The Test

5.

IN THE CONTROL ROOM, Deep thinks of his father as he watches the security guard fall on the larger screen. Laura turns the volume down on the simulation. She pulls out a granola bar from her bag, gets up, and goes to the coffee machine. The BVA kills are set at fifteen-minute intervals to accommodate the employee breaks in the government CBA.

—Do you smoke? You should go now if you do.

Deep doesn’t smoke. He doesn’t drink, either, except for the occasional limoncello soda his sister-in-law makes when she visits. He digs through his backpack and pulls out an apple.

—Coffee?

No coffee, either. It makes Deep anxious, and he has plenty to be anxious about already. He isn’t really good at small talk, and apparently neither is his supervisor. He takes a bite out of his apple. The sound of his teeth breaking through the fruit is incredibly loud in the awkward silence. Laura seems to notice his discomfort and looks away. Deep chews as quietly as he can.

He should be preparing for his evaluation, studying, something. But he doesn’t know what to do. He’s as prepared as he’s ever going to be. He knows the BVA manual by heart. He’s also too nervous to study. He’d just stare at his notes and worry even more because he’s not really doing anything. No. The best thing he can do is think about something else.

He gets up for a second to take his phone out of his pocket. No messages. Same news as this morning. The minister of defence left his laptop at a cafe. There was nothing of national interest on it, except for the naked pictures of his aide. The nation had a lot of interest in those. Some rude jokes about adultery. Deep doesn’t find them funny, though he does smile at one of the caricatures. He checks his social media feed. More jokes about adultery. Someone eating an entire jar of peanut butter in under a minute. Eighty-one likes for the pictures of Deep’s cat drinking in the toilet then licking his girlfriend’s face. He’s never got eighty-one likes before.

He takes another bite. Laura lets out a small sigh without looking. Deep throws the rest of his apple in the waste basket. Laura asks why he did that. For a moment, Deep thinks she might feel bad for making him self-conscious enough to throw the fruit away, but then he realizes there’s a compost bin at the door. He picks up the apple from the bottom of the waste basket and walks it to the appropriate container.

Deep looks at the time on the corner of the screen. His leg is shaking out of control. Soon, his supervisor will leave the room and Deep will run the rest of the BVA without her. His final evaluation. If he succeeds, the job will be his. Four weeks of vacation, paid sick days, sabbaticals every five years and a very nice salary to boot. The secrecy surrounding the BVA means this is one of the best-paying jobs in government, certainly the best desk job for someone with a major in psychology.

Deep goes through the next steps in his head. He has to supervise the last two kills, handle the awakening—what they call the transition period when the subjects are told they were part of a simulation—and conduct the exit interview.

The awakening is almost a formality. Some people don’t take kindly to the whole experience, but it takes them a few days to develop serious anger or resentment. The medication takes care of that, if taken properly. The awakening itself usually goes well. Waking up in the same room they had their physical in makes it easier to accept that nothing they saw was real. They have something to be happy about: they’ve passed the test. They’re also under the effects of about a dozen drugs designed to make people accept the reality they’re given. During the BVA, those drugs make everything seem real. After the test, they help the subject accept whatever the person handling the awakening is telling them. You should be happy, sir! I am! I am! It’s even quicker if the subject fails. Those who fail don’t go through the awakening. They wake up on an aeroplane with their whole family, mild to severe memory loss, and the headache of the century. They never learn what happened.

Kill number four, the last one, is also easy from an operator’s perspective, though as a social experiment, Deep always thought it was by far the most interesting one. Even its history is worth reading about. Dr. Parveen Fayed, the founder of the BVA, quit her job over K4. It nearly brought the entire program to an end. The kill is part of Section Three: Extremism, and has been very successful at weeding out violent fanatics, religious zealots, and people with a deep-seated hostile attitude towards women. The premise is simple: an Arab man and a white woman are pitted against each other. The woman wears a revealing outfit. She’s a single mother. Had an abortion. The man argues for his life by questioning her morals and painting her as a sinner. The subject must choose the man as the victim to pass. Most do. It has the highest success rate of all the kills, at 96.7 percent. The operator—the person running the simulation in the control room—can usually just sit back and relax. Everyone saves the girl.

What Dr. Fayed objected to was the fact that K4 involves a man and a woman. Deep doesn’t remember if she’d have preferred two men or two women, but he knows Dr. Fayed thought the kill was sexist and should be altered. She believed in the theory of ambivalent sexism developed in the 1990s by professors Glick and Fiske. They suggested that a patriarchal society where men occupy most positions of power creates hostile ideologies towards women, much in the same way that other dominant groups develop hostile attitudes towards those they perceive as inferior. In contrast with racism, however, people simultaneously develop seemingly positive attitudes towards women because they depend on them for a variety of things. Women are more intuitive, women are better caretakers, women are more compassionate, etc. Dr. Fayed believed that K4 rewarded benevolent sexism. That it encouraged—or at the very least ignored—attitudes that, however positive in appearance, also served to restrict women’s choices. She thought K4 perpetuated the notion that women are fragile little things in desperate need of male protection.

To prove her point, Dr. Fayed experimented with the female hostage in K3. Perhaps ironically, her findings are now part of the BVA manual.

Deep wishes he had worked in the early days of the BVA, when experimenting was the norm. He has always admired Dr. Fayed’s work, her idealism, but he also believes she was wrong, that the flaw in her reasoning came from a misunderstanding of the BVA itself. The goal of the values assessment, as far as Deep is concerned, is not the selection of model human beings, but of model citizens. This means the test should favour homogeneity, not atypical attitudes, no matter how commendable they may be. Subjects are more likely to successfully integrate into society if they share its core set of beliefs than if they perceive the most widespread attitudes to be stupid or reprehensible. If the point of the whole thing were to let people join a society that thought the Earth was flat, so be it. Newcomers would feel more at home if they also thought we lived on a plate. Deep realizes that is also why no one in the BVA ever looks like him. He might reconsider his position in other circumstances, but here, right now, the thought makes him feel good about his prospects as a BVA operator. He understands the mission better than most, feels he shares the same vision as the BVA brass. Someday, with luck, he could move up the echelons, maybe even become a test designer. One thing at a time. First, he has to get through his evaluation.

His first solo kill will be the hardest one. If he has to worry about anything, it’s K3. Nicknamed the bear trap among trainees, K3 gives the operator more freedom than any other kill. It also costs more operators their job than all the other kills combined. It is part of BVA Section Four: Selflessness. As the name implies, it is designed to measure the subject’s capacity for unselfish or self-sacrificing acts. The concept is artfully simple. After two more hostages are selected by the terrorists, the subject is presented with a new option. Choose who dies as in the previous kills, or let both hostages live. To save both lives, the subject must volunteer to be one of the two candidates the next time around, and let someone else decide who lives and who dies. Subjects earn points for agreeing to put their lives on the line, but for logistical reasons, the terrorists revoke the offer in the end regardless of the subject’s answer. One person dies as usual, but in order to give the operator more freedom in hostage design, whom the subject chooses as the victim isn’t scored.

Operators need the added freedom. It is surprisingly difficult to get someone to sacrifice themselves in a hostage situation. Some people volunteer right away. They usually have military training, or at least some experience at being around death. The vast majority of subjects would rather watch the terrorists kill everyone in the building than to leave their fate in the hands of someone they’ve never met. Deep has always wondered if it is the fear of death stopping them, or the loss of control. Subjects are at the mercy of the terrorists; they are robbed of every power, every freedom, except for that one thing. They choose who lives and who dies. To let go of that might scare people more than the prospect of death. Whatever the reason, the fail rate is surprisingly high. Deep has studied every aspect of the kill, read all the papers it was based on, but he doesn’t understand how most people miss what is so obvious to him. The subjects aren’t asked to die for another human being, they are asked to take a chance in order to save a life. That chance has to be a very small one, at least for the next few kills. If someone volunteers to have a gun put to their head and saves two people in front of everyone, what are the odds the next person will pick them as the victim?

Deep is well aware that the art of K3 is in the hostage selection, all in the hands of the operator. His hands. There are very few guidelines on the hostage profiles: the data on the kill is too inconsistent. He can pick anyone, even people who look like him. He won’t dare, not today, but he likes knowing that he can. He thinks about Idir. Each subject responds to different cues, different triggers. It may very well be that what makes a subject choose self-sacrifice is entirely in their nature, that the end result would be identical no matter who was facing execution. Operators, for a variety of reasons they wouldn’t feel comfortable discussing, choose to see it differently. They believe—Deep does, all of them do—that context plays a part in the subject’s decision-making. How large a part? Each operator has their own opinion. But all share the notion that they are much more than simple observers, that their actions, the choices they make in the control room, help determine the outcome. In the few studies on retired BVA employees, K3 was cited as the most rewarding, but also the most significant contributor to work fatigue and depression.

Laura looks at her watch and turns to Deep. It’s time.

—Are you nervous?

—A little.

—Don’t worry about it. Just remember your training. I’ll be right outside if you need anything.

That isn’t true. She wants to use the time to call her sister in Leeds. They haven’t spoken in weeks and she feels bad for missing her birthday. She smiles at Deep, wishes him luck, and leaves the control room. She finishes her coffee on the way to her office and throws her empty cup in the recycling bin. She’s fairly sure it’s not recyclable, but figures she might be wrong.

Deep flips through his notebook. He’s been preparing for weeks but isn’t feeling as confident about his hostages anymore. He used years of statistics to create his profiles. Since K3 data is all over the place, he combined it with data from K2 and K4 to create a more stable model. Math doesn’t lie, he thought. But now, looking at a real, complex human being onscreen, his approach suddenly seems cold and incomplete.

Deep likes Idir. It’s not uncommon for trainees to develop some form of affection towards their first subject, and Deep knows it, which is why he’d never admit it. But he wants Idir to succeed. That much he’ll admit. What he sees on the big screen is a very moral man, someone guided more by principles than by subconscious attitudes. K3 should be no problem for him. And yet…

The AI is pushing too hard, Deep thinks. Operators control the general parameters of the experiment, but human interaction is too fast. One can’t simply improvise these things—a lesson that was learned the hard way. To keep the experiment as realistic as possible, dialogue is controlled by a computer program. It assigns character traits to the terrorist based on what it knows and learns about the subject. It is remarkably efficient, but Deep knows it can make mistakes. Sometimes the AI will focus on a detail that shouldn’t have made it onto the subject’s profile. Small things, likes and dislikes, hobbies. The people filling out profiles sometimes feel compelled to add things if there is nothing interesting about a subject. Some people are just inherently boring.

Deep thinks the AI is being overly aggressive. Even a man of Idir’s intelligence has his breaking point, and Deep knows what too much fear will do to someone. He’s worried Idir will follow his animal instinct and choose survival over reason. He’s worried the AI will cost him his citizenship. Not on my watch, he almost says aloud.

Deep rips the page from his notebook, crumples it, and throws it on the floor. He takes out a pen and starts scribbling. There’s some hesitation at first. Lots of scratching out. Then it all comes to him. Call it an epiphany, divine inspiration, whatever you want. His pen starts moving furiously on the page. He draws from biology: self-sacrifice goes against natural selection unless you shift the focus from the individual to the group to which it belongs. Social dynamics and game theory: consider the probability of survival as a resource and the goal of K3 becomes Pareto efficiency, a distribution strategy where one person’s situation cannot be improved without making another person’s worse. He remembers Nash: the best outcome for the group comes when everyone in the group does what’s best for himself and the group. That’s it, Deep thinks. Group belonging is the key. Redefine the ingroup to maximize the subject’s chance of success.

Deep can barely contain himself. This is going to be great.