A LITTLE OVER AN hour later, and poorer by eight dollars spent on a ham-and-cheese croissant and coffee and the bus, he looks up from the iPad and watches a raven-haired woman apply makeup to her pale face with help from a pocket mirror. Pale like his mother on that last visit, her head rolled to the side, dry lips, Jeremy wondering whether someone is going to say something, anything, something conciliatory, or maybe one last twinkle-eyed parry on her way to eternity.
The bus slows. It mostly emptied on the circle through downtown. Just two stops now until Jeremy’s condo. To-do list: shower, change clothes, get a backup battery, walk to the office, go deeper into this list of variables.
He looks down at the iPad to see what it’s already given him: a list of 327 parameters that the computer used as a basis for predicting an attack.
A column on the left on his iPad window lists the name of a particular variable; then, in the next column, the variable is quantified, like the number of oil barrels shipped or produced. Next to that, one column shows how much the variable has changed in the last day and another column shows how much it’s changed in the last week. Finally, a column on the far right shows the extent of the change.
Jeremy scrolls up and down this dense, text-heavy list. Three of the lines of variables are blinking. The blinks signal that the variables in those columns have seen a significant change.
When he first built the program, he’d look every few hours at the variables, looking for changes, like a hopped-up day trader or, Emily once observed, like a psychology student who, a few days into class, starts imagining that he and everyone around him has everything from clinical depression to multiple personality disorder to attention deficit disorder. Jeremy was seeing conflict everywhere.
But he quickly realized that changes to the variables, even the most weighty ones, did not mean imminent war. Much more important is the combination of changing factors, their relative and combined influence in waging war. The precise right amount of sugar and butter and chocolate makes delicious cookies; the wrong amount, combined with arsenic, makes something that tastes like shit and kills you.
What he’s looking at, even if the data is accurate, still is likelier to be a predictor of a hoax than a war. Jeremy looks at the blinking columns. One is Tantalum. The second is Conflict Rhetoric. The third is the Random Event Meter, known as REM.
The bus slows again. Jeremy glances at the countdown clock: 57:40:00.
57:39:59.
57:39:58.
He looks up. The raven-haired woman applying makeup, sitting two rows ahead, is looking at Jeremy in her tiny handheld mirror. Isn’t she? As the bus comes to a stop, she hastily stuffs her makeup kit into an oversize purse and exits out the front.
Jeremy looks down at the variables. Rhetoric. The computer is programmed to see changes in language, to see nuances even in hyperbole. But from the column, Jeremy can’t tell exactly what has changed in the language of conflict, or where in the world it has changed. That level of abstraction will take more research, another command to the server, easily enough done when he gets upstairs. He’ll be able to tell if the language of conflict has intensified in, say, the Baltics, or among Greek bankers, or in Iran.
All Jeremy can tell from this column is that there has been a sharp uptick, 14 percent, in the last few days, of the collective language of conflict, material enough for the computer to care about.
There’s been an even sharper rise in the Random Event Meter. It’s up 430 percent. Jeremy shakes his head, mostly annoyed, vaguely curious. The meter measures whether there has been some event or series of events that, in historic terms, would seem far outside the standard deviation. And the event can be anything. An alien landing. All Major League Baseball games being rained out, or being decided by more than 30 runs. Statistical anomalies unconnected to any of the other variables.
Jeremy clicks on the line. He sees a story from the Associated Press. The headline: “Lions Freed in Three Zoos, Leaving Man Dead.”
He glances at the article. Lions loosed over the last twenty-four hours at the San Francisco Zoo, and in Oakland and in San Diego. Outside the San Diego Zoo, an old man found dead from claw wounds. Police speculate the man was responsible for freeing the lion. They speculate further that this is some effort coordinated among animal rights activists but acknowledge the theory is for the most part conjecture.
An off-duty cop shot the lion in San Diego. Animal control managed to dart and subdue the one in Oakland. The one in San Francisco remains at large.
It’s all Jeremy can do to stop himself from rolling his eyes. Fucking computer, fucking nonsense. Maybe the three lions were planning to get together to cause the apocalypse.
He turns to the third variable, tantalum. That’s up 4,017 percent.
The precious metal is integral to the making of cell phones. Tantalum, Jeremy recalls, comes from refining a raw material called coltan, which, in turn, is found in mines in Uganda and Rwanda.
Jeremy knows this not because he knows a bunch about how cell phones are made. Rather, he knows it because he knows how wars are made. And the battle for coltan once set off a rash of insurgencies in Africa. Simply: demand for coltan, which sold for $100 a pound when Jeremy was studying the conflict region in 2001, fueled efforts to control the mines and the municipal governments by guerrilla bands. It was a great place to explore localized conflict, even inspiring a visit once by Jeremy, then a graduate student, to try to put a human face on the variables measured by his computer brain.
Jeremy looks at his phone, sensing this might be further evidence of a hoax. After all, what difference could tantalum make in the possible onset of conflict? The sharp rise in shipments of tantalum, 4,017 percent, is certainly material, at least to the brokers and buyers and suppliers of the metal. But it could hardly, Jeremy imagines, have anything to do with tipping the balance of massive global conflict. He can imagine himself talking to some news blogger explaining he thinks that there’s going to be a nuclear holocaust because of an increase in the shipment of some precious metal. Oh, and the release of some zoo lions.
You don’t say, Mr. Stillwater? And for a follow-up question: are you still taking your meds?
The bus slows. Jeremy hears the doors open. He hears the rain. As he stuffs his iPad into the backpack, he sees the slick streets and smells the wet air.
How long has it been raining? How come he hadn’t noticed? Is the rain the forest, or the trees?
Fucking Emily. And that guy, there was something about him that didn’t add up. Something deliberate. A too on-point jacket, like a costume, but with shoes from an entirely different circus trunk. No, that’s too simplistic. Was he too handsome, too deliberate in some way? Jeremy can’t pinpoint what nags him.
Outside the bus, he walks against the foot traffic to his apartment complex cum condo, all steel and sharp corners, modernity that is San Francisco’s version of urban renewal. The high-rises near the ballpark are beautiful. What twenty-something could resist? Perfect for the Trustafarians and stock option babies able to throw down 20 percent of $1.2 million for nine hundred square feet that serves as nest and résumé. The elevator ride is a networking opportunity and a speed date.
He takes the elevator to the eleventh floor. Feels a sense of relief as he reaches his floor, maybe a chance for a catnap. He slips his key into 1117 and opens the door.
He sees the innards of the couch. They’re strewn all over the living room. Someone has gutted his sofa.
He peers to his right. Papers tossed on the counter that separates the kitchen from the living area. A stainless-steel spatula, an egg beater, lying beside their ceramic container, upside down on the counter.
Magazines on the floor.
Someone has overturned his apartment.
Someone still inside?
Jeremy backs out and shuts the door. He catches the eye of a young woman coming out of the apartment next door. He hates this woman, a bubbly, friendly, sycophantic thing, working at a startup in South Park. Actually, he once made an overture. When she first moved in. She rejected him in such an indirect but insurmountable way that it infuriated Jeremy. He has slipped several notes under her door asking her to turn down her music and, one night when particularly piqued at the sounds of sex coming from next door, anonymously warning her that it will set a bad precedent in her relationship if she fakes orgasms. Lies beget lies, he warned her. The building manager got involved but nothing could be proved, quiet warnings issued; the thing blew over.
“Excuse me, Tara.” He sees that her umbrella, while it isn’t unfurled, has Mickey Mouse prints.
“Hey.”
“Strange question. Have you seen anyone coming into my apartment? Anyone not me?”
Her mousy face, beneath a bob haircut, registers a genuine concern.
“Did someone take something?”
He shakes his head. “I’m not sure. I might have had an unwanted visitor. Did you see anyone? Last night?”
“Did you tell Aaron?” The building manager.
He’s starting to get irritated. Just answer the question, pixie. She must sense it.
“No,” she says. “I went to bed after Idol. It was quiet last night.”
He looks at the lock. No signs of forced entry. He opens the door.