Memory inheritor

Chapter 98 Mutual Benefit

After addressing Shen Qing's needs, Wei Hongsi returned to refining the project proposal.

He logged into the company's central server, found the standard model for preprocessing audio files, imported the eight audio files recorded by Shen Qing, and processed and organized them into eight data packets.

Next, we'll need to use the CMS again. As a core technical member of both the IR1 and CDU project teams, Wei Hongsi already has access to the system's regular functions and doesn't need to apply for anything further.

WD-CMS is an AI system based on cognitive science. Its core function is to simulate, analyze, and reverse reconstruct human cognitive behavior.

Previously, Wei Hongsi's use of CMS mainly revolved around the "see" function, but now he wants to make it "hear".

In fact, CMS has more than just that. It also has a sense of smell and taste, but these functions are not fully developed and are still at an infantile level, far less powerful than its senses of sight and hearing.

The reason for such a large difference is mainly due to the constraints of hardware conditions.

In comparison, sensors for vision and hearing, such as cameras and microphones, are technologically mature, with very convenient signal acquisition and easy digitization into pixel arrays or sound waveforms, and the data format is relatively uniform.

Smell and taste involve complex chemical molecule recognition, requiring high-precision, high-specificity chemical sensor arrays to convert odors or tastes into quantifiable, multi-dimensional data. The sensitivity, stability, and interference resistance of these sensors, as well as their broad-spectrum recognition capabilities for massive amounts of compounds, present significant challenges.

Wei Hongsi handed the preprocessed data packets to CMS for parsing, which was the standard format that it could "understand".

Since it was only to supplement the research methods to be adopted, there was no need for a very complicated analysis, so he chose the more basic "core two-dimensional" analytical model.

In this model, emotion is expressed using only two continuous core dimensions: pleasure level, ranging from -1 to +1, and arousal level, ranging from 0 to 1. Emotion can be represented as a vector, with the greater the distance from the origin, the stronger the emotion.

Not long after, the analysis results came out. Judging from the data, the eight musical excerpts provided by Shen Qing were quite representative.

比如对编号1的音频文件的分析结果为：愉悦度（-0.9至-0.8），唤醒度（0.8至0.85），处于「愤怒」情绪的区间。

编号2：愉悦度（-0.2至+0.2），唤醒度（0.8至0.9），处于「惊讶」的情绪区间。

The remaining six audio tracks correspond to fear, joy, disgust, calm, sadness, and ecstasy/excitement, respectively.

This is not an analysis of the emotional semantics of musical fragments, although on the surface the two may seem the same. Rather, it simulates the complex responses of the brain after receiving auditory cognitive signals. The emotion in the final conclusion is simply a label attached to that series of complex responses.

For Wei Hongsi, the emotions expressed by those eight musical fragments are not important; what he cares about is how his brain reacts after listening to a particular musical fragment.

For example, rock music can make people feel happy and excited, while those who don't like it may feel annoyed or even angry. Wei Hongsi doesn't care about the emotions that rock music itself expresses; he only cares about the psychological reactions of the specific listeners.

However, the purpose of this study is to explore a treatment plan, so it needs to have representative musical emotional semantics.

Subsequently, Wei Hongsi increased the number of simulated subjects and performed simulated analysis on the eight audio files again. The final conclusion was basically consistent with the previous results.

He organized the relevant information and included it in the project proposal, proving that using specific music for controlled emotional regulation is feasible.

How to integrate it into the "emotional reshaping method" is what this project will study.

It was already past 3 p.m. when he finished writing the document. He checked it again, found no problems, and sent it to Gu Shiming, adding explanations for the revisions and additions.

His body felt a little stiff after sitting for a long time, so he stood up and stretched, then went to the tea room outside to get a cup of water.

After returning home and not receiving a reply from Gu Shiming, and with nothing else to attend to for the time being, I remembered the personal homepage of Professor William Finch that I had saved yesterday, so I opened it to take a look.

The report above includes an update on the academic seminar hosted by Professor Finch yesterday afternoon; all other content remains unchanged, and everything appears to be normal.

Has the task force not yet taken action?

Wei Hongsi pondered for a while, then decided that this was unlikely. The task force had already obtained crucial clues from Mu Jie and couldn't possibly stop there. Perhaps it was because the preparations weren't complete, or perhaps they didn't want to alert the enemy?

……

……

Shen Qing slept for less than five hours during the night and woke up at eight o'clock in the morning.

She first checked her chat history with Wei Hongsi, but saw no reply, which made her a little disappointed. After lying in bed for a while, she still couldn't fall asleep, so she got up to wash and get ready.

It was past 9:30, and there was still no word from Wei Hongsi. She couldn't hold back any longer and sent him a message. Although it didn't mention anything important, it was still a reminder.

Unexpectedly, after exchanging a few words, Wei Hongsi made a rather strange request.

Although Shen Qing didn't know what he was going to do, she figured it must be something meaningful, so she went to the workroom and spent more than ten minutes recording eight musical phrases and sending them to him.

Not long after, she received Wei Hongsi's suggestions for revision.

This made Shen Qing happy, but she couldn't help but pout. "You should have said you wanted help sooner! Did you have to wait for me to contact you first? Did you think I wasn't even awake yet?"

Seeing that Wei Hongsi had provided two different revision options this time, she couldn't help but feel a little puzzled. But just looking at the text, it was really hard to imagine the details, so she started to revise it accordingly.

The first solution wasn't difficult. After modifying it, I listened to the recording twice, and it did sound much better than before.

Shen Qing estimated that this version should be able to pass the company's approval, but she didn't feel happy; even her previous excitement had dissipated considerably. It was as if she had finally embarked on a path she didn't want to take, even though everyone else thought it was the right one.

She sat there for a while, then looked at the second revised proposal given by Wei Hongsi.

This revision was significantly more complicated than the first one. To her surprise, not all the suggestions were aimed at a smooth and gradual process; some were even clearly designed to create a sense of abruptness.

Shen Qing immediately became interested and began to revise the document while thinking about it.