quitting. But I can’t burn through my savings too.”

It’s fair. More than fair. Alex has already given eight months.

“Okay,” David says. “Thank you.”

They sit in silence for a moment.

“What’s the plan for January?” Alex asks.

David pulls up a document. “Teacher outreach. We have five teachers using it. If each one shares with five more, that’s twenty-five teachers. If each teacher has thirty students…”

“That’s seven hundred fifty potential users.”

“Yeah.”

“Still no revenue.”

“No. But proof that it works at scale. That’s what we need for investors.”

Alex nods slowly. “You really think investors will care?”

“I don’t know. But I know we can’t keep doing this without money. So we need to try.”

“Have you thought about what you’ll say? How you’ll pitch?”

David hasn’t. The thought of pitching—of packaging this into metrics and projections and slides—makes him tired.

But the alternative is giving up.

“I’ll figure it out,” he says.

That night, alone, David opens a blank document.

Title: “Investor Pitch - Draft 1”

The cursor blinks.

What does he say? “We have sixty-seven users and no revenue but trust me it’s going to change education”?

No investor will care.

He thinks about Marcus. About Sarah Chen’s student. About Alex’s little sister spending two hours following her curiosity.

How do you turn that into a pitch deck?

David stares at the blank document for twenty minutes.

Then closes it.

Opens his email instead. Scrolls through the messages from users. From teachers.

“This is the first app my son actually wants to use.”

“My students are asking if they can use it in class.”

“I’ve been following connections for an hour. I’m a teacher and I’m learning new things.”

The investors won’t care about these. They want growth metrics. Revenue projections. Market size.

But these are why it matters.

David saves the emails to a folder. Labels it: “Why.”

Then opens the blank pitch document again.

This time, he types:

“The education system measures students by how well they memorize and compete. We’re building the opposite: a tool that starts with curiosity and shows students their interests matter.”

It’s not a pitch yet. But it’s true.

That’s enough to start with.

David types until midnight. The document grows. Still rough. Still missing the numbers investors want.

But honest.

When the clock hits midnight—January 1st, 2026—David is still typing.

A new year.

Nine days until rent is due.

Sixty-seven users who believe in something different.

And a pitch deck that might convince someone to believe too.

Or might not.

Either way, David is still here. Still building.

Still in love with the why.

The year ends. The new one begins.

The work continues.

End of Chapter 9