Company News

Hi！I’m a part-time professional in software testing. You can call me Aaron.

In order to learn AI, I voluntarily took on the role of the AI Ambassador of the company. Every month, I give a sharing course on AI skills to my colleagues. At the same time, I will also use AI to polish some of the contents of my sharing courses.

Then let’s start this year’s first sharing theme of AI collaboration: The more powerful AI becomes, the more precious human beings are. This is also a speculation and preparation for the future integration of SRM-PRO with AI.

“Teacher, how should I pronounce this word?” The eight-year-old daughter held up the picture book and asked. At this moment, her AI tutor had automatically recognized the picture and was presenting an explanation segment related to the word in five ways that children could understand, while narrating it for her.

At this moment, I happened to receive an urgent inquiry message from a client, requesting me to urgently purchase a customized electronic component. So, I opened the SRM intelligent procurement assistant of ESiC Technology and quickly completed the sourcing work.

This is not a science fiction film. Instead, it is the wonderful future that we are about to embrace.

When we talk about the future of AI, many people get panicked about “job replacement”, but they overlook a more important issue: educational methods, knowledge acquisition and career values are undergoing unprecedented innovations in the past century.

How should we understand this innovation? It cannot become a reality overnight. During my AI learning in the past year, I saw an AI cognitive model – the MUSE model.

In 2023, generative AI led by ChatGPT emerged suddenly, ushering in the era of imperative large models.

In 2025, DeepSeek’s inferential large-scale model emerged unexpectedly, further promoting the evolution of AI technology.

In the future, breakthroughs in underlying technologies will keep coming. AI will evolve like “growing neural synapses”, and even usher in the era of “silicon-based life”.

After 2023, I have clearly felt that my working mode has been gradually changing. When encountering problems, I switch from searching on Baidu to searching on Secret Tower. AI enables me to find answers faster. When wanting to read a book, I no longer purchase physical copies. Instead, I buy e-books and chat with AI about them. At work, emails are generated by AI, code is assisted written by AI, and test plans and use cases are assisted written and checked by AI. When family members treat AI as a toy, we have already become dependent on AI tools in our daily work.

On mobile phones, various AI assistants are increasingly installed. From ZhiPu Qingyan, DouPao, KIMI to Deepseek and now Tencent Yuanbao, in just two years, AI has been integrated into all aspects of our work.

Since 2023, major global tech companies have been pouring resources into the development of large-scale models, and in China, a “hundred-model battle” has emerged. Every day, there are various breakthrough news reports around the world, whether it’s publicity or marketing, it’s all very lively.

For large enterprises, they are always enthusiastic about making innovations at the transformation level. For the application-oriented industries, this also means that the existing solutions can be redone all over again. Even more imaginative products can be developed.

Share a picture for everyone’s reference.

The development of AI is also an evolution process of the collaboration between humans and AI.

Let’s do a deduction together. If SRM and AI are combined, what will happen?

Phase 1: Humans as the main force, AI as the auxiliary component.

Usually, a large model is utilized in a certain work process to generate output, thereby enhancing the quality of the output and reducing the workload. For instance, SRM-PRO integrates a personal conversation assistant, which provides supplementary information and operational assistance to the purchasing staff when they are using the SRM system for work.

Phase Two: Collaboration between Humans and AI

Integrating large-scale models into each work process to assist all business operations and achieve cost reduction and efficiency improvement. During this stage, certain business modules of SRM-PRO will be restructured. For instance, the supplier information entry can be done by AI retrieving public network supplier information, automatically filling in the details, and then being reviewed and confirmed by humans.

Phase Three: AI Completes Work Independently

Let the large models handle most of the tasks, and the final decisions will be made by humans based on the results produced by the models.

Here, it can be further divided into two sub-stages.

• Sub-stage 1: SRM-PRO-Agent Assistant. Manus is at this stage. For instance, let the AI assist customers in sourcing independently, provide dynamic sourcing reports, and ultimately it is up to the purchasing staff to decide whether to follow up on the business.
• Sub-stage 2: SRM-PRO-AGI Assistant. This stage is not merely an independent business module, such as being able to handle supplier sourcing and information entry. Then it connects with the Agent of the supplier (if the other party also has an Agent), automatically completes the authentication process and makes rating recommendations. Finally, the initial rating of the supplier is confirmed by human intervention.

Then, as individuals, we must also keep pace with the times.

Many people are worried again: “With the advent of AI, will I be out of a job?”

In fact, you don’t need to worry. Currently, AI can only do these few things: identifying problems, obtaining knowledge, breaking down tasks, executing commands and providing reference answers for you.

But it cannot fully accomplish two things independently:

1. Define the goals (set goals, conduct opportunity analysis, perform ROI assessment)
2. Result-based decision-making judgment (evaluation of the quality of the result, etc.)

Although Deepseek can actively infer and analyze the questions you raise, even a single sentence, it can provide you with a framework for thinking. However, the final generated content still requires your iterative and updated processing because AI is not God. It cannot obtain all the information you have in the real world and has no way to help you make judgments.

What we need to do is to leverage the strengths of AI and deepen our expertise in value-driven decision-making, cross-domain innovation and other high-dimensional capabilities.

Therefore, I have the following three suggestions:

1. Enhance cognitive upper limit: By applying the MUSE model, understand the evolution trend of AI technology and maintain a thirst for new knowledge.
2. Strive to cultivate fundamental capabilities: Deepen expertise in value decision-making and cross-domain innovation, etc., which are high-dimensional capabilities.
3. Build your own knowledge system: Let AI be your “accelerator”.

The essence of large models is that they are “super text-sequencing masters”, calculating the probability of the next character based on trillions of parameters. However, they are still merely a historical knowledge base capable of having conversations, a comprehensive encyclopedia.

Looking back from the threshold of 2025, one can notice an interesting “intelligence paradox”: The more powerful AI becomes, the more precious human unique creativity, empathy and value judgment become. Just as calculators did not eliminate mathematicians but made them more focused on solving cutting-edge problems.

ESiCer possess creativity and empathy, and they truly aspire to improve their products to meet the latest and fastest demands of enterprise procurement. SRM-PRO faces both opportunities and challenges. Its future is bright, but the road ahead should also be tortuous. Otherwise, how can it be worthy of future success?

This point, I know it. We all know it.

Finally, we sincerely hope that the MUSE model of this article, the AI technology evolution roadmap and the personal action guide can bring you a different perspective.