The research wing smelled faintly of antiseptic, solder, and fried instant noodles. Takumi liked that smell — it meant things were being made, fixed, improved. The lab he'd carved out of an old maintenance depot had become a cathedral of tinkering: racks of quantum drives, a bank of diagnostic holo-displays, and a soft play area with plush toys that doubled as neural-input testers for the kids.
Muroto Kin hovered over a slab of molecular models, hair perpetually mussed, eyes bright with overcaffeinated glee. The telekinesis gifted by the serum let her rearrange atoms in virtual space with a flick of her wrist and a mutter. She was the kind of scientist who would stay up three days straight, hate her own lack of sleep, and then be elated because she'd discovered something elegant in the 72nd hour.
Takumi watched her for a moment, then turned to the three girls — Lan Yuan Yuju, Bushi Cui, and Tina — who were still bubble-eyed from being whisked in by the small transport drones.
"Have a seat," he said gently. "We need a few samples so we can test a treatment. It's nothing scary. Think of it as a health check."
Yuju hopped up on the stool with the overconfidence of a child who already trusted Takumi like a sun. "Do we get cookies?" she asked, because children asked the important questions first.
"Of course," Takumi answered, conjuring a small pack of cookies, placing them beside the diagnostic scanner. The robot nurse smiled politely. "One cookie after the sample. Brave girls get bravery cookies."
Bushi Cui fiddled with the ID bracelet on her wrist. "Are we really going to fix the sick parts? Will I… not feel weird anymore?"
Muroto sat down, rubbed her temples as though to center herself, and leaned forward. "We'll study your DNA, yes. The Gastrea infection in your world resulted from a failed gene therapy. The virus is effectively a broken script. We'll complete and correct the script so cells stop producing the… undesirable phenotype. But —" she glanced at Takumi, "— the real work is subtle. It's not magic. It's careful editing and rehabilitation. You'll be closely monitored. You won't be guinea pigs."
Tina, usually taciturn, crossed her arms. "Then do it fast."
Takumi felt a small warmth in his chest. He was not made of empathy and yet, this fragile, noisy, starlit group of children had become his world.
Herrscher of Reason at Work
Takumi closed his eyes and let the Herrscher of Reason hum in the background of his mind. Creation by cognition. Construction through thought. He pulled blueprint-knowledge from his Imaginary Space — what his avatar called "infinite dimensional storage," but in practice looked like a calm ocean of files and half-completed latticeworks.
He did not "operate" on the children with his hands. Instead he built a cognitive scaffold — a virtual Petri net inside the lab's holo-auras that could model genome edits in real-time. He instructed the system in plain language, and the Herrscher translated thought into structure: error-correction subroutines, protein folding predictors, and safe delivery vectors. Around them, the lab responded — robotic arms moved to prepare microtubes, a clinical AI harmonized sample labeling, and the room's lights dimmed to a soft, non-threatening blue.
Muroto guided the phlebotomist drone; it drew a few milliliters of warm, trusting blood. The vials slid across a tray; the scanners hummed. Data poured into Takumi's construct like rainfall. He saw sequences, broken codons out of step, the ghost of a viral scaffold misfiring — and then he sketched a correction, a completion of the sequence that turned a destructive loop into a harmless protein regulator.
This is where the Herrscher decision mattered: Takumi did not rewrite randomness. He applied reasoning, constraint, and a principle he'd learned the hard way — designs that work must be explainable. Whatever he created had to be reconstructable by material means. He would hand humanity tools they could understand, not an inscrutable miracle. That, oddly, settled his conscience.
As the scaffold validated each edit, Muroto's telekinetic interface assembled a delivery vector — an engineered nanoparticle with a harmless viral shell and a benign payload of corrected mRNA — modeled to seek specific stem cell niches. The AI simulated immune response fluctuation, flagged potential red lights, and Takumi adjusted cognitive parameters until the risk sat comfortably below the threshold for ethical clinical trials in any sane world.
"Ready?" Muroto asked.
Takumi looked at the three girls who were now playing gentle games with a robot that could morph into foam animals. Their laughter punctured the lab's clinical air.
"Ready," he said.
Slice-of-Life Broadcast: Group Chat Lives
Takumi, normally inscrutable, had allowed a secure, read-only livestream for the group chat — a carefully curated view, designed to reassure friends and administrators, not to spectacle the children.
The chat exploded.
Fujiwara Chika:
[OMG TAKUMI-ONIICHAN LIVE?? I brought snacks!!!] 🍪
Sagiri:
[Please be careful, Takumi-oniichan!! Don't do risky stuff!!] 😣
Bronya:
[Monitoring. Noninterference required. Keeping an eye.]\
Himeko:
[If anything goes sideways I'll personally blow it up (in a good way).]\
Zhongli:
[It would be prudent to ensure informed consent protocols are followed.]\
Takumi (replying):
[Everything is within ethics. Muroto-san is certified. No one is forced.]\
Megumi Kato (quietly):
[…If they consent, it's fine. I'm watching too.]
A cascade of hearts and worried emojis filled the feed. The group's personalities—Chika bubbly, Sagiri anxious, Bronya practical, Himeko theatrical, Zhongli measured—provided the emotional punctuation to the scene. Takumi found the banter oddly comforting.
The Procedure (Fictionalized / Non-Technical)
Muroto and Takumi administered the corrected vector in microdoses. It was a slow, careful procedure; not one injection that promised immediate miracles, but a phased therapy with checkpoints.
For each child:
Baseline physiology measured (immune markers, organ function, cognitive status).
Low-dose vector administered to a narrow target group of stem cells.
Close monitoring for immune reaction; supportive care prepped.
If all stable, escalate dosage under supervised protocols.
In the fiction-safe language Takumi and Muroto used with parents and the chat, the therapy was a "corrective regulation" — a completion of a broken code.
The first twelve hours produced nothing dramatic: slight redness, a spike in cytokine markers (expected and mild), sleep and a tendency to dream of floating toys. The AI flagged everything within predicted parameters. Takumi slept at the lab overnight — eyes open to streams of data when needed, asleep to the soft breathing of the children in adjacent rooms.
At dawn, the first sign arrived: Tina, who had had chronic fatigue turned radiant and jumped up to chase a robot by her bed. She ran, laughed, and a tiny cheer rose in the clinical ward.
Muroto clapped once, delighted, exhausted, eyes wet. Takumi felt something like relief, and not the cold kind — actual, human relief.
Psychological Effect on Takumi
The achievement bolstered him, of course. But more insidiously, it did something quieter: it normalized control. Takumi could save lives with cognition. He could architect reality so people didn't have to flee broken worlds. The power was intoxicating — a slow, dangerous sweetness.
He watched the children play and felt the soft knot in his chest. For all his rhetoric about building civilizations and collecting people, this was personal. He had been alone on a ruined Earth — the idea that he could create refuge, and mend broken biology, made him both monstrous and saintly in his own head.
He pushed the dark thought away. Practicality first, he told himself. Make a template, run trials, scale ethically. The Herrscher's precise logic answered back with blue-light certainty: constructs must be iterated, documented, verifiable. Do it that way and you stay human.
Aftermath: Small Joys, Big Planning
By evening, the preliminary report was clear: the corrected construct functioned as intended in the trial cohort. Improvements were measurable but incremental — cell markers normalizing, energy levels rising, and cognitive metrics stabilizing.
Takumi announced the news in the group chat.
Takumi:
[Prelim results positive. No catastrophic side-effects. Muroto-san is ecstatic.]
Fujiwara Chika:
[YAY!!! PARTY WHEN?!?!] 🎉
Bronya:
[Start slow. Continue monitoring. Don't let enthusiasm override data.]\
Muroto (posting a lab selfie, hair worse than ever):
[I love you all. I will not sleep. I will analyze. I will protect them.]
Ganyu:
[I will help document ethics and long-term welfare.]\
Zhongli (private):
[If you need resources, ask. Liyue will assist with funding and materials.]\
Takumi read the messages and finally let himself smile without irony. A million small, human things stitched his evening together: a cookie shared with a girl who could run, a lab report that validated months of frantic work, friends in a chat throwing silly emojis and solid support.
He looked up at the stars simulated above the city and felt both the loneliness of infinite responsibility and the warm hum of purpose.
The Herrscher of Reason had fixed a broken code today.
The world had grown a little safer.
And somewhere beneath the western ruins, something older still whispered.
He was ready to listen — but now he would listen with a plan.