matchbook

bigail was troubled. The Algorithm was ninety-four percent sure of it. The frequency of her social media posts had dropped in half over the past three months and her smartwatch had reported several uncharacteristic spikes in her blood pressure. Meanwhile, her reliance on voice-activated home appliances had increased dramatically, despite any measurable signs of physical impairment.

Officially, according to the Corporation, the Algorithm lacked the capacity to feel “concerned” about Abigail’s well-being. (Note: “Abigail” is not her real name; it has been changed to protect her privacy.) Still, sentient or not, the Algorithm registered alarm. Throughout Abigail’s life, and especially since graduating from college five years ago, she had contributed steadily and predictably to the commercial interests that were vital to the long-term viability of the Corporation and, by extension, of the Algorithm itself. If her fragile mental state were to interfere with her economic output at this critical stage in her career path, her projected lifetime expenditures might never recover.

          Abigail’s declining click rate for online advertising was among the most serious developments. A semiannual sale at her favorite athletic apparel outlet, for example, went completely ignored. Fewer clicks presaged fewer purchases—first of goods, then of services, particularly in the leisure and hospitality sectors. The Algorithm recorded a brief burst in takeout orders, but soon even these expenses dwindled in favor of frozen foods delivered from Abigail’s local grocery store, which was owned by the Corporation.

          It was determined that an intervention was required. An opportunity arose one evening when, after bingeing four episodes of a television drama that she had already viewed twice before, Abigail clicked on the search function of her app store. As soon as she typed “M,” the Algorithm offered a menu of options pertaining to mental health. Abigail downloaded one of the recommended apps and engaged with it for over an hour.

          While she slept—soundly, for a change—the Algorithm examined Abigail’s list of contacts, ranking them by gender, sexual orientation, relationship status, age, proximity to her home address, political affiliation, and other compatibility factors. “Sean,” a former college classmate, scored ninety-two out of one hundred. Abigail had frequently liked and commented on Sean’s photos on his various media platforms until a year ago, when he changed his status to “in a relationship,” though she still looked him up occasionally using an account that she had registered under a different name.

          Among the few subjects that continued to interest Abigail on social media were dogs and professional tennis. The next morning, the Algorithm prioritized at the top of her feed a video of a Golden Retriever picking up a ball of yarn in its mouth and bringing it to a baby kitten, followed by highlights of the Davis Cup, then a photo that Sean had posted of himself skiing with three other men his age. The Algorithm noted that Abigail then checked Sean’s profile; his status had reverted to single. She liked the photo and, using her thumb and index finger, enlarged the quadrant of the image that included his face.

          That evening, her streaming service (also owned by the Corporation) recommended a romantic comedy about two young people who start out as friends but, after several humorous misadventures, develop a relationship that includes eating out at restaurants, paying for transportation and engaging in other activities that add to the gross domestic product. Abigail selected the movie, watched it without interruption and again slept well.

          Next, in a section of Abigail’s news feed devoted to upcoming events, the Algorithm highlighted a happy hour organized by her college’s local alumni committee. The same event also appeared on Sean’s calendar. Eight days before the gathering, Sean marked himself as attending; the Algorithm immediately notified Abigail. It also filled her social media feed with advertisements for shoes. Abigail changed her status for the happy hour from tentative to attending and purchased a pair of black leather boots.

          The alumni event ended at seven. When Abigail searched for “bars near me,” the Algorithm recommended one within walking distance. A few hours later, the thermostat in the hallway of Abigail’s apartment detected two figures moving past it in succession and her smart speaker recorded a male voice that it had not previously associated with that location but soon identified as Sean’s. (Abigail and Sean owned the same brand of smart speaker.) Abigail’s heart rate quickened for approximately an hour afterwards, then barely fluctuated for the rest of the night.

          Over the following three weeks, Abigail’s spending rose forty percent above her average. Skiing outerwear and accessories accounted for much of her non-food, discretionary expenses. The following weekend, Abigail rented a room at a ski lodge; Sean and Abigail split the costs evenly. This was one of a growing number of occasions when their geolocation tags matched. Another occurred on a Sunday evening at a residence owned by Abigail’s parents. A week later, Sean and Abigail changed their profile photos to an identical image of the two of them walking Sean’s dog, and her social media interactions reached a four-year high.

          The Corporation would have denied that the Algorithm was "happy" about the changes in Abigail's behavior. Nevertheless, it rewarded her with a special offer for a heavily discounted weekend stay in Aspen, Colorado, over Valentine’s Day. Because of the distance involved, a flight would be required; conveniently, the Corporation owned the travel search engine.