Chapter 2 - Learn and Adapt
Jeff Chen
The rooftop restaurant was all glass and candlelight, perched high enough over the city that the streets below looked like a glittering circuit board. Jeff still sometimes balked at the cost of restaurants like this. Growing up, money had always been tight, every dinner bill weighed against what it meant for the rest of the week. Adapting to the high-end lifestyle had been an adjustment—easier than he’d expected, if he was honest. It didn’t take long to get used to being pampered and catered to. But every once in a while, when the check arrived, some old reflex stirred and he felt the pinch, the quiet instinct to wince. Not tonight. Not with Camila. She leaned forward across the table, her black silk dress catching the amber glow from the overhead heaters. She was laughing at something he’d said about Ewen’s latest eccentricity, her hand brushing his, and for a moment Jeff let himself savor the thought: a year ago, this would’ve been unthinkable.
Camila was the kind of woman who belonged in glossy travel magazines, not swiping right on guys like him. But a sudden surge in his net worth had felt like a call-up to the majors, and here they were — champagne on the table, filet mignon on the way, the city at their feet.
Halfway through the entrée, his phone buzzed. A text from Rina Suresh, one of the senior engineers:
AIFusion just dropped Sauce v4. Big update.
Jeff’s pulse ticked up. Sauce — their top rival hadn’t even waited a day from their earnings announcement to release. A calculated move no doubt. Omni had been the undisputed leader when it launched. The gold standard. The AI everyone else measured themselves against. But in recent months, Sauce had started stealing headlines—flashy new features, splashy demos, clients migrating over just to be associated with the buzz. It was like the new kid in school who was instantly cool, not because they’d earned it, but because they were new. Jeff told himself Omni still had the depth, the enterprise backbone—but the press didn’t care about backbone. They cared about shiny.
Another text followed before he could respond:
Text-to-video with audio. Full minute. Realistic voice. Insane.
He slipped the phone under the table and glanced back at Camila. She was describing some villa in the Maldives, the turquoise water, the open-air showers. He smiled, nodded — but his screen lit up again.
New legal brief generator. Writes airtight contracts in seconds. Financial forecast tool passes analyst accuracy benchmarks.
Camila’s smile faltered. “Am I boring you?”
“Not at all,” he said quickly, sliding the phone into his lap. But then it buzzed again. This time the sender made his chest tighten:
Ewen: All hands. One hour. In person. No exceptions.
Jeff exhaled. “Camila… I’m sorry. I’ve got to run into the office. Work emergency.”
Her brow arched. “Now?”
“Now. But — look, order whatever you want. I’ll make it up to you. We’ll do that Maldives trip you were just talking about. First class. Promise.”
Her lips pressed into a line, but she nodded. “You’d better.”
An hour later, Jeff was standing shoulder-to-shoulder with fifty other engineers in the company’s largest conference space. Ewen Frank paced at the front like a field general, sleeves rolled up, hair even wilder than usual.
“We are in the fight of our lives,” Ewen said, eyes blazing.“AIFusion thinks they can steal our thunder? No. We own this space, and we’re going to prove it.”
He stabbed a finger toward the whiteboard where a list of bullet points glowed on the screen.
“Two months. Here’s what I want: Omni to match and surpass Sauce’s text-to-video—but I want real-time rendering. A multimodal personal agent that doesn’t just answer, but anticipates—across legal, medical, finance, and creative fields. Full context memory across all user interactions, no matter the time gap. Instant multilingual translation in voice and text with zero latency. And I want an adaptive reasoning layer that learns from each user and improves on-the-fly.”
The room was silent except for the scratching of pens. These weren’t goals — they were moonshots.
Ewen grinned like a man announcing a sure bet. “Do this, and we don’t just stay on top. We own the decade. And when we do, the bonuses will make last year look like a rounding error.”
The team erupted into murmurs. Jeff’s mind was already racing through timelines, architecture changes, data sourcing. It felt impossible — which was exactly why Ewen expected them to pull it off.
Cory Yates
When Corey finally learned the reason he’d been fired, the anger came fast.
The partners had signed a licensing deal with an AI company specializing in legal services. The math was simple: pay for cutting-edge software and cut junior staff to balance the budget. His position had been collateral damage.
Once the anger ebbed, curiosity took over. He’d been so immersed in mastering statutes and case law, he’d barely noticed the AI revolution happening outside his bubble. Now, reading about it, he couldn’t look away.
He marveled at what the tools could do — generate airtight contracts, predict litigation outcomes with 90% accuracy, cross-reference decades of precedent in seconds. Half of him bristled at the idea that a machine could do in minutes what took him years to learn. The other half thought: If you can’t beat ’em, join ’em.
So, he tinkered. At first it was the kind of silly, surface-level stuff everyone did when they stumbled onto these tools — typing in prompts to see a picture of himself as a medieval knight or generating a video trailer for a fake courtroom drama starring his college buddies. He even had an AI clone his voice reading Shakespeare just to see if it worked. Harmless distractions, the kind of parlor tricks that made him laugh and shake his head in disbelief. But then curiosity nudged him further. He began automating tedious forms, drafting arguments, cleaning up citations. And soon, bolder experiments: chatbots that could field client questions, little programs that drafted invoices or scheduled reminders. The shock wasn’t that the AI could do these things, but that he could coax it to. Corey Yates, the lawyer who couldn’t even figure out how to sync his phone contacts, was suddenly “writing code.” Sure, it was a stretch to call it that—the computer did most of the heavy lifting—but he was able to take an idea from concept to execution. Watching a script he’d pieced together actually run left him gob smacked. It was like discovering he’d been given a second career he hadn’t trained for.
Somewhere in the middle of all this, he met Janet. They’d matched online, and he’d been drawn to her simplicity — no designer handbags, no exhausting small talk about climbing corporate ladders. She worked as a nurse’s aide, loved hiking on weekends, and laughed easily. In the chaos of his life, that grounded him.
But dating without a steady paycheck was like carrying an overly filled glass of water across a balance beam. Every extra dinner or weekend brunch shaved down his severance. He’d started with ninety days. A handful of outings, a weekend trip, and before he knew it, he was down to forty.
That was when he reached out to Martin Weiss, a mentor who ran a two-partner law firm out of a converted brownstone downtown. They met at a café near the office, Martin in his usual off-the-rack suit with a loosened tie, Corey in something just polished enough to pass.
Over coffee, Corey laid it all out—losing his job, how blindsided he felt, the months of debt looming over him, and the way he’d thrown himself into learning AI. “I don’t know where I fit anymore,” he admitted. “Everyone’s saying this is the future, but the more I learn, the less sure I am what a lawyer like me is even supposed to do.”
Martin listened quietly, sipping from his mug, his eyes narrowing just slightly in thought. Finally, he set the cup down and said, “You want something real to work on? Automate our legal billings. Right now, I’m doing it myself, and it’s eating into my billable hours. If you can free me up, I’ll pay you for it.”
Corey blinked. “You’d trust me with that?”
Martin gave a wry smile. “You’re smart, and you’re hungry. That counts for a lot. Besides, it’s not glamorous work. If you screw it up, worst case, I still do it myself like I’m doing now. Best case, you save me ten hours a week. That’s worth something.”
It felt a little like charity—Martin had always looked out for him—but Corey didn’t care. Charity or not, it was a paying gig. He needed the money. More than that, he needed a win.
And suddenly the stakes felt higher. Messing around with AI to generate knightly portraits or goofy video clips was one thing; creating a product someone was actually paying him for was something else entirely. This wasn’t a toy anymore—it had to work. At first, he didn’t even know where to begin. He had never thought of himself as someone who could build anything technical, but AI shouldered most of the heavy lifting. His legal background let him steer it through the subtleties: how time entries should be categorized, how to apply retainers correctly, what invoice language reduced the odds of disputes. Even so, the cracks showed. AI often spit out half-baked code or hallucinated functions that didn’t exist, and Corey, brand new to the whole world, would sit staring at error messages he barely understood. The combination wasn’t quite the blind leading the blind, but it wasn’t far off.
He’d prompt the AI for a script, run it, and hit an error. The first time it happened, he almost quit. But he learned that if he uploaded screenshots of the error messages, the AI would diagnose the problem and suggest a fix. He’d copy the solution in, run it again, and either move forward or hit the next roadblock.
It was slow, frustrating work. He’d never had to think like a programmer before, never had to keep track of so many tiny, interlocking details. But each small win — each bug squashed — felt like a little jolt of momentum.
Still, there was pressure. Martin had offered to pay, and Corey couldn’t hand him something half-baked. It wasn’t just about the money anymore; it was about proving to himself that he could do it.
Finally, after a string of late nights, he thought he had it. The program was supposed to pull case time logs, generate an invoice, write a cover email to the client, send it, and log the transaction in the firm’s records. He’d probably failed a hundred times already, fighting one stubborn glitch after another.
Is this what it’s like to be a programmer? he wondered, flexing his sore wrists.
He clicked the button. For a heartbeat, nothing happened. Then, one by one, the programs kicked into gear. An invoice populated in PDF form. An email drafted itself, addressed to the right client with the correct retainer balance. A “sent” notification appeared, and a second later, the billing log opened in Excel.
Corey blinked twice to make sure he was seeing it right. No errors. No missing fields. It had worked.
Dolores Holmes
Sitting at the kitchen table long, Dolores stared at the same columns of numbers she’d been staring at for days. The spreadsheet on her laptop glowed with unforgiving precision: rent, utilities, groceries, medical costs, clothing, insurance, gas, car payment. No matter how many times she adjusted the figures, the math refused to bend. She had already cut everything she could — canceled the streaming services, put off buying new shoes, stopped grabbing her morning coffee. There was nothing left to trim.
It was like standing on the edge of a cliff with no rope, no net. Every new expense was a gust of wind threatening to push her over. She felt the familiar weight in her chest, the dull ache of stress that never quite left her. Was this it? Was this her life now — divorced, scraping by, forever running in place while the world around her surged ahead?
In a moment of bitter honesty, she admitted to herself: she was out of ideas.
Her eyes drifted to the open browser tab she’d been ignoring for weeks — an article about AI assistants. The headline promised “Your New Digital Partner: Smarter Than You Think.” She’d rolled her eyes when she first saw it. It sounded like just another overhyped fad, like those diet shakes or miracle budgeting apps that claimed to change your life and never did. But now, with nothing left to lose, a reckless thought slid into her mind: Why not give it a shot?
With a sigh, she clicked the link and downloaded Sauce, the AI everyone was raving about.
She started small, almost embarrassed, as if the program could judge her desperation. “How do I pay off unexpected medical expenses without sinking deeper into debt?” she typed.
The response came instantly. Suggestions popped up on the screen, practical and detailed. Some of it was stuff she already knew, but a few ideas actually made sense — things she hadn’t considered, like negotiating directly with the hospital for a payment plan or applying certain tax credits she hadn’t thought she qualified for. It wasn’t magic, but it was more helpful than she expected.
Her heart gave a little jolt of cautious hope. Maybe, just maybe, there was something here. She leaned forward, fingers trembling as she typed her next question, a little bolder this time: “How can I save more money for a house?”
That answer was even better. Sauce broke down her spending, showed her what to cut and how much it would save, and even offered a plan to funnel the savings into a separate account. A simple, concrete plan — the kind she hadn’t been able to sketch out on her own.
And then, the question she was most afraid to ask: “How can I make more money on the side?”
The screen filled with suggestions: freelance bookkeeping, reselling unused items online, tutoring math, using platforms to pick up quick gigs. Dolores clicked one link, then another, and before long she was scribbling notes like a student cramming for finals.
She tried one of Sauce’s ideas the very next day. Listing old baby clothes and kitchenware she no longer used. Then, with the AI’s help, she streamlined her accounting work — shaving nearly an hour off her daily tasks by automating repetitive entries. That gave her just enough time to test out another side hustle in the evenings.
It wasn’t glamorous. It was exhausting. But it was working.
By the end of the month, she’d scraped together almost $800 — not the full $2,000 she needed, but far more than she would’ve believed possible a few weeks ago. More than anything, she felt something she hadn’t felt in months: proud.
Sauce became her quiet companion, her co-strategist, her secret weapon. It planned her budget, reminded her of deadlines, even suggested workouts when she complained about being too tired. Every time she sat down at her computer, the AI was waiting, ready to help her chip away at the mountain in front of her.
When Greg called to say he’d gotten one job offer but at 10% less pay — which meant reducing child support — Dolores braced herself for the familiar wave of panic. But it didn’t come. Instead, her mind clicked into gear. She already had a plan. And now, she could make a new one.
For the first time in a long time, she didn’t feel like she was drowning without a life preserver. She flipped open her laptop and typed into Sauce:
“How do I make up for another 10%?”
Chapter 2 - Commentary
After revising chapter 1 results with different character motivations and expanding key moments, I moved on to chapter 2.
The Good: It took the prompts and delivered mostly what I asked for. The quality of the writing was very good.
The Bad: If I didn’t provide the drama, it didn’t “fill in” those gaps. As an example, after reading Corey Yates’ chapter 2, it was missing something so I had to go back and create the opening of the Janet storyline.
The Ugly: For this chapter, the AI did a very good job overall. It didn’t do anything terrible. This was the least rewritten chapter of the entire story.
If I’m being nitpicky, I’d say that I had to go back at the end and then insert elements that made the entire story more cohesive, which is a bit of a hole, but it isn’t AI’s fault. It doesn’t see the overarching story. It only sees what you tell it is right in front of it.