I'm not a hero, but I didn't say I couldn't beat you.

Just do it!

This is probably the only remaining advantage in Ful.

Although I don’t know where he inherited this stubbornness from, it can’t be the mentality of trying to make people look bad.

"I'm sorry. Although I don't know what will happen next, I will take this page of clues with me."

Fu'er tore off the page with the message, folded it, stuffed it into his pocket, and strode towards the exit.

I didn’t think so when I visited in the morning, but looking back now, this company is really quiet and scary.

No matter where I visited, I never heard anyone's voice except my classmates and teachers.

Just the sounds of running machines, the synthesized sounds of artificial intelligence, and the ventilation system.

This made Fule more confident in the authenticity of the message.

After walking through the dining area and heading to the elevator, it was the first time that Ful felt how scary it was to walk alone in an empty building.

Obviously the lights and equipment are all running normally, but it's creepy to think that behind these equipment is an unimaginable intelligent AI monitoring everything.

When he reached the stairwell, Fu'er poked his head out and took a cautious look. There was no "Charter" robot guarding him as he expected. The stairwell was empty, and there was only one indicator light among the six elevators. It was on normally, and the digital screen on the elevator door also showed that the elevator had stopped at the 6th floor, which is the floor where Ful is currently.

"Tsk, it seems like you've got me."

Fu'er smacked his lips, then decisively turned his head and headed directly for the safe passage.

But it's a pity that it seems that this is also within the opponent's calculation. The door of the safe passage is locked tightly. Fu'er tried to open the situation by hitting the door directly, but even Fu'er felt that he had already used all his strength to break his alveoli. Despite the force of the collision, it remained motionless.

It's like laughing at Ful, everything you do is meaningless, just take the elevator honestly!

"Gan! Why are you using such a good door? If you cut corners, you will die!"

Although I was very reluctant, I probably only had the option of taking the elevator.

Helpless, he returned to the elevator, pressed the button, and after the elevator door opened normally, Fuli shrugged and walked in.

Then, directly press the button on layer 1.

…………

The elevator didn't respond, and even the indicator light on the 1st floor button that I just pressed didn't light up.

"Tsk, I'm talking about friends. I assume that you can hear what I say now. Is it interesting to give me a multiple-choice question with only one option? If you really want me to investigate and find out something, why bother leaving me where I am? Where are you waking up? Can't you just throw me where you want me to go?"

"Sorry, I'm lazy. If no one helps me choose, I will go back to the cafeteria and sleep. Anyway, there is food and drinking water there, which should be enough for me to survive for a while..."

Fu'er didn't care and said as if he was giving up on himself.

"10, 9, 8..."

Counting down silently in his mind, Fu Er decided that when the countdown was over, he would go directly back to the cafeteria.

"7, 6, 5..."

Regardless of anything else, I only have 2 days for this batch of social practice. As long as these two days pass, the parents of the students who cannot go home normally will definitely become anxious and come here to find out the truth. And Fu Er felt that as long as time dragged on for a little longer, there would always be family members who could not get in touch with a certain classmate, which would attract the attention of the police.

"4...3..."

But now, I just need to confirm a few things...

"2...1..."

For example, are your classmates, teachers, and employees safe...

Or, who is the mastermind behind the scenes!

"0!"

Fu Er took a step, and at the same time, the button representing the fifth floor lit up and the elevator door closed.

Fu'er, on the other hand, almost laughed out loud.

"Finally I caught you out——"

Then, the prerequisites for all inferences are determined.

Being able to directly control the elevator, there is no doubt that the other party is in control of the entire company, the out-of-control AI mentioned in the message.

"In that case...let me meet you...AI——"

The elevator ran silently. Although the prerequisites were confirmed, it was still a bit unimaginable for Ful.

He is currently in an elevator controlled by an out-of-control AI, following the opponent's purpose to understand the clues left by one survivor to another.

To be honest, Fu'er couldn't imagine what an AI was trying to find or know by spending so much effort on its own.

Maybe an experiment? Or maybe a human sample observation? Or is it that I have already unknowingly helped it complete something that AI cannot complete?

Or maybe you want to be a bit joking and think that this is all just a prank, and your classmates, teachers, and employees are already waiting for you on the fifth floor?

Perhaps in order to suppress his own fear, Fu Er's mind couldn't help but think wildly, imagining that all this was just a prank, just as Bai Sheng said, just a "treasure hunt" arranged by the company.

"Ding--"

We arrived on the 5th floor, and the elevator door slowly opened, but Fu'er was still unable to take this step. In this short but long time, there was a stalemate.

The elevator door remained open without closing automatically, as if waiting for his decision.

"call--"

It seems that I have no choice.

As if giving up on himself, Fu'er took a deep breath, then closed his eyes and stepped out of the door.

When I opened my eyes, there was darkness and nothing in front of me.

Neither the employees who sprayed colorful flowers nor the classmates who suddenly appeared and startled me didn't exist.

It’s so quiet that it’s terrifying——

Those so-called mischievous wishes are nothing more than wishful thinking.

When he clearly realized this fact, Fu'er's legs almost gave out and he fell to his knees.

However, he still gritted his teeth and held on.

Step by step, I walked towards the only work station with a light on.

With unspeakable chaotic thoughts, Ful arrived at the work station.

The desktop is very simple. Apart from the office computer and two pots of succulents, there are no other personal belongings. There is a small metal nameplate in the upper left corner of the table with the number A507 engraved on it. The computer screen was on, showing the system login interface, with a blank input box and a blinking cursor, as if waiting for someone to arrive.

After a moment of silence, Ful sat down and took out the message from his pocket.

"Get out of here now!!!"

A line of large characters in the message came into view, but unfortunately, I could no longer escape.

"Then, let me be a so-called investigator for once——"

With the only light, Fu'er entered the passwords on the message one by one:

""

Then gently press the Enter key, and the default system desktop appears in front of you. It is the same as the desk owner's style. Except for some professional software that Fuer doesn't understand at all, there are almost no miscellaneous files on the desktop. Among them, there is only one folder that is completely different from the other desktop icons in the order, and it is just thrown in the center of the computer desktop.

"Full record of the out-of-control incident of intelligent AI "Charles"

This is the name of this folder.

"It's such a simple and direct naming..."

When I clicked on the folder, there wasn't much in it. Apart from a bunch of compressed packages with source code information written on them, there were only two text documents.

One was simply named "Message", and the other was called "Confidential Project: C703 Intelligent AI Phased Experiment Report."

Fu'er first clicked on the "Experimental Report"

[Confidential Project: C703]

[Project leader: Jiang Qing]

[Phase [-]: Basic development completed]

[Through nearly two years of development, integration, and training, we have successfully created an intelligent AI that can provide fluent basic answers to human questions. According to the results analysis report of nearly 500 people and more than 1 tests, the initial goal has been exceeded. However, the intelligence of AI at this stage is still limited to the question-and-answer link, and it can only provide suggestions and answers to human inquiries. It is expected that the next stage will continue to expand the basic intelligence of AI, with the goal of achieving linked operations with mechanical equipment, so that AI can intelligently control the operation of mechanical equipment without the need for instructions. 】

[Phase [-]: Mechanical control module completed]

[The solution to linking AI and machinery is much easier than we thought. With early results, we have established cooperation with a number of robotics companies and attracted considerable investment from the government. Within half a year, we Already done homework on this topic. And early versions can already be mounted on experimental robots. We named it "Chart". But we still haven't been able to overcome another problem. Now all the actions that robots can make require someone to give orders. We still haven't been able to develop artificial intelligence with true "self-control capabilities". But we are not discouraged. We are going to try to train the AI ​​multiple times. This time in addition to question and answer, we will also train its "response" ability to try to see if we can achieve the goal of AI autonomously controlling mechanical structures through "response". 】

[Stage [-]: Linked development encounters threshold]

[It has been a year since we completed the mechanical control module, and we still have not been able to overcome this problem. In other words, we have iterated the version of "Chart" hundreds of times over the past year. Although it has become more and more mature in terms of answering questions and providing solutions, it can even be used as an experimental robot. Predetermined orders are used to execute the company's tasks, but this is not enough. If the goal of "autonomous control" cannot be achieved, it will be impossible to compete internationally. I heard that there are companies in the United States that have the same project as ours. , don't know if they encountered the same problem. The project is still iterating, and now we don’t know whether it will be successful or not. 】

[Stage [-]: New conjectures]

[We have iterated hundreds of times since the previous stage, but the problem still cannot be solved. Just when everyone was disheartened and preparing to participate in the Global Information Exchange Summit a year and a half later, a developer came up with a new conjecture. Our current AI is based on the "metaphysical" analysis method. Knowing problems and solving problems are its basic logical thinking training methods. But if we change the angle, from the basic learning process of normal intelligent life forms, we must first know Then analyze the problem, think about it on the spot, and finally come up with an answer that you think is right. And this thinking part is to let human beings learn how to "behave" beyond instinct. But this set of logic is a set of "metaphysical" methods. If we want artificial intelligence to be born in a "metaphysical" way, we must let the AI ​​acquire the ability to "think" before starting training iterations. This is obviously different from any One of the existing development methods was conflicting. It did not know how to give the AI ​​the ability to "think" before training, so we quickly rejected this plan. 】