Chapter Five: Variables That Refuse to Stabilize
1049 Words
The problem with systems is not that they are cruel.
It is that they are patient.
St. Lucien did not punish me immediately after the demonstration failure. There were no public reprimands, no sudden isolation, no dramatic consequences designed to terrify the rest of us into obedience. Instead, the institution did what all confident systems do when confronted with anomaly: it observed longer, recorded deeper, and quietly adjusted its models.
They stopped calling it rebellion.
They called it instability.
I noticed the change first in the language.
Reports began using new phrases—emergent dependency, emotional interference, non-linear response. The nurses didn’t speak to us differently, but they watched differently, as if affection itself had become suspicious. As if connection were a contagion.
Liora and I were no longer assigned the same tasks.
That was deliberate.
Separation is the oldest control technique in existence. It doesn’t break people outright; it erodes them slowly, convincing them that loneliness is natural and solidarity is dangerous. The system didn’t forbid us from speaking. It didn’t have to. It merely ensured that every interaction was inconvenient, monitored, conditional.
And worst of all—it framed this as care.
“Healthy individuation,” Dr. Vantier explained during a private consultation. “Excessive reliance on peer bonding can distort developmental trajectories.”
“You mean love,” I said.
He paused, just long enough to remind me that words mattered only when power allowed them to. “Love is a subjective classification,” he replied. “We prefer measurable terms.”
That was when I understood the real danger.
They weren’t afraid of my abilities anymore.
They were afraid of what I felt while using them.
The next experiment confirmed it.
They placed me in a cooperative task—with someone else.
Not Liora.
A boy named Aron. Cognitive amplifier. Brilliant, compliant, careful in the way people are careful when they believe survival depends on obedience. He smiled politely, avoided eye contact, followed instructions perfectly.
We were instructed to synchronize.
My role was to manipulate spatial variables; his was to calculate probability pathways in real time. Together, we were supposed to demonstrate “optimal system harmony.”
It worked.
Too well.
The structure responded instantly. Sensors stabilized. Data streams aligned. The observers behind the glass nodded with satisfaction.
I felt nothing.
No resistance. No friction. No urgency.
And that terrified me.
Because the system loved us together.
Afterward, Aron thanked me. Genuinely. “You’re efficient,” he said. “They’ll be pleased.”
“They already are,” I replied.
That night, I found Liora in the stairwell—the only place cameras still struggled with angles and shadows.
“They paired me with someone else,” she said immediately.
“So did they,” I answered.
We stood there in the half-dark, not touching, not trusting the walls.
“They’re testing substitution,” she whispered. “Seeing if attachment is replaceable.”
I nodded. “And if it isn’t?”
Her jaw tightened. “Then we become a problem.”
The answer came sooner than expected.
Two days later, during routine monitoring, they declared Liora a destabilizing influence.
Not publicly. Quietly. Clinically.
She was removed from general circulation “for recalibration.”
No one told me where she went.
That was the moment the system made its mistake.
Because until then, I had been learning restraint.
Afterward, I learned refusal.
The first sign was small. Sensors in my wing began reporting contradictory spatial data. Objects were where they were supposed to be—and also not. Doors delayed opening by fractions of seconds. Lights flickered only when certain people passed beneath them.
Maintenance blamed calibration errors.
I blamed grief.
The second sign was harder to ignore.
During a high-level evaluation with external representatives, the room failed to maintain structural alignment. Not collapse—nothing so dramatic—but misbehavior. Chairs resisted movement. Glass panels resonated unpredictably. Sound arrived a beat too late.
The observers shifted uneasily.
“Subject 14,” the woman in tailored clothes said coolly, “explain the variance.”
“I don’t know,” I replied honestly. “Maybe the system is under stress.”
Her smile was thin. “Systems don’t feel stress.”
“No,” I said. “People do.”
They isolated me after that.
Real isolation this time. No shared air. No routine. No Liora.
The room was designed to suppress—not my power, but my context. Blank walls. No stimuli. A place where nothing could be interrupted because nothing happened.
They thought deprivation would stabilize me.
Instead, it clarified everything.
Without objects to move, without people to react to, I felt the system itself more clearly than ever. Its rhythms. Its redundancies. Its assumptions.
It assumed compliance was the default.
It assumed rebellion required spectacle.
It assumed love could be redirected.
It was wrong.
When they returned me to testing, I didn’t resist.
I misaligned.
Commands were followed—but not in the order expected. Variables responded—but not to the dominant signal. The system recorded success while experiencing failure.
That kind of contradiction is lethal to institutions.
Eventually, Dr. Vantier asked the question they had been avoiding.
“What do you want?” he said.
I thought of Liora. Of the archive. Of the children labeled non-viable.
“I want you to stop pretending this is care,” I said. “And admit it’s production.”
His expression hardened. “You are not in a position to negotiate.”
“No,” I agreed. “But you are in a position to lose control.”
They moved Liora again after that.
Not to punish her.
To protect the model.
Because what frightened them most was not my power—it was the evidence that connection made it unmanageable.
I don’t know where she is now.
But I know this:
They have begun redesigning the program.
And programs only change when something threatens to break them.
I lie awake at night, counting the seconds between system pulses, listening for irregularities. The orphanage hums differently now. Less confident. More alert.
They have classified me as an unstable variable.
They’re right.
And instability, I am learning, is not weakness.
It is potential that refuses to behave.
- Writer
