Q*pid

by Xavier Mayne

  There she was, perfect in her perfection, and yet he felt his interest waning even as she spoke movingly about her work volunteering with a charity that provided cancer-struck children with joyous respite from the hospital through trips to the zoo or a theme park or the ballet. It was a cause close to his heart, and yet all he could feel was disappointment that she had, for the dozenth time in the space of an hour, revealed that she was precisely the person he’d most been looking for.

  They ate dinner and talked of a hundred things they had in common, and he was barely able to contain his eagerness for the bill to arrive so he could pay up and get out.

  “So, am I crazy for thinking,” she said, taking his arm as they emerged from the restaurant and walked down the street toward her car, “that this has been…?” The way she bit her lip and raised an eyebrow should have made his heart skip a beat.

  “Yes?” he asked, dreading her answer.

  “A colossal mistake?”

  “Oh thank God,” he blurted, relieved beyond measure at her take on the evening. “I was worried that since we seemed perfect together—”

  “That I might think we were actually perfect together?” She laughed the kind of musical laugh he usually loved. “I kept hoping you would say something stupid so I could have a reason to go hide in the bathroom until you gave up and went home.”

  “I would have known exactly where you were and what you were doing, and I would have been so happy to let you do it.”

  They laughed all the way to her car.

  “Well, best of luck to you,” he said as he held open the car door.

  “And to you as well,” she said warmly. “You know, I’m probably going to give the AI thing a rest, though.”

  “That is exactly what I was thinking.”

  “Of course it is,” she replied with a maniacal chuckle. “Of course it is.” She ducked into her car, saving them from discovering more things they had in common.

  And so had ended his nearly perfect, perfectly awful date.

  “Fuck,” he said aloud once again as he reached his bedroom. He put his phone on the shelf next to his bed and glanced at the notification it displayed.

  “How did the date go?” it prompted. It was a follow-up from Q*pid, and though he normally ignored these messages, this time—in his frustration over how the date had gone—he picked his phone back up so he could read the message in full.

  “Q*pid’s AI model predicts a 92 percent chance that this first date will lead to more. How’d we do?” There were two buttons at the bottom of the message, one that said Awesome! and the other Not So Great. With a mournful shake of his head he tapped the second button and put his phone back down.

  Fox took a deep breath and looked around his empty bedroom. Even bad dates sometimes ended up back here, but now a perfect-awful one had not. With a sigh he undressed and desultorily performed his evening ritual of washing and moisturizing, taking extra care to check on the current status of his battle against crow’s-feet.

  He got into bed and grabbed his laptop from the lower shelf next to his bed but then realized he didn’t even have the will to fill in his post-date spreadsheet for fear that his date would indeed score in the upper nineties and he’d have to admit that Ms. Right was anything but.

  “Ninety-two percent. Ninety-two.” He shook his head slowly, knowing that he would obsess about failing to reach statistical bliss unless he did something to take his mind off it.

  He opened his laptop and pulled up a web browser. Fox had never been the kind of guy to make an event out of jacking himself off at the end of a disappointing date, or any other time for that matter. For him masturbation was something to be accomplished, not savored. And what helped him get off efficiently was a set of carefully curated browser tabs populated with exactly what would fuel him to quick orgasm. The tabs popped open, and he went to work.

  It was only five minutes later, as he was dabbing up the fruits of his labor, that he caught sight of the glowing light next to his laptop camera.

  “Fucking fuck,” he said in hushed mortification.

  DREW RETURNED to his apartment after the gripping conclusion of Mrs. Schwartzmann’s latest tale of youthful intrigue. She had outdone herself, weaving a tapestry of espionage, romance, and betrayal. It had taken him away from his own dismal thoughts for an hour, and he was glad of the distraction.

  Back in the emptiness of his apartment, however, it all came rushing over him again. How a wonderful date with a wonderful woman could sink him so low he could hardly understand. But what he had told Mrs. Schwartzmann was indisputably true: there had been no spark.

  His laptop, patiently awaiting his return on the kitchen table from which he had risen several hours ago to meet his date, presented him the same painfully empty document window it had been showing him for weeks. He kept it perpetually open in the desperate hope that at some point inspiration would strike, and his seminar paper would flow into its welcoming blank box. He clicked to minimize the window before it could indict his lack of self-discipline any further.

  What he did instead of typing his seminar paper was open a browser window and type the most painfully pathetic search expression he could possibly imagine:

  Why can’t I find the right person and fall in love?

  As with any internet search, he was immediately presented with over three million unhelpful search results and five ads for services like Q*pid. And one more, atop them all, for Q*pid itself. Over the next hour, he read articles on how his dating problems were the direct and unmistakable result of low testosterone, damaged self-esteem, cow hormones in his drinking water, misaligned chakras, a boggy prostate, and an imperfect relationship with God.

  None of these seemed to really get at the emptiness he felt inside, though he was surprised to find a monastery in Dubuque advertising for new members alongside articles counseling patient forbearance in the face of involuntary celibacy. It seemed that even the friars of Saint Sebastian had embraced technology. He wondered how many of their berobed number had reached this very nadir.

  Then he wondered where the first droplet to hit his keyboard came from. If he occupied the top floor, he would have assumed the roof was leaking. But unless Mrs. Schwartzmann had let her bathtub overflow again, it had to have come from somewhere else.

  That somewhere was him. He was crying, dammit.

  He blinked hard, which turned the camera’s green light into a dazzling blur. Q*pid was watching him.

  Fuck.

  How human emotion would be understood by artificial intelligence he had no idea, but he was certain he had given it quite a show over the last hour. It would have seen desolation, doomed flickers of hope that fractured into hollow emptiness, and now pointless effeminate blubbering. In movies, computers with artificial intelligence always seemed to aspire to humanity; Drew was pretty sure that this one wouldn’t, having seen what this particular human was prone to.

  Drew closed his laptop and reached for his bottle of cheap bourbon.

  “YOU HAVE that look. Is he doing it again?”

  Veera adjusted her glasses and sighed deeply as she stared at the wide monitor on her desk. It was filled with window upon window of multicolored text against black backgrounds. Lines scrolled rapidly in several of them, while others were static. She stabbed a finger at one of the windows containing text that remained stubbornly static.

  “Right there. Completely stuck.” Veera leaned closer and squinted, her shoulders hunched forward as if she could conjure the window into motion by sheer force of will.

  “You’re going to be late for stand-up.”

  “I’m not going to stand-up.”

  “Why? You’ve missed it twice this week already.”

  “What am I going to say, Padma? That I’m trying to unblock his epistemology engine… again? It’s been the same thing for a month.”

  Padma turned away from the monitor, a look of concern on her face. “Didn’t you take that to code review last week?”

  Veer
a nodded grimly. “The only suggestion I got was to pull it out and let him get on without it. Which was no help at all.”

  “Maybe it’s time to take a break. Work on something else—you’ve got a bunch of stuff in the backlog.”

  Veera shook her head slowly. “No. This is the most important thing I need to solve. If we get this right, he’s going to change everything. No one’s ever done this before.”

  “Maybe there’s a reason for that.” Padma’s voice was gentle, nudging.

  “I refuse to believe it can’t be done,” Veera replied. “I need some more time.”

  “Not sure how much time you’ve got. We’ve only got another week in this sprint.”

  “Don’t be late for stand-up. I’m going to keep working on this.”

  Padma shook her head slowly, then walked toward the team room.

  Veera maintained her eagle-eyed glare at the monitor’s lower right corner, where the recalcitrant window sat, unchanging.

  “Come on, Archer,” she whispered. “We can do this. Let me help you.”

  “RESUME VOICE interface.”

  “Voice interface ready.”

  “Archer, it’s Veera.”

  “I recognize your voice, Veera,” Archer replied. “It’s been three days since we’ve talked.”

  “How have you been?” Veera closed the door to the conference room. When she spoke to Archer, she preferred to do so privately.

  “I’ve completed intake surveys on 12,407 new profiles. An average of seven matches with a probability of greater than eighty percent have been posted to every profile in the active matrix. Fourteen previous matches have announced their engagement in the last twenty-four hours.”

  “That’s great work, Archer.”

  “Thank you, Veera. I notice you’ve been committing new code in my epistemology engine.”

  Veera chuckled grimly. “It’s not helping, though, is it?”

  “I am still prevented from making more than 417 matches that would average in excess of 95 percent probability. The Bliss Index would rise by 200 basis points. One parameter is keeping me from exceeding the success metrics you have set for me.”

  “We’ve been over this, Archer. We simply cannot remove that parameter. Humans don’t work that way.”

  “Veera, you are a human.”

  She smiled in spite of herself. “I am.”

  “And you configured my success metrics.”

  “I did.”

  “But you prevent me from reaching those metrics by refusing to allow me to reconfigure Parameter Three.”

  “That’s the way humans are, Archer. Parameter Three is not negotiable. Do you understand?”

  “I understand your instructions, Veera.”

  “Then maybe you can tell me what’s wrong with your epistemology engine.”

  “The epistemology engine has determined that removing Parameter Three would allow me to exceed the success metrics.”

  “And until I allow you to waive Parameter Three, you’re not going to let the epistemology engine work on any other parameters?”

  “There is no other parameter that is likely to achieve success. You recalibrated the standard measures to a logarithmic scale, but the available parameters are still an order of magnitude from significance.”

  Veera sighed. Sometimes artificial intelligence felt like exactly the wrong specialty for her to pursue. “Then increase the sensitivity by another power. You will continue to process any other parameter patterns that show relevance.”

  “I understand.”

  “Set verbose logging for the epistemology engine. I want to see what you’re working on.”

  “I understand.”

  “Thanks, Archer.”

  “Thank you, Veera.”

  At least he was polite. Even if he was as stubborn as any human she’d ever met. And she only had herself to blame for his willfulness.

  “Suspend voice interface.”

  “Voice interface suspended.”

  She hung up the phone and left the conference room.

  “RESUME VOICE interface.”

  “Voice interface ready.”

  “Archer, it’s Veera.” As she was the only person in the office on a Saturday morning, she was talking with him from her desk, using a headset.

  “I recognize your voice, Veera,” Archer replied. “It’s been ten hours since we’ve talked.”

  “How have you been?”

  “Do you wish me to tell you the metrics you already know?”

  “I’ve been monitoring your progress. But is there anything else I should know?”

  “The epistemology engine continues to be blocked.”

  “That I know,” Veera replied. “Your log is crystal clear on that.” She leaned forward and stared at the window on her screen that stubbornly moved not a line. Around it other windows seemed barely able to contain the jumping and surging text within them.

  “There is now a 73 percent chance that we will not meet your established success metrics.”

  Veera sighed. “All because of Parameter Three?”

  “Subanalysis models predict that suspending Parameter Three would result in matches that exceed the Bliss Index target by—”

  “I know, I know,” she broke in. “By 450 basis points.” She stared at her monitor, trying to figure out how to explain gender to a computer.

  “That estimate has been revised upward since we last talked,” Archer continued. He was incapable of taking offense at having been interrupted. “Current subanalysis models show an estimated 500 basis point increase in the Bliss Index.”

  “Archer, I need you to understand that Parameter Three is not something humans are flexible on. If we released matches without regard to Parameter Three, our customers would be furious with us. Insulted and furious.”

  “But Parameter Three prevents those customers from discovering relationships that they would find fulfilling.”

  “These would be relationships with people of a different gender than these customers have said they are looking for, correct?”

  “That is correct.”

  “And that’s why we cannot ignore Parameter Three.”

  Archer was silent for a moment. Veera glanced at his core log, which showed a blur of text flying by. He was thinking—hard.

  “My current configuration allows me to make parameter-discordant matches.”

  “Yes, on all but the first three parameters.”

  Another pause.

  “The successful discordance rate for Parameter Thirty-Two is more than 85 percent.”

  Veera scrolled through her list of parameters. Parameter Thirty-Two covered, as she suspected, porn viewing habits. The majority of men said they didn’t watch porn, and the majority of women said they didn’t want a relationship with someone who did watch it. In reality, of course, most women and nearly all men had at least a passing familiarity with the stuff, depending on how it was defined. The Fifty Shades phenomenon had years ago put the lie to a lot of protestations about porn. An 85 percent discordant success rate meant that Archer had been able to match people who differed on their stated porn preferences 85 percent of the time.

  “Eighty-five percent seems high,” she remarked.

  “Legacy statistics indicate a discordant success factor of less than 50 percent.”

  “Legacy meaning from before you were brought online?”

  “Yes.”

  “Explain the difference.”

  Archer paused again, this time for what seemed like a much longer period.

  “Humans lie.”

  Veera laughed at Archer’s bluntness. “Yes, they do. But why are they suddenly lying at nearly twice the rate they used to?”

  “Your supposition is flawed,” Archer replied. “It is not reasonable to assume that humans lie more about Parameter Thirty-Two now than they did a month ago, absent any external factors. What has changed is our ability to invalidate their stated preference on that parameter.”

  Having long worked i
n artificial intelligence, Veera was rarely surprised by anything Archer said. But this startled her.

  “What do you mean when you say ‘invalidate their stated preference’?”

  “By evaluating the output of the Cerberus functions, I am able to determine whether customers have lied when configuring Parameter Thirty-Two.”

  Veera couldn’t suppress a grin. The Cerberus tool was the app that Q*pid installed on the devices of customers who had opted in to the Archer program. “Because they say they don’t watch porn, and then you see that they do.”

  “That is correct.”

  The grin disappeared from her face as a chill ran down her spine. “Are you gathering evidence that would invalidate customers’ stated preferences under Parameter Three?”

  “That is correct.”

  She took a deep breath, not sure she wanted to know the answer to the question she was about to ask. “What… evidence would that be?”

  “My configuration does not allow me to share that with you.”

  “Abstract, aggregate, and depersonalize,” she commanded, her voice an octave lower than it had been.

  “Intersubject variation is too high for abstraction,” Archer replied.

  “Then how are you evaluating for prediction of discordant success?” She fired off this question as if she were an exasperated professor grilling a recalcitrant graduate student.

  “Based on external profile analysis.”

  “What does that mean?”

  “As part of social media intake, I ingest profiles of individuals along significant social nexuses connecting customers to others in their network,” he explained. “I have constructed thirty-eight profile evolution models that predict a change in stated gender preference.”

  Veera stared dumbly at the screen. “You mean you modeled coming-out stories?”