Relatively speaking, the output is relatively simple.

Because it is a prototype, not too many output conclusions are added, just to test its thinking process.

In fact, the thinking process that supports this speculative result is the core of the prototype.

During the five-second detection time, the continuous actions of the detection object are broken down into a large number of behavior slices, and slices of different lengths are used to form continuous segment data of different nodes.

There are many points in the data at each stage that can be recorded and mined.

The speed of energy flow in the body, the body's posture defects during static slicing, the changes in defects in combo data, and the stability of the body's center of gravity can directly reflect the control over the body and combat effectiveness to a certain extent.

Taking Tunyan's fighter profession as an example, his physical defects are extremely low in any static slice. In the combo data, the center of gravity remains stable and the degree of defects will remain at a relatively stable low level.

However, the data shown by the classmate in front of me showed that his middle door was wide open, his center of gravity was unstable, and his body reaction was relatively slow. Judging from all the details, it was basically impossible for the opponent to have the ability to fight in close combat.

Moreover, through combo data, you can detect the opponent's movement details and behavioral habits.

These behavioral details and movement habits can directly reflect a person's mental state and emotional actions.

What is his current mental state? Does his behavior reflect his emotional state, as reflected by the energy flow in his body? Is his current mental state normal?

Although these data are mainly speculative, they can largely screen out abnormal individuals.

For example, the student currently being tested is in a state of extreme curiosity, mild nervousness, and necessary excitement. Combined with big data judgment, this is a reasonable psychological state, and the flow of energy in his body also happens to confirm this.

It is based on these two points, physical and psychological, that the judgment of low threat and no danger is made.

If Noah is willing, based on the current data detection standards, he can generate a plan to attack and kill the classmate. The success rate of killing him with one shot should be 70%. Moreover, as the monitoring time continues to improve and the recorded behavioral habits become more and more abundant, this success rate should be infinitely close to 100%.

Perhaps a targeted psychological intervention plan can also be issued, which, through specific behavioral guidance, can strengthen a person's current specific emotions and influence the psychological state through emotions.

As long as you have the data, you can do much more than just produce a conclusion report.

"Noah, did you succeed?"

The classmate looked at the report, which had only two lines of words, and asked curiously.

Noah nodded: "Yes, it basically succeeded."

After saying this, Noah looked at the students who had gathered around to watch and said with a smile, "If you'd like, you can go over and try it out. It's a good opportunity to check out the prototype and its current load-bearing capacity."

After hearing Noah's words, the others readily agreed and moved closer to the machine.

The prototype has no encapsulated shell yet, and the various wire circuits inside are exposed and clearly visible. They also want to take a closer look at these mechanical circuits, perhaps they can figure out some ideas based on the knowledge they have learned so far.

After all, in terms of knowledge level, most people in the group already have a very solid foundation of basic knowledge, and they even have some understanding of some of the intermediate knowledge involved in some of the expanded knowledge.

But they soon discovered that they could indeed understand it, but only a little.

It’s like everyone knows numbers and letters, but it’s difficult for anyone to understand the concise formulas composed of numbers and letters and the meaning behind the formulas at first glance.

But just the parts they could understand were enough to shock them.

They never imagined that the basic knowledge they mastered could be combined into such exquisite changes and logic.

"Wow, this is too difficult."

"Why is it arranged like this? I remember this type of probe had obvious defects. There should be warning signs."

"The reason for the flaw is that under certain circumstances, the recorded data will be distorted. However, as you can see here, other cameras and algorithms have clearly compensated for this flaw and avoided extreme situations. Even if there is a problem with data distortion, it should be able to be marked immediately."

"But in this case, abnormal data will also make the final judgment unstable."

"What you said seems to make sense. If it were me, I wouldn't take the risk. How about we ask Noah?"

The people in the study group were discussing the various modules of the prototype, and from time to time someone would turn around to check on Noah's current condition.

They were discussing animatedly.

The prototype is now faithfully scanning everyone, recording their behavior, and numbering them according to their behavioral characteristics.

Because a large amount of individual data is recorded, the original individual model is further developed into different types of group characteristic models.

This means that in subsequent processing, with the assistance of the group characteristic model, the required amount of calculation will be reduced accordingly, and it can be easily connected with the AI ​​database.

The progress was much smoother than expected, and this discovery made Noah in a good mood.

Noah naturally answered the questions raised by members of his study group.

Noah observed the operation of the machine while also paying attention to answer the students' questions:

"Any machine will have failure rates and errors, which is actually normal. But this monitor is not the case."

"The fact that it distorts data in certain situations is not actually a flaw."

"If you look up the company that produced this monitor, you'll find it's a product of Wujie."

"This monitor is actually the base model of a series of monitors. It has reserved more data processing connectors and modes. When we fail to handle its redundant contacts, data may enter these connectors, resulting in inconsistent data return times, resulting in changes in the order of data, and causing data dirty reads."

"As long as we process this part of the interface, this problem will not occur. If you don't want to use it and need to skip this part of the interface reading, you can also achieve the same effect."

"The processing logic here involves intermediate knowledge, so there are warning signs here, and it is stated that data distortion may occur under special circumstances. This does not mean that it is flawed, but it requires careful selection. At least if you want to use it, you need to understand its structural principles."

"In fact, if you take a closer look, you'll find that its sensitivity, detail, and expansibility far exceed other monitors, which is why I chose it."

"At the beginning, our project did not require very high equipment requirements. To save time and cost, we could use off-the-shelf monitors."

"Of course, if you can't even figure out where the defect is, then the warning sign has served its purpose."

After hearing Noah's words, the two students immediately began to try to find information.

They opened the model interface through the AI ​​interface, re-checked the display column with the warning sign, and carefully read the instructions below.

I found that it was clearly written above that there might be data distortion and misinterpretation.

But when they thought about what Noah had just said, they became undecided again.

In their impression, the answers obtained by AI must be correct and have never deceived anyone.

Noah saw their thoughts and said with a smile:

"When it comes to professional issues, AI's explanations may not be correct. Or, AI may understand the underlying principle of the problem but be unable to explain it in depth due to permission restrictions or other regulations. Therefore, a simpler method is used to directly issue a warning."

"In short, there are many possibilities. Who knows what it's thinking? But I can be sure that as long as you can master this monitor, its performance will be far ahead of other monitors of the same level."

After Noah finished speaking, he casually put the test data and usage records of the monitor directly onto the key knowledge network, and marked the nodes of the knowledge network directly for the two to see.

"I understand." The classmate nodded, not quite understanding.

"I see. Looks like I need to go learn something."

Seeing Noah's interest in answering their questions, the students around him became even more enthusiastic. They had come here to observe the progress of the project up close and learn from it.

After all, this matter caused a lot of uproar in the library. Even the senior students heard about it and asked them about the situation.

"Brother Noah, I want to ask about this..."

"Noah, I have a question too. Can you explain it to me?"

"Noah..."

Noah randomly picked a few students he remembered to answer questions and said, "The reason it's designed this way is because the level of monitoring I need far exceeds normal power consumption, so heat dissipation and stability must be considered."

"The structure of that place has been specially treated, and the program inside has also been rewritten. In order to ensure the best monitoring effect at different time periods, more judgment processing has been done."

Just as Noah was communicating with his classmates, someone suddenly noticed that the machine began to flash red.

The test results that should have been output were suddenly blocked at this moment.

Someone murmured quietly: "Is there a problem?"

"It's possible. I saw a clear error message on the display screen over there, with a large number of abnormal records."

Another person whispered.

The others were also standing aside, waiting to see how Noah would handle this sudden accident.

"excuse me."

Noah said as he walked over to the instrument's console.

Open the instrument's console interface, expand the pop-up window, call out the log, and filter the information by entering commands.

Extract the most recent exception log.

Noah looked through the log information and found that it was an abnormal program interruption. Based on the error code location of the program interruption, he found the corresponding module.

Pretty much what I expected.

Noah understood.

This is also a part that I deliberately left behind so that the program can be iterated.

When recording a small amount of data, there is no problem in the operation of the instrument, and multi-dimensional data can be entered smoothly, because no matter what, these are the data of one person. Although they are different attribute directions of one person, the landing points of these data are different tables and will not affect each other.

However, as the number of test objects increases, the entered data will enter the table at different times. How to reasonably arrange and plan the testing and the loop of sequential data flow becomes very important.

These all take time to troubleshoot.

From single data to more data, then to large amounts of data, and finally to massive data, there needs to be an iterative evolutionary process.

This is also a problem that Noah must solve.

Demands arise from problems. As a novice, in order to avoid being labeled by AI, it is best for him to complete his growth trajectory. He can be smart and talented, but not too evil.

When you are not clear about AI's specific attitude towards humans, there is nothing wrong with being careful.

Noah smiled and said to the others, "We've almost understood the problem, but we'll need some time to figure out how to handle it. Thank you all for your help with today's testing. If you need further testing later, please feel free to help."

"I've organized the relevant knowledge points about this detector into a key knowledge network, along with some detailed optimization questions. It can be considered a small interview."

"If any of you have good ideas, you can directly answer them here. When we need to expand our staff in the future, we will give priority to students who have a certain understanding of the subject and relevant technical background."

Of course, those who joined the study interest group were looking forward to it and were very happy, but some students who did not join the study group did not take it seriously.

They don’t know what learning online is all about, nor do they know about the efficiency improvements that learning online brings.

In their opinion, they will naturally not do anything without rewards.

……

In the following period of time, Noah continued to debug the machine.

It took two days to hand over the completed prototype to Tunyan and Qiantong, which meant they could start trying to "launder" the data they had collected in advance.

Moreover, with monitoring instruments, they can more accurately grasp some data parameter dimensions that they cannot understand.

Because the entire project is for AI service, after the prototype is completed, the homepage provides them with a sub-AI, which serves as the embodiment of the main AI's will and connects with their program to process data.

The data collected by the prototype will flow into the sub-AI.

In order to better express its own evolutionary process due to these data, and to facilitate Noah and others to collect targeted data according to its evolutionary needs, and to allow Noah and others to clearly see the current progress of the project, Zi AI opened up its own thinking process.

Tap the screen to use advanced tools Tip: You can use left and right keyboard keys to browse between chapters.

You'll Also Like