Simone was correct. The Rabbit’s Foot was always looking for processors.
Theo, perspiring, wasn’t certain whether he ought to be excited. He had called the number the following morning, cross-legged on his duvet. The person who answered was terse and professional. Yes, they had some vacancies; yes, he could come in. They gave him an address, an appointment, and sixty minutes to arrive.
He had expected something glamorous, mysterious. Instead the office was in a plaza near the motorway, squat and flat, across from a field of weeds. Signs advertised the offices of a community radio station, an artisanal salsa lab, an improbably confident ribbon manufacturer (The “Best” since 2015!). A long-languishing bowling alley trembled at the end of the lot.
Suite 500’s buzzer was unnamed on the intercom. When he hit it, the lock thunked open and Theo took the steps two at a time, bicycle helmet under his arm. The staircase ended at the fifth floor: a heavy door, shiny black, with the numbers 5•0•0. Theo hesitated for a moment and then left his helmet in the stairwell, on the second-to-highest step. He ran his hands through his hair.
The door revealed an open-plan office that seemed to go on forever. The light fell evenly over glass desks and flat-screens, cyan-tinted water coolers, men and women in voluptuous, translucent office chairs, their work-stations divided by short blue partitions. Nearly everyone he could see, as many as a hundred people, was wearing headphones. Nobody was listening to their own typing, clicking, scrolling; heedless of the clack and whirr, they consigned these sounds to him.
“Yes?” The man behind reception had taken off his headset.
“Theo Potiris. I have an interview?”
“An interview?” The man had a birthmark across his chin, like an ink spill.
“For the, uh, processor gig?”
“Oh, you’re here to take the test. Use the terminal under Mr. Peanut.”
Theo looked around. There were a few free workstations in the row nearest the entrance. The last of these, next to the wall, sat below a large black-and-white poster. A peanut-shaped biped leaned on his cane, winking through a monocle, under a bold heading that read MR. PEANUT FOR MAYOR.
“There?” Theo asked, pointing.
“Yeah.”
Theo went and sat down. The screen came to life as soon as he pressed a key.
THE RABBIT’S FOOT
Processor Screening
build 4.1
(Spacebar to begin.)
Theo shifted in his seat, trying to get comfortable. He spacebarred.
The test was in fact a series of tests, each one incorporating its own tutorial. At first the tasks were simple: count the number of white circles in a green colour-field; press a key when an animated figure crosses a dotted line; identify the pentagon in a series of hexes. Next came a segment focused on spreadsheet skills. Theo was already adept, an expertise acquired through years of bookkeeping, but even if he had not been the exam might have been okay, since the software explained exactly how to obtain the result it wanted. He looked up from the computer at a clock. About twenty minutes had passed. No one was paying him any attention. The dog-eared corner of Mr. Peanut’s poster quivered in the exhalation of Suite 500’s ventilated air.
The third and final section of the screening test required the user to study video footage from sporting events. Whereas the earlier segments had relied on text prompts, now the terminal instructed Theo to put on a pair of wireless headphones.
He did, and a woman’s voice began speaking: “Processors are interpreters. They interpret between two languages: the language of human life and the language of software. The first of these vocabularies consists of sense data, perceptions, clumsy human judgments. The other is made up of clean, quantitative data. This is our work. Converting what you – the aspirant processor – discern, into what our system – the Model – understands.”
The woman’s voice was undecorated and almost severe. She did not need to be liked or to spark the listener’s interest. She sounded middle-aged, a native speaker; the way she emphasized certain words — movement, discern, the Model – gave the impression that she completely understood what she was saying, that perhaps she was not even following a script.
She continued, “Here is the first example. A tennis match. Click the mouse whenever either player strikes the ball with their racket.”
Theo did this. With every click, a numeral in the corner of the screen increased by one.
“Serves should not be counted the same way as other strikes. Instead, indicate a serve by pressing the S key on your keyboard.”
Then: “The only events that matter are the events in the match that are designated as significant.”
And, “When a player misses the ball, press the spacebar.
“Try to press the keys or click the mouse as soon as possible after the event occurs. The more accurate you are, the better your screening score.
“Good. Continue processing the match until you hear a long, uninterrupted tone.”
So Theo learned to process a tennis match. He clicked the mouse button and touched the spacebar and pressed S. His hands found the most efficient finger positions without his ever needing to think about them and after a few minutes he found that he was not watching tennis at all, not in the traditional sense: he was watching the two players – it was Nadal, it was Djokovic – without admiring or considering them. All he was seeing was the movement of the ball and the instants when the ball was met or missed, or served.
A tone sounded, long and low, like an engine catching. The video flickered and was replaced with a racetrack. Greyhounds. Theo couldn’t remember ever having actually seen a greyhound race. It seemed like a trick, that these lean, eerie mammals should be configured at a starting line and made to compete. “These are greyhounds,” offered the voice. The image freeze-framed. “Click once on each greyhound.” Theo did. The starting pistol was fired, like an inner tube popping. The greyhounds leapt into life, fervid and absurd. Dogged, Theo thought, and then the video paused again. “Click once on each greyhound.” Theo did, squinting to make out the blur of their contours against auburn earth. The race resumed. Then another pause, the same instruction, and on and on until the fastest dog nosed across the finish line. Freeze-frame. “Click once on each greyhound. This helps the Model to distinguish the competitors from their environment.”
From greyhound races the screen transitioned to basketball, with keys to press for dunks or fouls. “A for a successful steal, Z for an attempted steal.” The voice explained what a “steal” meant; the software didn’t require a familiarity with the sport or its rules. The rules, in fact, seemed irrelevant: to process a game you didn’t need to follow it. Once real life had been processed into events and time-stamps, ones and zeroes, the algorithms could do the rest, deducing the ways these numbers fit together. Perhaps steals presaged losses. Perhaps a greyhound’s subtle movements could forecast its future powers. It didn’t matter whether the Model understood the rules or strategies of basketball, tennis, racing: just this is a basket and this is a serve and this is a purebred nosing over the finish line, and what are the ways that each of these things lays bare what will happen tomorrow?
A long, low tone, like a ship’s groan. The basketball game snapped off, leaving the words, “Screening complete.”
Theo waited for the woman in his headphones to say something. She remained silent. Eventually he got up, removing the headset. It felt strange to do it unbidden. All through Suite 500’s broad office, he saw now, processors were processing video. There were screens with dog races and screens with horse races and screens with race cars and soccer players and gymnasts, and in front of them men and women wearing headphones, clicking, watching, waiting, touching A or Z or P or; if the events demanded it.
“I’m done,” he said to the man with the birthmark on his face.
The man looked up. “What?”
“The screening. I’m done.”
“Oh.” He craned his head around. “You were on number four? Under Mr. Peanut?”
“Yeah.” Theo clasped his hands.
“All right,” said the man. He pressed some keys. He had kept his headset on and this bothered Theo. Didn’t Theo deserve the man’s full attention? Was he trying to send him a signal? And then the thought occurred to him: at this very moment, while Theo stood with clasped hands, the receptionist might be in the act of processing him, using keystrokes to record his actions and reactions, and to predict what might come next. Press C if the applicant coughs.
Theo unclasped his hands.
“Okay,” said the man, “you scored an eighty-five.”
Theo was pleasantly surprised. “That’s good, right?”
“It’s out of a hundred and thirty.” He noticed Theo’s disappointment and added, “But it’s not bad. It’s within the bounds of tolerance.” He consulted a chart. “If you can start today, I’ll put you under Mr. Matisse.”
Theo assumed he was referring to another poster on the wall, a print of Goldfish or The Dance. But after his silence was taken for assent the man picked up a telephone receiver and enunciated into it, “Mr. Matisse? I have that processor for you.”
He saw Theo staring.
“Fifteen dollars an hour,” the receptionist whispered, “excluding lunch.”
Mr. Matisse turned out to be a reedy man in slacks and a button-up shirt, with a grand pile of corkscrewing hair. He was young, with the tanned/clammy mien of an elite postgraduate student. He seemed thrilled to encounter Theo. “A rat for the maze!” he exclaimed. “Just kidding. It’s more of an analytical warren. A series of caverns. You learn your way around. Come this way. A maze is by its nature confusing, no? Or else it ceases to be a maze. A labyrinth must be labyrinthine. Yes? No? Perhaps I’m wrong. Something to think about. Don’t step in the recycling bin.”
As Matisse conducted him through the rows of terminals, Theo began to distinguish a mild shabbiness to the office, the sense of a place that was installed in top condition and then neglected. Chair wheels whined, partition walls were frayed, screens hunched on their mounts. A couple of nineteen- or twenty-year-olds sat murmuring at a conference table, huddled over what looked like a phone book. They didn’t look up as they passed. Matisse halted at an unkempt desk. He motioned for Theo to sit.
“Your throne,” Matisse said.
Theo hesitated; he had assumed this was Matisse’s workstation.
“Do you have a name?” Matisse asked. “Or do you prefer to make one up?”
“Theo.”
“Pleasure. I’m Roberto. I go by ‘Matisse.’ It’s my surname. I do appreciate Matisse’s canvases but the moniker is honestly earned, not an affectation. May I join?”
“Sure?” Theo said, confused.
“Thank you. I will fetch my chair.”
Theo lowered himself into the voluptuous, translucent chair. He would call the store at lunch, he decided, and get Esther to take care of his chores. For now he surveyed the workstation. A computer, headset, keyboard, mouse. A pad of foolscap and a dull orange pencil. Rummaging through the desk he found a few more pencils, pens, a half-erased eraser in the shape of a raspberry. The drawer at his knee concealed empty Kit-Kat wrappers and an over-saturated photograph of a man and a woman. It appeared to be, but was not necessarily, the sample photograph that comes with a frame. A solitary one-panel comic had been thumbtacked to the partition on his right: an anthropomorphized key having a conversation with an anthropomorphized lock, asking, “You busy Friday?”
Finally Matisse returned with his own voluptuous, translucent chair. He threw himself down on it. Men and women all around them continued working without looking up from their monitors – horse races mostly, baseball, soccer, someone processing what appeared to be synchronized swimming.
“To be perfectly candid, I wasn’t expecting you,” Matisse said. “Dolores said it would take at least a week for me to be assigned any processors. ‘You can’t rush incoming,’ she said. But lo and behold, fifteen hours later, here you are. Hello and welcome. How are you liking it so far? Of course you don’t even know the story! The fact is: you are my very first processor. The first! Hello! Welcome! Believe me, I’m as nervous as you are. Most likely even more nervous. Why should you be nervous? You shouldn’t be nervous! Processing is passive, procedural. It does not require independent thinking. At least it shouldn’t! If you’re thinking independently, you’re doing it wrong. ‘We can’t have the processors getting creative.’ Isn’t that the case in so many industries? Isn’t that a late-capitalist status quo? ‘Ask the bricklayers! Ask the —’ Oh, haha, that’s pretty good.”
He was looking at the comic with the lock and key. “Is that yours? Yes, well, it’s yours now! Right? I tell you: in a perverse way, it’s nice to be back at the ole’ 500. I spent the past four months auditing the Dublin systems. Half of that chez Dolores, mind you, but two full months on the Emerald Isle. Something weird in their greyhound spread. But we found it. It just takes patience. ‘Doctor, we need some patients!’ Sometimes I postulate that this is the most important criterion for this job. The capability to sit with a problem. To roost until it hatches. As Jackson Pollock once said, ‘The truth will never divulge itself but the falsehoods will.’ For a little while I encouraged the other specialists to call me Pollock, until one of the code guys took it for a reference to fake crab. I suppose Matisse makes more sense. It is, after all, my name. But sheesh: his paintings are so easy to adore. I would prefer an artist whose work was more abstruse. Those are the greatest, are they not? When the story’s a little harder to work out? Defer narrative and representation: devise a work that functions at the level of apprehension. A chasm, a mirror, a burning bush. If there’s any logic it should be the logic of dream. Tea?”
It took Theo a few beats to recognize that this was a genuine question. He knew this type from comedy clubs. He was okay with it. “No, I’m good,” he said.
Matisse yanked a courier bag from the back of his chair. “I keep a thermos,” he explained. “It’s green. The tea, not the thermos. Haha. I discovered this tea in Hong Kong last year. Have you ever been? The Rabbit always puts us up at the Palace of All Workers’ Aspirations. It’s a fancy hotel – goose-feather pillows, chrysanthemums in vases, a Rodin in the lobby – but the tea at the bar’s like bathwater. I think it’s deliberate, to compel the tourist class out of the hotel, into the streets. Down the hill you find one of the city’s oldest tea markets, rows of stalls with baskets and bags and barrels of teas, popcorn genmaicha and black keemun, twisted saffron bundles, emperor’s buds, orange-blossom and Fujian white and the cagada-like pellets of divine, aromatic oolong. A paradise. And scales everywhere – if you have an interest in scales, as I do, you could wander the tea market just for that, a history of weight and measures, perfect little instruments, some that collapse like Swiss army knives to hide in a tea-seller’s coat. So you go from tea stall to tea stall, drawing deep breaths, inhalations not aspirations – am I boring you? It’s not necessarily that I talk a lot but when I’m interested in a subject I try to exhume it totally. The other side is that when I am working I find I plunge myself into the material. I dive, you understand? Captain Nemo, Jacques Cousteau. Did he actually climb into the submarine or did he stay in the boat? No, of course Cousteau went in the submarine. I’ve seen the films. YouTube. It’s not my fault. I’m a millennial. I appreciate the term because it assigns me an era, situates me in time. Pliocene, Neolithic, Post-Impressionist. I fit within a taxonomy. What about you? No? Were you born before 1980?”
“1981,” Theo said.
“Ah, near-millennial,” Matisse declared. “Cusp. Picasso in his Blue Period. Blogs? Yes. Emojis? No. How many fingers do you use to type?”
“Listen,” Theo said, “are we supposed to be working together?”
“Oh, sure,” said Matisse. “Sure we are.”
And so it unfolded. Roberto Matisse recounted and digressed. Roberto Matisse showed Theo the ropes. Over the coming days, the Rabbit’s newest recruit learned the why, how, whither of this role. He watched tape and processed it. He clicked and counted, weathered Matisse’s soliloquies. Whereas most of the Rabbit’s processors worked in isolation, obeying the instructions that came from their headsets, Theo had a real live boss. Experimental research, skunkworks; Matisse was on a mission and Theo was his subordinate. He processed what Matisse told him to process, using schemas that Matisse had designed, investigating the inklings that Matisse had dreamed up, months prior, before Dolores sent him over.
“Who’s Dolores?” Theo asked, after the umpteenth mention.
Matisse began to laugh. “The Dolores is a place, my man. It is headquarters, the Dolores Building, the edifice in which everything of import occurs.”
Even from Simone’s description, Theo hadn’t expected the organization – the association – to be so vast. Matisse’s job title was “specialist,” way up the org chart from the processors, but he still seemed a long way from leadership. The group’s upper echelons – attorneys, data scientists, software engineers – worked at the Dolores and other offices off-site, far from the processors’ suite. The scale was bewildering, as was the Rabbit’s Foot’s tone: studiously legal, meticulously bold. It earned its revenues by placing winning bets, in bulk – thousands a day – with bookmakers across the world. Not a single one of these bets was placed by a human being. Every decision was made by an immense and complicated computer program, a network of predictive algorithms, housed on an encrypted server. “The Model,” said Matisse. “She’s the prettiest woman on the catwalk.”
The Model decided what bets to place – which horses or greyhounds or 100-metre dashers to back. The Model decided how much to bet and when to take a pass. Assembled line by line, module by module, over years, it was the association’s crown jewel and its hidden advantage, the source of all its edge. The Model knew that right-handed pitchers were more likely to strike out left-handed batters. It knew that teams that won a lot were more likely to win in the future. It knew the age at which tennis players, or thoroughbreds, begin to fall off. It learned these principles, and unimaginably more complex ones, by parsing the data, running the numbers. The Model also contained the numbers, all of them, in a gargantuan codex of sporting statistics. The bulk of these statistics came from proprietary sources: on-base percentages and home run rates from official MLB databases, racing records from bookmakers, prospect rankings published in the trades. The next data set came from processors: workers reviewing video, tallying whatever they were told. They also helped the Model to learn to help itself: training the software to see the things they saw. Therefore Theo at his terminal, distinguishing greyhounds from the dirt.
Most of Suite 500’s processors had been repeating the same tasks for months. Perhaps the Model had uncovered that the length of a basketball player’s stride had a bearing on the odds of his team winning their game. A processor would be assigned to watch the matches, pausing and playing, clicking footfall to footfall, repeating the same operations every day, or most days, for weeks on end. Piece by minuscule piece, the Model gained confidence in what was likely to happen. It calculated odds for every wagerable event and whenever these odds exceeded the bookies’ – even by a small amount – the Rabbit’s Foot’s software automatically placed a bet. Volume mattered most: a coin-flip you can predict even 50.1 percent of the time, over thousands and thousands of tosses, across millions and millions of dollars, turns pocket change into fortunes.
“Here’s the trick,” said Matisse one lunch break. “Win more than you lose.”
Matisse’s big idea was weather. Employees of the Rabbit’s Foot were always trying to find new sources of edge: it was the way to earn standing and to get a big payday. If one of your findings was incorporated into the Model, you’d receive a proportion of the corresponding winnings.
“You mean me, too?” Theo asked.
“Processors don’t count,” said Matisse. He gestured out at the glassy-eyed workers. “You’re the cannon fodder. To cash in, you need to have come up with the idea. Most of the time it’s a linkage we hadn’t noticed yet, something hiding in our statistics. Maybe certain soccer clubs should avoid taking corner kicks. Or teams play better when they’re wearing red uniforms.”
The other approach was to find something useful that nobody was counting. “There is a literal infinity of attributes to observe in the world.” Matisse was twirling cold soba salad around a plastic fork, shedding peas and cranberries and shards of tempeh. “So how do we know which ones to record? The knife, the fork, the lunch? The number of calories? Of bites? Of colours in the bowl? It’s an act of discernment, curation. The painter and his subject, choosing colours from a palette. The scientist and her hypothesis. One choice among an infinity of choices – just so for the gambler trying to ascertain what’s significant.”
Theo said, “So you thought to yourself, ‘Weather!’ ”
“I thought to myself, ‘Weather!’ and I convinced Mitsou to let me give it a shot.”
Theo paused in his banana-peeling. “Mitsou?”
Matisse pitched his fork into the garbage. “Mitsou San Marziale,” he said. “Mistress of the Model, queen of the gamblers, Our Lady of the Numbers.”
Simone’s sister. From Mitsou’s office on the Dolores Building’s thirty-sixth floor, she oversaw the association’s wins and its losses, the division and multiplication of stakes. She had made her first fortune predicting two of the Final Four in the 2004 NCAA basketball tournament, using a computer system that grew up into the Model. “She was a one-woman Computer Group, an ingénue Ranogajec,” Matisse insufficiently explained. “They say she grew up fiddling with Markov chains, stochastic calculus. Romping Bayesian and non. Dogged. Self-taught. Everyone was like: ‘Holy shit!’ ”
After years of loyal service, Matisse had won the Lady’s permission to pursue his interest in the impact of weather on sporting events. No more missions to Dublin or Hong Kong, no more all-nighters scraping Olympics data or squishing the bugs in the Model’s rugby sevens code: Mitsou had allocated him a single junior processor and three months to ascertain whether heat waves and rain clouds could be baked (or soaked) into the Rabbit’s predictions. Did weather nudge gaming outcomes? Could it help predict them? “We’re sprinting through as many meteorologies as we can,” Matisse said.
Theo’s main job was logging the weather conditions on archival sports tape. Every morning he biked away from Provisions K, sat at his computer (headphones on), and turned his attention to mud on racetracks and sunshowers on baseball diamonds and the rare, semi-apocalyptic hailstorm on a soccer pitch. Sometimes the apposite question was whether an arena roof was open or closed – skimming a broadcast for overhead angles. For two days he watched tennis. For three he weathered NASCAR. Then he spent an interminable week of fact-checking international weather databases, evaluating the accuracy of their hourly local temperatures. A chime would bell in his right ear, followed by a message superimposed over some field of uniformed athletes. Dataset states “VERY HOT.” Press <Y> if appears correct, <N> if erroneous. He’d watch until he witnessed a nonaerobic rivulet of perspiration, a player removing his cap and shaking his head at the sky. Then he’d press Y, and the game would change to another, with a new chime and a fresh inquiry. Dataset states “AVERAGE COOL.” Press <Y> if correct, <N> if erroneous.
“Average cool,” Theo murmured to himself.
After work he’d cycle back to the store, folding unannounced into the fray. He’d check that the specials were restocked and the pushcarts squared away. He’d touch base with the fish-boys and the section managers, Esther and Mo, and the white-coated curators of the cheese pit. He’d speak with Mireille, if she was around, though usually she wasn’t. He avoided the basement, his mother’s untidied desk. Bit by bit, day by day, as Theo made his presence in Provisions K less critical, the store’s lieutenants took over.
I’m chasing a bigger return, he told himself. Never mind the hastiness of his career-change, its worrying resemblance to a midlife crisis. Never mind that this gambit didn’t seem sensible – an international gambling ring, $15 an hour. Of course it didn’t seem sensible: the best bets rarely do. But looks can be deceiving, as most successful gamblers know. There’s no reason to put it all on red except, intriguingly, that there’s a chance to win.
Theo was standing at an overlook when his phone rang. It was a Sunday morning and he had decided to bike up the mountain, a half hour each way, taking advantage of cool air and honeyed light. It would be hot later. Now, looking out over the city, he felt sweaty, satisfied, a fully functioning human being. His bicycle was leaning against a guardrail and the scrubby bushes behind it were studded with berries. Luscious or poisonous he wondered, luscious or poisonous, luscious or poisonous.
His phone rang; some convoluted number. He answered.
“Hello?”
“Theo?” said Lou.
He forgot at once the berries, the bushes, the sky, the view, the polished disarray of his hometown. “Lou?” he said into the phone.
“Yes, hi.”
“I can’t believe it. Are you okay?”
“I’m fine. I’m still here.”
“Where?”
“Morocco.”
“The retreat.”
“The retreat,” she agreed.
“I thought there weren’t any phones.”
“There’s one, a landline at the campus. For emergencies.”
“Is this an emergency?”
“Not like that. I wanted to talk to you. It’s good to hear your voice.”
“It’s good to hear yours,” he said.
“You sound okay. You sound well. I thought about calling when your letter…I’m so sorry about your mom.”
“Thank you. Yeah, I’m okay.”
“Are you at the store?”
He moved his phone to the other hand. “I’m on the mountain. It’s the weekend. I mean, I have a new job – did you get my letter?”
“I don’t think so. I don’t remember anything about a new job.”
“I’m doing forecasting,” he said. He tried to describe the Rabbit’s Foot, calling it a lucky break, a lucrative opportunity. He did not tell her about the commute, the cubicles. He told her it drew upon his untapped skills. Theo had been talking for some time when he registered that she was not saying much herself. It was not like her.
“Lou,” he said.
“Yeah?” she said.
“What is it?”
She didn’t reply at first.
She said, “I’m extending my stay.”
“You’re not coming home?”
“Not yet,” she said, in the same cautious tone. “Maybe at the end of the summer.”
“ ‘Maybe’?”
“This is a lot, Theo.”
“I know,” he said, though he didn’t really know.
“Will you keep sending me letters?” she said. “I read them all.”
It was only after they hung up that he considered whether Lou sounded different than she had before she left. Whether she sounded sad or happier, or more equivocally changed. Had there been someone else in the room? He found he could not remember precisely the way she used to sound, before she went away, when they had walked by the tracks or lain together in her bed.