(AN: Chapter longer usual than enjoy)
1992
Age 13
The lecture hall smelled like chalk and warm plastic. Every seat was taken. People sat on the steps and along the back wall, notebooks balanced on knees. I arrived early enough to choose a spot near the front where I could see the board and the projector without turning my head. Paige slid into the seat beside me a minute later. She set her bag down, capped pen aligned with the zipper, and looked at the blank screen. She didn't say anything. She didn't need to.
Dr. Asher stood at the front, hands folded behind his back, posture that said he trusted the room to settle itself. He waited out the last bits of noise. Someone laughing in the doorway, a backpack zipper that refused to cooperate. Then he tapped the edge of the podium twice, the sound small and precise.
"Good afternoon," he said. "Thank you for coming. Today our top entries will describe their models. You have ten minutes each and the same two obligations. Be clear and be honest."
He didn't announce the order until the projector warmed and the fan stopped sounding like an old car.
"We'll begin with team Navarro-Pruitt. Then Singh-Chambers. Then Cooper-Swanson."
Paige's fingers tightened once around her pen and relaxed. She glanced at me, quick. I nodded. Fine.
The first team's slides were clean and blue. They modeled the spread of heat through a metal plate with a defect, then compared analytical and numerical solutions. Their code output lined up with a textbook curve so perfectly it looked drawn after the fact. Asher asked two questions about boundary conditions. They answered like they'd practiced in a mirror. The room approved.
The second team studied rumor transmission in a dorm with a single elevator and a broken stairwell on weekends. Their graphs moved like sound waves. Peaks of interest, valleys of indifference, a strange wobble whenever the elevator failed. They laughed once at their own joke about social entropy, then caught themselves and apologized to no one. Professor Kim leaned forward and asked how they measured interest without lying to themselves. They admitted the proxy was weak and pointed to the error bars. The room forgave them because they had measured the weakness instead of pretending it wasn't there.
Between the second and third teams, Dr. Li erased the corner of the board with a felt eraser and wrote 10 in small numbers where the clock was out of view. Ten minutes. Not optional.
"Asher," she said quietly, "let them run their demo."
He nodded.
Our names appeared on the screen. Cooper-Swanson. Paige stood first. I followed.
Facing a room makes it larger. The lights on the projector washed out the first row, turned faces into pale shapes with dark eyes. I set our notebook on the ledge, as if paper could hold the rest steady. Paige plugged the cable into the department's computer. The image flickered once and settled.
"Our phenomenon," she said, "was the flow of people through a cafeteria during peak hours. We chose it because it's bounded and observable, and because it contains both conservation and randomness."
Her voice was even. She pointed without stabbing the air, palm open to the slide the way good teachers do. I advanced to the next image. Our floor grid with entrances, serving line, and tables marked. Sources, resource, sinks. Simple enough to recognize in a breath.
"We built a probabilistic agent model," I said. "Movement rules depend on local density and a small memory parameter that can be set on or off. Noise enters through shocks, spilled tray, early line closure, and through the clock. Arrivals follow a distribution we measured."
Paige clicked to the plot of density over time. Rise, wobble, plateau. She had chosen a restrained font that made the figure look honest.
"We validated against five observation windows," she said. "We used independent counts to estimate throughput and mean time-in-system. The model held within our error bands even when we varied the seed."
Dr. Li's pencil tapped once against her notebook. Approval, or maybe just rhythm.
I started the demo. The grid filled with small dots. People entering, tending toward the serving line, splitting around obstacles, seeking tables, exiting. At first the motion looked chaotic, then it arranged itself the way flocks do. No conductor, just rules. A murmur moved through the room like a small wind and then settled.
Asher lifted a hand. "Pause at ninety seconds, please."
I paused. The screen froze with three narrow currents drawing around the center tables and a thin diagonal cut where only a few agents dared cross.
"What creates that diagonal?" he asked.
"Sound," I said. "When ambient noise rises above a threshold, some agents switch from shortest path to path of least attention. It's small but consistent."
"And you measured attention how?" Professor Kim asked, not unkindly.
"Proxy," I said. "We used a decibel reading near the entrance and assumed correlation with visual noise. We tested the rule by faking the sound, chair scrape, and saw the cut appear."
"Honest," Asher said. Not praise. A label.
Paige resumed the run. The plateau steadied. The colors cooled.
When we finished the ten minutes, Asher opened the room. A faculty member near the aisle, the kind who wears a tie when everyone else wears a sweater, asked why we hadn't used a queuing model instead of agents. Paige answered first.
"Queues fit arrival and service well," she said. "They don't capture path choice. Our interest was movement under crowding, not just wait time. We chose a model that could be wrong in more ways so we could be honest in more ways." She paused. "It cost us simplicity."
"It bought you realism," Dr. Li said, almost to herself.
A graduate student in the back asked about scalability. More doors, more rooms, multiple floors.
"It scales," I said, "but the parameter that acts like memory becomes harder to justify. We kept it binary for that reason. If you let agents remember too much, you're writing their biography instead of their behavior."
That got a small laugh. Not mine.
Time ran out exactly when Dr. Li's chalk 10 became a dot. We thanked the room. People clapped the way mathematicians clap. Politely, with a kind of relief that nothing exploded. We took our seats again.
Asher didn't announce the results right away. He let two more short talks, honorable mentions, fill the remaining slot, then stood at the podium and stacked the note cards he hadn't looked at once.
"Thank you," he said. "You were clear. Some of you were honest. That matters more than you think." He glanced at the clock. "Our judges have conferred. Third place, Singh-Chambers. Second place, Cooper-Swanson. First place, Navarro-Pruitt."
There was clapping again, and a small exhale beside me that I felt more than heard. I watched Paige's face out of the corner of my eye. No shock, no smile. Just a soft settling, as if an answer had landed in the expected range. The room asked first place to come forward for two additional questions. We stayed seated.
Second place is a clean number. It has direction without finish. While Navarro answered a question about heat loss at the edges, hand poised like a conductor, my mind reformatted the afternoon into a list. What we did right. What we hid. What we admitted. What we didn't. I felt no sting. Just a narrow line of wanting to test the model under something it hadn't yet seen.
When the hall emptied, the noise spiked and then thinned. Students had somewhere else to be. Faculty collected papers the way birds collect small stones. Dr. Li stopped near us and didn't pretend she had more time than she did.
"Well done," she said.
"Thank you," Paige said.
Dr. Li looked at me. "You left the memory switch binary."
"Yes."
"Keep it that way," she said, and moved on.
Professor Kim appeared with a half-folded handout and the look of a man who gets places by accident on time. "Your seed choice," he said. "Clock-based?"
"For the runs we showed," I said. "Repeatability in our logs."
"Good," he said. "Bad models need one seed. Good models can tolerate many."
Asher came last. He extended a hand. We shook. His grip was ordinary.
"Your defense was clean," he said. "You two disagreed when you wrote it."
"Yes," Paige said.
"Keep disagreeing," he said. "Collaboration without friction is choreography."
He headed for the door, windbreaker already half on. The room dimmed as the projector cooled. Paige and I stood there a moment like we'd come up for air and found the surface as expected.
"Second," she said.
"Second," I said.
"Better than fine."
"Better than fine," I said, and meant it.
We collected our things. The hallway felt louder after the insulated quiet of the lecture hall. Students passed with exam packets and coffee cups, the carry of their voices bouncing off the lockers. We walked in step because we always do. Near the stairwell, she stopped.
"You were good at questions," she said. "Kim likes you."
"Kim likes code that doesn't lie," I said.
"You don't lie," she said, not quite a compliment. "I'm going to eat. You need anything?"
"No."
She nodded once. "See you tonight. Library."
"Right."
I watched her take the stairs two at a time. The set of her shoulders looked different from two weeks ago. Still narrow, still carrying weight, but more consistent with what the bones were built for. If Spring Break had reset her, it had done it without fuss.
Outside, the light was wrong for the hour. Thin clouds flattening color. I crossed the courtyard and cut through the rec center because I always cut through now. The boxing room door was open. The timer sat silent on its hook, red light dark. A man I hadn't seen before taped a split knuckle with the indifference of practice. The coach was fixing a chain.
"You training today?" he asked without looking over.
"Not tonight."
"Then don't stand under the bags."
"Understood."
He went back to work. The chain clinked once, steady.
The library's atrium felt too bright after the gym. Paige had already claimed a table. Far wall, no traffic behind her. Our notebook lay open between two pencils aligned like rails. She looked up when I reached the chair opposite, then down at the page, then back up.
"We should archive the code," she said. "And the raw counts."
"Agreed."
"And write a one-page addendum about ethical limits," she said. "We were careful, but we didn't say we were careful."
"I'll draft it," I said.
"Okay," she said, and the word held more trust than agreement.
We worked for an hour without speaking. She handled figure captions. I documented the pseudo-random seed policy and the bounds on our shock rule. The room breathed. Cough, page turn, shoe scrape. All the usual noises of people trying hard. At some point my right hand found the pattern of the keyboard the way my left hand had found the rhythm of the bag. Timing is a transferable skill.
At seven, she set her pencil down and leaned back, eyes on the ceiling grid like she was counting squares.
"Hungry?" I asked.
"A little," she said, surprised it could be true.
"Cafeteria," I said. "Ten minutes."
We walked in the light that looked like the end of a day that had been longer than the clock admitted. The cafeteria lines were reasonable. The currents we had modeled earlier in the semester moved with a familiar caution, students negotiating trays and tables without naming the negotiation. We ate near the edge of the room. The map we had drawn months ago still fit.
"Are you happy?" Paige asked, neutrally.
"Yes," I said. "It was honest work."
She nodded. "I'm happy," she said after a beat. "I wanted first. Second feels like reality behaving."
"Reality is a good model," I said.
She smirked at her tray. "Don't put that on a T-shirt."
We finished and didn't linger. Outside, the air had cooled. She looked toward the dorms.
"Library again?"
"I'm going to write for Kim an explanation of the seed choice," I said. "Then yes."
"Okay," she said. "I'll get the references right."
At the stairs to her floor, she paused.
"Stephen?"
"Yes."
"Thanks for the last weeks. The hard part."
"Better with friction," I said, borrowing Asher's line.
She almost smiled. "Night."
"Night."
In my room, the desk lamp made a small cone of steady light. I turned the boxing wraps out to dry where the heater could find them, then opened the notebook and wrote a single paragraph about random seeds and trust. The idea that a model that survives different starts is worth more than one that only survives the lucky ones. The words came easily. Simple claims write themselves when you've already done the proof.
I packaged the code and counts into a tidy archive, named with the date. The computer fan spun a dull circle that matched the sound in my head. I stood and stretched until my shoulders unknotted. The clock hands sat where they always sit at this hour.
I thought about the room earlier. The board, the timer, the faces going blank while they listened. I thought about second place like a coordinate. Not a loss, not a win. A point with direction. In the back of my mind, a straight jab placed itself cleanly where it belonged and returned on the exact same line. Not violence. Geometry with consequences.
Before I shut the light, I wrote one line at the bottom of the day's page.
Control is not winning; it's staying inside your own form.
I closed the notebook. The room held still. Tomorrow the system starts again. That's the work.
Thanks for reading, feel free to write a comment, leave a review, and Power Stones are always appreciated.