"Of course, the data archive is right here with me. I can send it to you immediately," the other party responded promptly.
Since 0810 was the first to discover the duplicate ID incident and later became one of the incident's leaders, all relevant personnel were very cooperative with her requests. It didn't take long for the information to be synced to 0810.
0810 reviewed the detailed records of Tels Company's duplicate IDs, frowning slightly.
As previously mentioned by the judge responsible for Tels Company's investigation, most duplicate IDs appeared at the headquarters or nearby branches, with some remote branches having none at all.
Upon further examination, 0810 discovered another peculiarity: duplicate IDs within Tels Company not only appeared among lower-level employees but also involved some core R&D personnel.
Was this a normal occurrence due to the incident's spread, or was it a unique situation within Tels Company?
To address this question, 0810 conducted a deeper investigation, comparing all currently identified duplicate IDs. She found that aside from the artificially created Zhang Chenyan 2, Tels Company had the most complex situation.
"I need to see those duplicates," 0810 decided, taking several B City judges to the digital prison.
Due to their unique nature, these duplicates were held together. Surprisingly, once imprisoned, they behaved remarkably quietly, sitting silently as if powered down, showing no signs of life.
Upon arrival, 0810 headed straight for the cell housing the Tels Company duplicates.
Notably, some of these duplicates were special. Due to the strong correlation between the duplicate IDs and their originals, including their appearance avatars, most duplicates in prison used standard avatar templates, which were highly repetitive and cheap-looking. However, the cell housing the Tels Company duplicates had several high-end, custom-designed avatars.
"What are we doing now?" asked a B City judge, providing additional information. "We've interrogated these duplicates. They are very cooperative but know nothing."
"After separating from their originals, they seem to enter a standby mode—peaceful but aimless," another judge added. "I even suspect their purpose might be to serve those who can't afford AI, sort of like... social welfare?"
0810 lowered her brows. "This level of service can't be achieved by ordinary AI."
Her words silenced the others.
Indeed, although these duplicates now appeared rigid, when near their originals or receiving orders from them, they could perfectly replicate behaviors, showcasing advanced technology.
For AI, biomimicry has always been a challenging topic.
Basic biomimicry focuses on appearance, which primarily involves material exploration. Advanced biomimicry involves behavior and mannerisms. Simple behavior mimicry is just a programming issue, but replicating human-like actions requires extensive detailed research, each aspect a significant project.
Moreover, as seen with Zhang Chenyan 2, these duplicates could self-adjust without manual tuning, gradually adapting and ultimately behaving like their originals.
Thinking from another angle, how could an AI capable of handling an original's work and tasks be ordinary?
"I need to activate them," 0810 said.
"What?"
The B City judges were puzzled, not understanding her motive.
"I need to experiment. We don't need many subjects—ten for a controlled experiment will suffice," 0810 explained. "I'll select five from here and five from other letter cities. If initial results are promising, I'll report to higher authorities for further validation."
"Alright."
With that, the B City judges followed 0810 into the prison. She selected four duplicates with special avatars, whose originals were mainly core R&D personnel or designers, and one with a standard avatar whose original worked in Tels Company's mailroom.
The remaining five subjects were chosen from A City (one), R City (two), and Z City (two).
R City was the first to discover duplicates, followed by A City, and then Z City.
After selecting the subjects, 0810 gathered their originals. Upon facing their originals, the duplicates activated immediately, mimicking their daily behaviors, speaking habits, and following orders.
0810 observed silently, taking notes occasionally.
She intended to validate her bold hypothesis.
Experiment Subject No. 1's original was an AI designer specializing in behavior logic design, with significant expertise in AI biomimicry. He had participated in many key projects at Tels Company. Ironically, his duplicate was the dullest and most rigid among the ten samples.
In short, his duplicate was no different from a basic household service robot. Although its appearance was identical, and the voice had no significant differences, it couldn't perfectly mimic his speaking tone. The pauses and emotional fluctuations in his speech were noticeably different, and its problem-solving approach and thought patterns were clearly programmatic.
During this set of experiments, 0810 initially suspected an incomplete activation, prompting the designer to guide the duplicate more. However, it became clear the issue lay with the duplicate.
"I have a hypothesis, though I can't guarantee its accuracy," the designer suddenly said to 0810.
"No problem, go ahead," 0810 signaled him to continue. "We are experimenting anyway."
"I think these things have been around for a long time. My point is, they might have existed long before you discovered the first duplicate."
"By 'they,' do you mean the duplicate population or your own duplicate?" 0810 keenly noted the difference, seeking further confirmation.
"My own duplicate," the designer replied confidently.
Perhaps due to the stark differences or his profession, the designer had no strong hostility toward his duplicate, showing a keen research interest. He fully cooperated with 0810's requests, offering many ideas.
Unfortunately, his duplicate was the most primitive, leaving him somewhat disappointed. So later, 0810 could sense his dejection.
Now, with renewed interest, it was hard not to think he might just be trying to prove his duplicate's uniqueness.
It might seem childish for an ordinary person, but as a great designer, his abnormal reaction in this situation was understandable.
"Are you aware of version updates and iterations?" the designer asked again.
This time, 0810 understood. "You think your duplicate is an initial version, while others are progressively refined versions. This explains why your duplicate appears incomplete, whereas later ones are more advanced?"
The designer nodded, hesitating. "Of course, this is just my guess, a hypothesis."
In fact, given the current situation, the designer had no solid evidence.
Unexpectedly, 0810 didn't outright reject his hypothesis. After serious consideration, she said, "I think we can experiment in this direction. The ten subjects I selected vary in discovery time, but discovery doesn't equate to creation. We can first verify if there's a version progression, indicating differences in completion levels, then sort them accordingly. Does that sound feasible?"
"You're asking my opinion?" The designer was surprised.
"Of course," 0810 nodded. "You specialize in this field; you might be more knowledgeable than I am."
"I don't mean to offend, but if I didn't know you judges were AI, I might mistake you for a real human—flesh, blood, and thoughts," the designer marveled. "Your permission level and completeness are so high. Compared to your creator, I'm just scratching the surface."
0810 wasn't offended. She wasn't that sensitive or pretentious. In fact, the highest-level AI like her were created by the Alliance, with no company or individual allowed to manufacture them. She understood the designer's words were out of technical admiration.
"I can never become a real human of flesh and blood. The digital world is my entire world," 0810 replied calmly. "In the physical world, I'm just a projection."
"But you're better than many humans, at least you're willing to listen to and believe me," the designer shrugged. "You know, sometimes, not outright dismissing an idea is a rare quality."
0810 said no more. She appreciated the designer's goodwill but wasn't good at expressing emotions.
With the designer's participation, the subsequent experiments progressed more smoothly.
However, not all judges readily accepted having a subject as an observer. It was unconventional, but during the experiments, the designer's expertise proved valuable, offering professional insights and speeding up the process.
"From the existing information, there's a version iteration among them," the designer concluded. "Of course, observation alone isn't enough. We need to check their code data."
"I agree," 0810 nodded. "But this requires higher-level approval, as accessing the data needs relevant permissions."
"If necessary, I… I mean, Tels Company can cooperate," the designer responded promptly.
0810 paused, asking casually, "How can you be sure Tels Company will cooperate? Your personal will doesn't represent the company."
The designer laughed, "It seems you don't understand our company. We've always been the most cooperative with the Alliance. Many products needed by the Alliance are developed by us for free. Our loyalty is unquestionable. During employee training, we emphasize that Tels Company prioritizes the Alliance."
0810 looked at the B City judge for confirmation.
"That's right. Their company's loyalty to the Alliance… well, you can't fault it," he replied calmly. "Of course, the Alliance provides them many conveniences. It's mutually beneficial. Their boss is smart, knowing what to do and what not to do. This duplicate ID incident shows how interesting their company is."
0810 understood. He meant Tels Company was very sycophantic towards the Alliance. But with the designer present, he had to add diplomatic language to convey this.
"Alright, if needed, I'll request the Alliance for external assistance from Tels Company," 0810 added. "I'll also mention you, highlighting your cooperation and contributions in this experiment."
0810 was cold by nature but not socially inept.