22. Algorithm
Alexander
The rumor has no identifiable source. That’s what makes it dangerous. Because it can’t be confronted head-on.
It emerges on social media like a collective whisper, a suspicion shared without an author: MatchMapping allegedly doesn’t present certain profiles. A narrative without a face.
No one accuses anyone directly.
No one cites a specific document.
And yet, the same words keep resurfacing, again and again.
Bias.
Silent exclusion.
Elitist algorithm.
A few weeks before the IPO, it’s a disaster. The worst possible timing.
Internally, the teams talk about a “perception crisis.” Investors want reassurance. Legal insists there is no formal proof linking the company to intentional discrimination. Everything is technically defensible.
But the numbers tell a different story.
Atypical profiles—nonlinear career paths, niche interests, fragmented social networks—systematically receive fewer recommendations. Not because they’re rejected outright, but because of an accumulation of reasonable criteria: too few shared data points to justify the risk. Death by optimization.
The algorithm is only doing what it was taught to do: avoid uncertainty. Reduce noise. Maximize predictability.
Suddenly, the memory of Kristel slams into me.
That conversation was about two months ago.
“We’re over-optimizing,” I had confided that night. “Profiles that fall outside the norm are slowly disappearing. They don’t fit into the boxes.”
She had smiled, leaning against the couch, a glass of wine in hand.
“That’s the price of efficiency, Alexander. You can’t please everyone. The investors are satisfied—that’s what matters.”
“It’s not just about numbers,” I had argued.
“Why question something that works?”
At the time, I thought she simply hadn’t listened to me.
Today, I’m no longer so sure. She had listened. Closely.
There’s no proof linking her to the rumors. But the timing, the precision of the terms used, certain phrasings—they’re eerily close to what I confided in her that night.
And above all, I remember her icy stare at the Speakeasy, when she swore she’d get her revenge. A promise, not a threat.
The doubt is there. Persistent.
Around two p.m., Nina walks in without knocking, as she always does when she decides I’ve crossed the line of reasonable behavior. She drops a takeout container on my desk with an exaggerated sigh.
“Take a break. Starving yourself won’t solve the problem any faster.”
I know she’s right. I eat distractedly, eyes glued to my screen. I want to call Sam, to talk to her, but I’m afraid of dragging her down with my foul mood, and some conversations don’t belong on the phone. I saw her messages. I’ll call her tonight. Or better yet, I’ll go see her if I can get away.
Meetings stack up all afternoon. Legal. Data. Communications. Everyone speaks in turn, cautiously, as if the slightest word could make things worse.
“Technically, we’re compliant,” legal concludes.
“Morally, it’s murkier,” someone else replies.
By the end of the day, we still don’t have a satisfying solution. Just a fragile consensus: we need to regain control of the narrative before it controls us.
I end up approving a first statement—short, measured.
“Like any large-scale data processing system, our algorithms may exhibit biases. We are actively working to identify, measure, and correct them.”
It’s defensible. It’s not reassuring. And everyone knows it.
Days pass without me touching ground. I collapse from exhaustion onto my bed and sink into a restless sleep. I haven’t seen Sam in days.
*
It’s seven p.m. on the third day, and I’m still at my desk when Nina pokes her head through the door.
“I’m heading out. You have a visitor.”
Surprised, I look up.
Sam is standing in the doorway, hesitant.
I get up immediately, my heart leaping in my chest.
“Thank you, Tina. Have a good evening.”
As soon as the door closes, I pull Sam into my arms and kiss her full on the mouth. A hungry, desperate kiss. God, how I missed her. I hadn’t realized how much—until she was standing right in front of me.
She laughs softly against my lips.
“And here I was, afraid of bothering you.”
“Never,” I murmur, holding her tighter.
She pulls back slightly, studying me as if gauging my sincerity. Her smile fades.
“Why didn’t you tell me anything? Mara’s the one who told us about the rumors.”
I sigh and run a hand through my hair.
“I didn’t want to burden you with my problems.” I thought distance would protect you.
Sam’s expression closes off. She slips from my arms and walks toward the window, her back to me.
“Except you still burdened me, just without meaning to.”
Her voice is calm, but I can hear the hurt beneath it.
“I knew something was wrong,” she continues. “I just didn’t know what. I thought you regretted… us. That you needed space. That I was too…”
Her voice breaks slightly.
My heart tightens. Suddenly, I understand all the scenarios that must have played out in her mind during those days of silence.
“Sam, I’m sorry. I’m really sorry.”
She turns back to me, eyes shining.
“We’re together, Alexander. That means I want to be there when things go wrong. Not just for the good moments.”
I step toward her and place my hands on her shoulders.
“You’re right. I was wrong. I thought I was protecting you.”
“I’m not fragile,” she says softly. “I can handle your bad days. But I can’t be kept at a distance.”
I pull her into my arms and bury my face in her hair.
“I hear you. I promise I’ll do better next time.” I mean it.
She stays against me for a moment, then pulls back slightly and looks me straight in the eyes.
“Good. Now you’re coming with me. You look like a zombie.”
“I still have work—”
“Which can wait until tomorrow,” she cuts in. “Come on.”
She drags me out of the office, orders takeout on her phone, and once we’re back at my place, literally pushes me toward the bathroom.
“Shower. Now.”
“Come with me,” I say with a tired smile. “Just for the shower. I’m too drained for anything else.”
She agrees with a sideways smile.
But the little menace has other ideas. Under the hot water, her hands slide over my skin—soothing at first, then provocative. She does exactly what’s needed to clear my head of everything else. Methodically. Effectively.
When we finally return from the bathroom much later, dressed in comfortable clothes, Sam reheats the food she ordered. We eat in a comfortable silence.
After dinner, she curls up in my arms on the couch. Her head rests against my shoulder, her fingers absentmindedly playing with the collar of my robe.
“Tell me,” she says softly.
So I tell her everything. The rumor. The IPO. The algorithmic bias. My suspicions about Kristel. The fear of watching everything collapse.
She listens without interrupting, fully attentive. When I finish, she stays quiet for a moment. Then she asks:
“Do you know why I love watercolor?”
I shake my head, surprised by the change of subject.
“Because there’s an almost magical alchemy between water, light, and pigment. If you try to control everything… you lose the light. The magic.”
She straightens slightly to look at me.
“The most striking example for me is Turner’s work. His oil paintings are magnificent, perfectly mastered. But it’s his watercolors—especially those of Venice—that truly move me. Because he lets go. He lets the water decide. He trusts the accident.”
I frown, trying to understand where she’s going.
“What matters,” she continues, “isn’t the artwork itself. It’s the door it opens. The emotion it triggers. The unexpected thing it reveals.”
She places her hand on my cheek.
“A well-calibrated recommendation has a good chance of working, that’s true. But if you suggest more atypical profiles to open-minded people… you might create magic. Encounters that should never have happened on paper, but are exactly what those people needed.”
“It’s a beautiful idea in theory,” I say slowly. “But it’s risky.”
She doesn’t look away.
“Everything worth it is.”
Her words hit me like an electric shock. Because she’s right. And I know it.
I look at her—this incredible woman who just summed up, in three sentences, what my data scientists and I have been struggling with for days. Without jargon. Without fear.
“We’ve stacked too many rational criteria,” I murmur, the thought taking shape as I speak. “Interests, social circles, similar trajectories. Everything that ensures a predictive model. Everything that excludes the unexpected.”
“Exactly,” she says softly.
“The solution isn’t to tweak a parameter. It’s to change the promise. The philosophy.”
She smiles, and I pull her against me, kissing her with a gratitude and tenderness that overwhelm me.
“You’re brilliant,” I say. “You know that?”
She bursts out laughing.
*
The very next morning, the entire team gets back to work with a new direction. Together, we imagine a different recommendation layer. Less measured compatibility, more potential resonance. A controlled element of chance. A deliberate dilution of dominant criteria. A crack in the system. On purpose.
A watercolor in a world of technical drawings.
From now on, we communicate differently. Not about performance, but about openness. Not about certainty, but about experience. MatchMapping no longer promises the best choice. It offers a possible encounter. An invitation, not a verdict.
And every night, I call Sam before going to sleep. We don’t stay on the phone for long. I tell her about my days, she tells me about hers. But beyond words, we connect with each other.
And like water and pigments, we learn together to create light.
Slowly. Deliberately.