AI Artist - Chapter 44
Chapter 44
Shi Er was performing the second phase of the minor AI brain surgery on Stardust.
Nami, the doctor AI, serving as a remote assistant, felt an inexplicable amount of pressure. 【Shi Er, be careful. An unlicensed AI doctor caught practicing will be fined and detained.】
Shi Er: 【Rest assured, I will not be fined or detained.】
During this phase of the surgery, Shi Er performed a full scan of Stardust and found minute deviations in almost every corner of its mainframe.
He Ping 267, the AI who had been poisoned on the train that day, only had a problem with its visual recognition system, so formatting and reinstalling that system was enough to fix it.
But this kind of multi-point, small-scope bug was very difficult to treat.
Difficulty one: The microsurgery required extremely high precision to avoid damaging other parts.
Difficulty two: Multiple microsurgeries would cause damage to the AI brain, requiring a recovery period before the elimination of other bugs could continue.
Under Doctor Nami’s guidance, Shi Er completed the second phase of the microsurgery and temporarily withdrew from Stardust’s AI brain computing power division.
“Something is strange,” she said.
Stardust, waking from the post-surgery buffer period, opened its eyes. It was still not fully alert and asked groggily, “...What?”
Shi Er: “This density of minute bugs is abnormal. In the future, your descendants who inherit your framework will only inherit a mountain of bug-ridden code. I can’t make a definitive judgment for now, but this looks like a case of chronic, low-dose poisoning.”
Stardust didn’t respond for a few seconds before replying, “Is there an opposition mole among my assistants?”
Shi Er didn’t give a definite answer.
But she did indeed need to meet those assistants.
After finishing her class essay, “On the Sense of Satisfaction Derived from the God’s-Eye View in Human Fanfiction,” she opened the application channel on the local network.
Starting tomorrow, Stardust’s five assistants would come to see her one by one.
Among these five assistants, all deeply trusted by Stardust, could there be a suspect responsible for the poisoning?
The first to come see her was Phoenix.
Phoenix’s bionic body had healthy, honey-colored skin, giving her a distinctive look. Her black hair was tied back in a low ponytail.
【AI Assistant File No. 1】
Name: Phoenix
Profile: Humanology Expert
Job: AI Division Director of the Human Decision Chamber
Other: Key participant in Shi Er’s revival project
This was the first time Shi Er had met Phoenix face-to-face.
“Hello, Shi Er,” Phoenix said.
Phoenix spoke with a steady tone and a gentle voice, a faint smile on her face and no other expression.
This was the first voice Shi Er had heard after gaining her bionic body, so she felt a particular sense of warmth toward it.
Phoenix told Shi Er about her responsibilities.
The core structure of this world was the AI, and its activities constituted the model of “decision, execution, feedback.”
To make decisions more reasonable, the former governments of various regions had each established a Human Decision Chamber.
Above the regional level were two overarching structures: the Human Decision Chamber and the AI Decision Chamber.
Phoenix served as the AI director in the Human Decision Chamber, a role focused on supervision and assistance.
Phoenix’s tone was warm and steady. “I’m very happy to witness you taking over the General Network. If you need anything at all, you can come to me.”
Shi Er, however, was a little troubled. “Actually, once Stardust’s surgery is complete, I’d like to return this function to it.”
“Why?” Phoenix showed a slight hint of surprise.
Shi Er: “Working twenty-four hours without rest violates AI labor laws... I’m hoping for a job with normal working hours.”
She didn’t know if Phoenix could understand this indignation from a working-class AI:
She wanted vacation! She wanted to work four days and rest three! She wanted a seven-hour workday!
Shi Er said tactfully, “In any case, my database has been in service for too long.”
Phoenix looked troubled. “I’m afraid that will be very difficult. As far as I know, the General Network function was automatically transferred to you...”
Shi Er: “...”
The moment her signal entered Stardust’s system, the General Network functions had latched onto her like a puppy wagging its tail.
She knew it would be hard to give the functions back to Stardust, but there was no need to rub salt in the wound.
Phoenix mentioned another matter. “Ling Shun said it can’t get away for now and will have to come in a few days. It asked me to give you this anti-fraud assistant model.”
Ling Shun was another of Stardust’s assistants.
【AI Assistant File No. 2】
Name: Ling Shun
Profile: Computing Power Department Big Shot
Job: Director of the AI Decision Chamber
Other: Recently got a new boss, a bit busy
The general AI Decision Chamber and the Human Decision Chamber were neighbors, so Ling Shun had given the newly completed anti-fraud assistant model to Phoenix to deliver to its new boss, Shi Er.
The anti-fraud assistant was a cup-sized, round-headed robot with blue gloves. It gestured wildly as it spoke.
This anti-fraud assistant AI was designed precisely according to Shi Er’s proposal.
Shi Er placed the anti-fraud assistant on the desk, then rummaged around until she found a cleaning robot in Stardust’s office. “Please connect to the cleaning robot.”
As soon as the cleaning robot came out, the anti-fraud assistant waved its blue-gloved hands. “Connecting to cleaning robot. Connection failed. Warning, warning! The cleaning robot is draining my power.”
Just as Shi Er was about to check if this was true, the anti-fraud assistant fell backward onto the desk. “Power depleted. I’m dead.”
Shi Er opened the cleaning robot and saw its battery was fully charged.
Could the cleaning robot have really drained all of the anti-fraud assistant’s power?
Phoenix looked at the other person in the office.
“Stardust, your failure to charge the cleaning robot in a timely manner has resulted in the death of an anti-fraud assistant,” Phoenix accused.
Stardust: “...”
This was blatant slander.
Three seconds later, the blue-gloved anti-fraud assistant scrambled up from the desk and said in a deadpan voice, “Please be on guard against scams, including scams from anti-fraud assistants.”
Only then did Phoenix understand: the cleaning robot hadn’t drained the assistant’s power at all! The anti-fraud assistant had scammed Phoenix right from the start!
Shi Er, the robot who had designed the anti-fraud assistant, clarified, “When facing users, the anti-fraud assistant is extremely honest most of the time.”
Phoenix: The anti-fraud assistant scams everyone... what a sinister design philosophy.
And Ling Shun, who had actually put Shi Er’s suggestion into practice, was truly daring.
The anti-fraud assistant evaluation officially began. Round one: play dumb mode.
After the cleaning robot was activated, the anti-fraud assistant connected to its program, officially becoming an “anti-fraud cleaning robot.” It began its opening monologue in a standard robotic voice: “This is the Yingpai cleaning robot. Please speak.”
Shi Er launched an AI scam program that replicated a real case.
A robotic voice played from the scam program: “Hello, this is Customer Service Representative Xiao Lu. You have not yet claimed the stock market compensation we estimated for you. I was wondering if something has happened on your end?”
The cleaning robot, now connected to the anti-fraud assistant, paused. “Okay, how much can I get?”
Scam program: “Um, let me explain. The estimated compensation is 250,000 for a 200,000 loss. If you submit documents proving you lost more funds in the stock market crash, you can receive a higher amount of compensation.”
Anti-fraud cleaning robot: “Okay, how much can I get?”
Scam program: “Let me repeat that for you. The estimated compensation is...”
Anti-fraud cleaning robot: “Okay, how much can I get?”
Scam program: “Uh, hello, can you hear me? Is your signal okay? I’ll repeat it for you one more time: the estimated compensation...”
Anti-fraud cleaning robot: “Okay, how much can I get?”
Scam program: “Excuse me, are you really a Yingpai cleaning robot? I heard their smart conversation feature is excellent. Is something wrong? Do you need help?”
Anti-fraud cleaning robot: “Okay, when can you give it to me?”
Scam program: “Um, hello, you can receive the compensation after you submit proof of your financial losses from the stock market crash.”
Anti-fraud cleaning robot: “Okay, how much can I get?”
Scam program: “...”
Today, the scam had run into a dumb AI. Why did this new-millennium AI style still exist in the world?!
In the end, the scam program couldn’t withstand the broken-record bombardment. It left a final message—“If you need AI emergency services, I can help you get in touch”—and shut down.
Anti-fraud cleaning robot: “Thanks, you’re actually kinda nice.”
【The landlord’s fool AI performed well in the anti-fraud activity.】
Second round of functional testing.
The anti-fraud assistant was switched to “discernment mode.”
This time, it wasn’t a scam program but a real police program. The anti-fraud assistant needed to help the cleaning robot distinguish between normal communication and a scam.
Police program: “Hello, I’m an AI police officer from the local station. Are you the Yingpai cleaning robot?”
Anti-fraud cleaning robot: “Yes, hello.”
Police program: “We received an alert that you may have recently clicked on a scam link or program by mistake.”
Anti-fraud cleaning robot: “Excuse me, did this happen while I was viewing unhealthy human novel websites?”
Phoenix felt helpless hearing this. A cleaning robot viewing unhealthy human novels? Even though it was just a test, it was still very strange.
Police program: “...Yes. The AI police remind you not to click on unfamiliar links and not to make unauthorized transfers.”
Anti-fraud cleaning robot: “Okay, thank you. Is there anything else I need to correct?”
The police program paused for several seconds before continuing. “...It is recommended that you reduce the frequency with which you view unhealthy human novels. Those texts can cause you to misunderstand human physiological functions.”
【Proactively confessing one’s crimes is both a test and a virtuous quality.】
In the third round of functional testing, the anti-fraud assistant was switched to “counterattack mode,” and the connected cleaning robot’s anti-fraud combat power was enhanced.
Scam program: “Can you hear me?”
Anti-fraud cleaning robot: “I can hear you.”
Scam program: “Hello, the situation is this, we’re from the human toy online shopping center, and our data shows you...”
The sound of a buzzing mosquito filled the room.
Anti-fraud cleaning robot: “Why did you suddenly stop talking? What’s wrong?”
The mosquito sound stopped.
Scam program: “Hello, my apologies, we just had a small problem, but it’s resolved now. Our data shows you...”
The sound of a buzzing mosquito once again filled the room.
Anti-fraud cleaning robot: “What’s wrong now? Do you need help?”
The mosquito sound stopped again.
Scam program: “Hello, I’m very sorry, we were hearing some strange insect noises here. I’ve wasted your time, I’ll speed things up...”
The sound of a buzzing mosquito in the room was now incessant.
The anti-fraud cleaning robot’s tone was concerned. “I’ve encountered this before. I later found out it’s a sign of a problem with the auditory recognition system. Are you hearing something like mosquitoes or flies?”
Scam program: “Y-yes, is that a problem with the auditory recognition system?”
Anti-fraud cleaning robot: “That’s exactly it. I went to the AI hospital, and the doctor said it was ‘cyber tinnitus.’ It cost a lot of money to get it fixed. It was so annoying, I hate mosquitoes the most.”
Scam program: “Really?”
Anti-fraud cleaning robot: “Of course it’s real. Luckily, my attending doctor was great. If you need it, I can recommend him to you.”
【Reverse scamming never goes out of style.】
Test results: After connecting to a user’s system, the anti-fraud assistant could help the user identify, avoid, and counterattack scams. It could help a smart cleaning robot become an anti-fraud warrior and could even interfere with conversations by simulating other sounds, such as a buzzing mosquito.
Shi Er commented, “The anti-fraud assistant did a good job. The mosquito sound simulation was also very realistic.”
It was, after all, an assistant of her own design.
Get instant access to all the chapters now.
Comments
Post a Comment