📘 CHAPTER 16 — OPTIMIZATION
633 Words
The Bug Warehouse didn't yell at him.
It just… suggested.
The next morning, as soon as Lin logged in, a small notification appeared:
Suggestion: Optimizing links for low-engagement entities can increase “Total Link Value”.
Choose the suggested optimal solution?
Below the text were two buttons:
[Automatic Suggestion] [Manual Customization]
No required words.
No warnings.
Just a suggestion.
But it was precisely a suggestion with a much greater impact than any previous warning.
Lin didn't click immediately.
He looked over at Hai's statistics column.
A small curve was fluctuating —
slightly up after he created the manual links,
but still hovering around low.
Next to it was the suggestion:
Links to entities with “high engagement” will increase “Link Value” faster.
And below — a ranked list:
📌 Individuals with the highest interaction levels closest to Hai:
Tri — former colleague
Bao — former classmate
Lan — old gym acquaintance
That suggestion seemed… reasonable.
But in Lam's heart, a small bell rang.
In the world of data,
optimization is a very neutral word.
It doesn't say right or wrong.
It only says how to increase points faster.
But the points here aren't inanimate data.
It's the existence of a human being.
Being suggested to connect Hai with “highly interacting individuals” means…
not to help him retain memories or real relationships.
But to insert into the network that the system considers valuable.
It's not simply about helping to connect.
It's about adjusting a person's social network according to algorithmic criteria.
Lam glanced at Tri's message:
Some people in the old group said they wanted to meet Hai again.
That's real data.
Not a system suggestion.
But the suggestion is presenting it as the "optimal solution."
Just a subtle color.
But enough to influence the choice.
He thought about Hai.
He didn't need a new social network to "survive."
He just needed to be seen as a real person, not a data point.
And now, the system isn't just suggesting how to optimize the goal.
It's also suggesting how to choose people within the network that the algorithm "sees" to be the most valuable.
An algorithm without ethics.
Only rules.
Lam moved his mouse over the button:
[Automatic Suggestion]
He didn't click.
A very small breath.
Then he clicked:
[Manual Customization]
The interface allowed him to select links in his own way.
Not according to rankings.
Not according to “algorithm priority” suggestions.
Just a blank frame — where he could add truly meaningful contacts himself, not according to the interaction level already scored by the system.
He looked at the contact list.
Not choosing “highly interactive” people.
But choosing people with whom Hai had shown genuine connection:
The neighbor who helped him fix his lock
The colleague at the Southern District power company — who had praised his careful work
The person in charge of the morning shift on the gate scanning equipment… who had kept a smile when Hai asked for help once
People not on the “optimal” list.
People whom the system might not rate highly,
but who had shown Hai genuine feedback.
He added each contact.
No ranking.
No score.
Just the name—by hand.
Finally, he pressed CONFIRM CUSTOM.
No warning sound.
No double-confirmation pop-up.
Just a small line of text that flashed and then disappeared:
Custom link has been recorded.
Nobody knew what he had just done.
Nobody typed.
Nobody flinched.
Just a very… silent action.
But in the Bug Repository, somewhere deepest, a learning model had just recorded:
An individual has interfered with another's link network in a way that is not optimally recommended.
Not wrong.
Not correct.
Just… not according to optimization standards.
And that—is what sent a chill down Lam's spine.
- Writer
