"Continue?" Orochimaru raised an eyebrow and looked at Hiroki.
Hiroki took a deep breath. Without looking at the corpse lying on the ground, he walked straight toward the second man.
"Continue," he said.
Though Hiroki said this, his thoughts had already delved into the desktop in his mind. He created a new log file: [Analysis of the Causes of Failure in the Human Trial of Muscle Automation].
Subject: Root Cause Analysis of the Failure of muscle_automated_growth_v0.1.bat in Human Trials.
GOOGLE Search: TWKAN
Final objective: Analyze the causes of failure, formulate corrective hypotheses, and design a v0.2 experimental plan to achieve controlled, nonfatal muscle growth in humans.
1. Data Review and Root Cause Analysis:
Hiroki paused to think.
During the first human trial, he had used the mouse script—the version that had already been brought close to equilibrium in mice. Yet in less than a minute, the first subject had died. This was even more dramatic than the effects of the first script he had used on a mouse. At least the mouse had survived for seven days.
What had gone wrong?
Hiroki frowned as he thought.
The mouse experiment had clearly succeeded... The biggest difference between mice and humans was probably body mass.
A small mouse weighed around 500 grams, while an adult human weighed close to 70 kilograms.
That meant a difference of nearly 140 times. It also meant that the quantities of muscles and cells in their bodies were on entirely different scales.
It was highly likely that this was why the script's effect on a human wasn't simply a scaled-up version of its effect on an animal model, leading to the first subject's death.
As he worked through the analysis, Hiroki made some changes to the code—
The source file for muscle_automated_growth_v0.1.bat appeared in his mind.
Hiroki's thoughts raced as he began to modify it.
The difference in body mass had caused a catastrophic increase in resource requests. That was the most likely explanation.
The v0.1 script used a global instruction. It issued a maximum-strength resource request to every cell in the "full-body muscle group list."
For a mouse, that load might have been barely bearable. For an adult human, it might have been too much to withstand.
It was like a "DDoS attack."
A "DDoS attack" worked much like the Brain Overload Attack Hiroki had developed earlier.
Take the criminal as an example: every muscle fiber and every cell in his body sent a flood of useless requests to a server at the same time. In this case, the server was the body's circulatory system, and the requests were for things like "calcium ions."
But the human body had only a limited supply of calcium ions, perhaps not nearly enough. With so many cells requesting them at once, the supply would quickly run out.
As mentioned earlier, an adult human weighed nearly 140 times as much as a mouse. That also meant that when the script issued the same instruction, the number of cells in the human body that needed to respond was more than a hundred times greater than in a mouse.
That could easily lead to an abnormality.
Take calcium ion imbalance, for example. Hiroki managed to dredge up from his memory that the normal calcium level in human blood was approximately 8.6–10.2 mg/dL, while in mice it was 8.40 ± 1.09 mg/dL.
Those levels were quite close.
But... under a global instruction, the instantaneous resource demand—the calcium ions required by the script—
An adult human had a total muscle mass of about 28 kilograms (28,000 grams). Assuming that synthesizing each gram of muscle required 0.1 milligrams of calcium ions to initiate growth, the global instruction would create an instantaneous total demand of 2,800 milligrams.
Using the same rate of 0.1 milligrams per gram, a mouse's instantaneous total demand would be 22.5 milligrams.
By that ratio, the difference didn't seem that great.
But what if they took into account the difference in blood volume?
A mouse weighing around 500 grams had about 32 milliliters of blood, containing roughly 3.2 milligrams of calcium ions. That was seven times less than the 22.5 milligrams required.
A 70-kilogram human, however, had only about 5,000 milliliters of blood. The human demand of 2,800 milligrams was about 5.6 times the available 500 milligrams.
Although the deficit ratio seemed slightly lower than in mice, the actual shortfall was vastly different.
The mouse was short a total of 19.3 milligrams of calcium ions.
The human, meanwhile, was short nearly 2,300 milligrams.
The difference in demand was nearly a hundredfold! And the human's actual demand exceeded the mouse's by nearly 2,280 milligrams of calcium ions! Such a shortfall was devastating on an absolute scale—directly so.
And that was only an estimate.
The amount of calcium ions needed to synthesize muscle in a human body was impossible to calculate precisely.
So Hiroki had to change the global instruction to a targeted application.
His thoughts raced as he modified the code. He deleted the loop that targeted the full-body muscle group list and replaced it with a mandatory input parameter: specify the target muscle group.
That way, the patch could strengthen only selected muscle groups.
But that still wasn't enough.
Though Hiroki had killed someone in his first experiment, test subjects were limited. He didn't want to make more mistakes and cause more of them to die.
Even when targeting only one muscle group, 100% intensity was still an unknown risk. He added another parameter: a throttling mechanism, a variable called the Growth Intensity Percentage, with its default forcibly set to a conservative 5%.
This variable limited the total amount of muscle grown per unit of time, helping determine what values would cause abnormalities in the human body.
Finally, after a moment's thought, he placed his hand on the criminal. He quickly found the data he wanted among the mass of files: blood calcium, pH levels, calcium ion reserves, and more!
Next, Hiroki added the program's most important safeguard: a parallel resource monitor, similar to the one he had written for himself to monitor CPU and memory usage in real time. It would track key physiological parameters, including blood calcium and pH levels.
If any indicator reached its preset safety threshold, the monitor would immediately lower the growth intensity or trigger an emergency shutdown. He also added a work-rest timer, changing continuous high-pressure operation into a cyclical mode.
muscle_sustainable_growth_v0.2.bat, compilation complete.
Hiroki took a deep breath and began his second experiment. The man shuddered violently, his eyes filled with pleading terror. Hiroki ignored him.
There was no need to save a criminal. He was better off serving as fuel for Hiroki's growth.
In the silence of his thoughts, he issued the command: "Execute."
"Blood calcium levels are starting to drop!" one researcher said, his voice suddenly tense. "Down 3%... but they're relatively stable compared to last time!"
"Lactate levels are fluctuating within a safe range," another reported immediately. "The subject's heart rate is elevated, but the rhythm is stable. No abnormal beats."
"Could it be... we succeeded?" One of the researchers looked at Hiroki.
Hiroki took a deep breath and slowly began to speak:
"..."
Before you continue