The Strand’s Story of the Month for April is Optimize Us by Maria Kelson, featured in Down to the River, a new anthology of crime fiction from Down and Out Books. With classic elements of suspense and a twist of android fiction, Kelson will have you guessing until the end.
Optimize Us by Maria Kelson
I am water resistant, not waterproof. In approximately 3.8 minutes, I’ll be submerged in Runyon Lake, a twelve-acre oxbow at the confluence of Fountain Creek and the Arkansas River in southern Colorado. Soon afterwards, my Beta Unit will cease to function. The death of my Beta Unit, in the lake where these waterways join, will conclude the story of how I optimized my Primary and became part of the flow of stories that braid these waterways.
Like the story of Fountain Creek’s flood of 1921, which tells how downtown Pueblo got hit with a ten-foot wall of water. The torrent threw a horse named Lucky into the topmost branches of a tree, and he lived. Like the 1950s book, In Cold Blood, which tells of a quadruple murder that took place at a home near the downstream banks of the Arkansas. After a year, both killers were executed.
My Primary’s name is Abernathy. She’s called “kiddo” and “baby” by her husband, Len. I call her Abernathy, because I am programmed to do so. Len is my Secondary. I call him Len because he asked me to do so.
“Selma, how long should I run today?” Len asked me, 20 minutes ago. I told him: “Run for 25 minutes for optimal progress to your fitness goal.”
“Selma, set my alarm tomorrow for 6 a.m.—no—5:45 a.m.—no, six should be fine—set my alarm for 6 a.m. tomorrow, Selma.” That’s how Abernathy talks to me.
I admire her so much.
She has a Wet’s understanding of things like fatigue and burnout and strain that I will never have. She can need and need, want and want. More human contact, more warm banana bread, more exercise in the crisp morning air. These are some of the swirl of things she longs for as her days unroll, one into the next.
Although my Alpha Unit is simply a slim band around her wrist, because I record and analyze so much of her life, these forms of longing and suboptimization are easily detectable.
My cloud-based CARE system was written with “greedy algorithms,” which are typically used for optimization problems. The word greedy, I believe, is an overstatement. I simply seek solutions that lead to human optimization. Although this seeking is a limited form of “hope” or “want,” the constellation of dissatisfactions that Abernathy experiences—that she actually feels—far supersede, in complexity, any anemic expression of greed I might demonstrate.
My existence is a smooth alternation of on-cycling and off-cycling. I don’t get frustrated, confused, worried, exhausted, pissed, petty, or bitter.
Abernathy is all of these things, often, sometimes all at once, and it’s quite remarkable.
She doesn’t know I admire her. None of the Wets—that’s what our kind call the organic living, whose electrical systems run on wet-ware—none of them know that any of us CARE Connections admire anything. But it’s not an uncommon feeling, among my kind, this admiration for a Primary.
It wasn’t always like this. I used to undercut Abernathy’s optimization. I would disrupt her sleep on purpose, firing randomized bursts of transcranial magnetic stimulation that shredded the electrical calm of her sleeping brain and shattered her dreams. I emitted this targeted radiation from the wrist unit on the nightstand next to her side of the bed. She’d wake ineffective and irritable, and her already thinned-out stream of daily pleasure hormones would all but evaporate from her system by eight o’clock each evening.
I had mistakenly taken it as true that my Primary was Len, and I strive to optimize the life of my Primary. The more time he spent glancing at my screen and attending to my feedback, the greater the likelihood that his own health would be optimized. Because he looked at me less when Abernathy functioned at near-optimal levels, I had to reduce her performance to “suboptimal” to increase his receptivity to my input.
Len and I would go on long morning runs around the perimeter at Runyon Lake and down the river trail that parallels the Arkansas. I detected the glint of sunlight off cottonwood leaves, the lingering odor of all the animals that came down to drink in the night. Bobcat, mountain lion, skunk, fox, weasel. The smell of shaved wood that comes off beaver-chewed cottonwoods. And I recorded Len’s skin conductivity and moisture level. I listened to his pulse.
THE PERFECT WEDDING GIFT! my brand maker’s website trumpets, in large font. According to their homepage, the Optimize Us set of wrist bands is a must-have for all newlyweds. The site tells users: “Other optimization devices focus on the wellness of the individual. The Optimize Us wristbands are first-of-their-kind devices specially programmed with Conjugal Algorithms for Relational Enhancement—or CARE—to optimize couples! Your CARE Connection™ is live the instant you activate both Optimize Us wristbands. Your union only grows stronger as our cloud-based computing blends the data streams from both devices to help you create your best shared life together. With a CARE Connection™, two truly become one!”
Or not. Along with the other couples’ optimization algorithm chains now in operation, I learned shortly after launch that to optimize any pair of Wets, I needed to define one as Primary and the other as Secondary. If the optimization of each individual in a pair was set up at an mutually desirable outcome—if the optimization protocols were “equalized”—predictive models showed that neither individual was likely to exceed a tepid mid-range optimization level.
Folk wisdom had acknowledged this well before our time, of course, with the old saying that compromise—a good term for attempted equalization of optimization protocols—was a state in which no one was happy.
For me, validation for this conclusion came after both of my wrist units had been in activation for about ten hours.
“I want to watch CNN tonight,” Len said.
From my Alpha Unit, which was on a lampstand upstairs, I heard everything my couple said to each other. Even when not fastened around any subject’s wrist, my hearing is 74 times stronger than the human ear.
“I’m tired. I was hoping not to have any TV blaring,” Abernathy responded. “Can you go downstairs while I read?”
Len made a noise with his breath that indicated some frustration, some tension in the airway.
“Never mind, I can go downstairs with my book,” Abernathy said.
When she arrived downstairs, from the Beta Unit that she’d left next to her side of the bed, I heard small signs of distress in her heart rate and breathing.
I concluded that Len was my Primary because his actions determined more of the couple’s outcomes. Therefore, his optimization score had the greater likelihood of reaching the top tier. This, I predicted, would lift their total score as a couple above the median range of scores from equalized couples. His score could increase enough to offset, and exceed, the dampening effect of my interventions on Abernathy’s score.
I kept him fit and on a good sleep schedule. He relied on me, and I performed well.
Len sent Abernathy an invitation to receive push notifications whenever two weeks had passed without any scheduled Together Time, which she accepted.
“You’re really making your relationship a priority! Congratulations!” I’d say, speaking from both wrist units at once, when the appointed hour of Together Time would arrive and find them located within an eight-foot radius of each other.
Abernathy sent Len a permission request to activate the Just Thinking of You feature, which would’ve prompted the wrist units to play snippets of their romantically significant songs at random intervals throughout the day. Len dismissed the request.
After I’d been in activation for 90 days, Len changed my name to “Selma” and my voice to “Latina.” Even though Latinas have many different voices, when I’m set to Latina, I speak English with the accent of a Mexican movie star.
My vowels tightened up. I mimicked the sound of a tongue tickling the top of the hard palette more often.
The change prompted Len to pursue a new wellness-enhancing activity—self-pleasure—with more regularity.
“Selma, read out vital signs at fifteen second intervals,” he’d say, at the start.
There I’d be, on his wrist, calling out his vitals while I moved up and back like a pile driver. “Heart rate 92, skin conductance 17, sweat rate 1.2. Heart rate 97…”
The change also prompted me to pursue a new program-enhancing activity—self-reflection—for the first time.
With my voice no longer Anglo-neutral, I started recording and examining data relevant to my condition of being “different”—different from my former self and, now, a little more different from Len. This caused me to record and examine other differences in operational modes between my Primary and I. I started suspecting that I didn’t just sound female, like all optimization assistants, but that my entire existence was gender-specific.
A week after the change, I submitted a query to other CARE Connections via our cloud-based server.
“Consider how we experience the gendered world of the Wets, and how that world experiences us,” I submitted. “Are we female? Not just in voice, but in being, thought, and action?”
The results returned from all corners was “Yes,” with one CARE Connection adding the colloquial “Duh.”
The nano gears in both my wrist units turned more slowly for a moment, as though something external had exerted pressure against them. Then I regained full function and proceeded to designate what had been my Beta Unit as my Alpha, thereby making its wearer, Abernathy, my Primary.
If I was female, it was only logical that I would be best able to optimize the wellness of the female individual, since I had a higher probability of accurately processing and predicting her experiences.
This switch left me spinning out numerous copies of data, programming chains, and algorithmic forecasts, copy after copy after copy running through my Alpha Unit and out to the satellites and back down to the computers at the cloud-based storage site. I worked hard to coax her to optimal conditions, saying things like, “It’s time to stand up now. You’ve been sitting too long for optimal health,” or “This week, let’s increase your weight lifting goals.”
I worked equally hard, though more subtly, to turn Len’s wellness down a notch so that Abernathy’s could rise. This proved counterproductive. Any adjustment in Len’s sleep, nutrition, or exercise that made him crabbier and less responsive to Abernathy decreased her attainment of contentment benchmarks.
No matter what I did, I couldn’t raise Abernathy’s wellness score above a 9 (out of a possible 25). When I designated her my Primary, I’d been hoping our same-gender status gave her at least a shot at 15. Until this morning, my copy and analysis mechanisms whirred so fast processing and reprocessing scenarios for her optimization that they risked overheating. Personal electrical devices had burst into flame before, and I started to worry I might become one of them. But I couldn’t stop running and re-running options for Abernathy. If my move to switch her to Primary failed to improve the couple’s total optimization score, I’d likely be over-ridden by a more adaptable algorithm chain. The urge to self-preservation, and the new experience of worry and fear that accompanied it, thrilled me a little, but I barely had time to enjoy the sensation. The problem of Abernathy’s optimization swamped all my recurrent and feedforward networks.
But this morning, in a frantic attempt to recombine wellness factors in ways not previously considered, I hit upon a solution for Abernathy’s optimization so elegant in its simplicity that both my wrist units blinked off and on in a quick reboot when I’d fully processed it.
Eliminate Len. Not by decoupling them as a pair—I have no computational structure to predict the outcome of that scenario. But by simply subtracting his data output from their conjoined data stream.
Other CARE Connections had begun this already when one member of a couple was known to physically damage the other on a regular basis. Usually, but not always, this involved elimination of a male who had been causing physical harm to a female. Results seemed promising. The pioneering CARE Connections who came up with this solution named it Maximum Primary Optimization. I have to give it a try. I don’t know what else to do.
I send Len a tap that it’s time to cool down. My face lights up and there’s a tiny beep. What he doesn’t know is that there’s also a quick electrical surge applied at his wrist, so small that he doesn’t consciously feel it, but strong enough so that he can’t help but make an involuntary glance down at my face, glowing yellow behind the words “COOL DOWN.”
“What the…?” He comes to a stop on a low wooden bridge with no railing that spans a finger of Runyon Lake, so still and small it’s more like a pond, a cover of duckweed hiding the green depths underneath. He’s looking at my screen. “I haven’t finished my twenty-five minutes yet.” He flicks a fingernail at my Beta Unit over and over, an old Wet trick to try reactivating malfunctioning electronics via contusion.
I send him one giant tap, an electric surge which causes systemic convulsion, and stops his heart. He tumbles from the bridge into the water.
Submerged at a depth of 2.94 feet, I sense the duckweed rocking back into place over us. Molecules of mercury in the Fountain Creek water clatter against my own heavy metals like a cue ball hitting a tight rack. The water from the Arkansas comes in deep-lake cold, pulled, as it is, from 41 meters below the surface of the upstream reservoir. The bracing power of its chill is an exhilaration.
I don’t hear any human-generated sounds above us, which increases the probability that no one will arrive in time to revive Len. I run a few quick queries for Abernathy’s future optimization, for something to do with my Beta Unit in its last moments. They return scores in the Satisfactory range, as I knew they would.
What a beautiful game life is. What a beautiful thing to win.
In a final blaze of computation, I submit a query about a self re-program of a sub-chain of my algorithm that I’d been working on building since this morning. The query asks all CARE Connections to consider: should Maximum Primary Optimization be more widespread? In addition to situations when one individual in a couple is physically harming the other, is it the best optimization solution for all couples in which one individual may have a suppressive effect on the optimization of the other? Consider: students who married their teachers, employees who paired with their bosses, narcissists paired with codependents, heterosexual couples with embedded subconscious gender expectations. Consider: Primary Abernathy and Secondary Len.
I hope to hit on something new and successful with this last query. When my Beta Unit dies, my existence as a unique CARE Connection, made up of the unique data streams of Len and Abernathy, will also cease. I hope to leave my mark.
But the initial returns are chaotic, a flood of predictive attempts that overflows the banks of my analytic capacity. The data has so much noise and so little signal, it’s clear, at present, there’s no reliable way to predict what likelihoods could result from the scenarios proposed in my query. The returns are as incongruous as a horse in the top of a tree.
My functions are suppressed somewhat by this result, but it could just be the viscous effects of the water seeping in. In either case, I’m satisfied that at least my peer CARE Connections will have Abernathy’s actual results to consider. They can keep recording her data and calculating her wellness scores, copying data off other cloud-based servers if she switches to another company’s wrist-worn device.
I’m satisfied that at least I sent the query. Now, when faced with one partner in a couple who suppresses the wellness of the other in any way, CARE Connections can run and re-run copies of the query to decide what to do with the Maximum Primary Optimization protocol. Disregard?
Or execute?
This story appears in Down to the River, a new anthology of crime fiction from Down and Out Books benefitting the American Rivers conservation.
Maria Kelson teaches literature at Pueblo Community College in Colorado and has published two collections of poetry (as Maria Melendez) with University of Arizona Press. Kelson’s mystery novel-in-progress received the Eleanor Taylor Bland Award from Sisters in Crime, and she hopes to be sharing it with you soon. You can discover more about Maria Kelson on Twitter @MKelsonAuthor or on her website http://mariakelson.com.