GuShiio.com 
(Photo source: unsplash)
Entering the peak job-hunting season for gold, three and silver, under the DeepSeek craze, many companies are eager for AI talents and have begun to compete for AI model talents with high salaries.
Xiaomi said that the talents it is looking for must have a master’s or doctoral degree in computer science, artificial intelligence, natural language processing or related fields, more than 3 years of relevant experience, and have a certain understanding of multimodal technology, mainly used in AI technologies including large models, Agents, multimodal, etc., to promote the dataization and intelligence of corporate operations.Prior to this, Xiaomi recruited Luo Fuli, a former researcher at DeepSeek and Alibaba Cloud, which attracted market attention.
At the same time, Huawei is also recruiting AI model technical talents at high salaries in Qingpu, Shanghai.The monthly salary is 35,000 – 65,000 yuan, and the annual salary is as high as 1.04 million yuan.
Since February, many leading companies such as Meituan, Didi, Ant Group, Jingdong, Yushu, DeepSeek, and Zhipu have been recruiting AI talents. Also, I noted thatWell-known investment institutions such as Sequoia Capital China are also recruiting large-model investors with salaries of up to 200,000 yuan.
The demand for AI talents has exploded.
Zhaopin recruitment data shows that in the first week after the Spring Festival this year, the number of AI-related job seekers increased by 33.4% year-on-year, ranking first in the industry. The growth rate of AI engineer positions reached 69.6%, close to 70%. On the recruitment side, AI engineers lead the ranks first in the profession with an average monthly salary of 21319 yuan, while the average monthly salary of AI engineers requiring more than 5 years of work experience even exceeds 35,000 yuan.
Obviously, the rapid rise of AI big model technology has a crucial impact on the job market, human resources (HR), etc.
On February 28, Zhang Yuejia, president of Zhaopin Recruitment Group, said thatIn 2025, China’s AI market will account for 20.9% of the world’s AI market, and the Internet industry’s AI penetration rate will reach 89%.The job market is undergoing structural transformation. The recruitment demand and employment demand of traditional enterprises are declining, while the demand for positions related to flexible employment and new quality productivity is rising. Therefore, enterprises need to make full use of AI capabilities to serve the entire organization.
DeepSeek allows companies to increase their demand for AI talents instead of decreasing
“We will definitely increase some AI algorithm engineers accordingly. The current recruitment volume has also increased a lot compared with previous years. rdquo; At a recent job fair, Zhang Dongdong, human resources manager of Zhejiang Yushu Technology Co., Ltd., said.
In fact, in February this year, GuShiio.comAGI noticed during the Global Developer Pioneers Conference (GDC) held in Xuhui, Shanghai that dozens of AI and robotics startups were recruiting a large number of talents related to algorithms, data, and large models.
At the same time, Shanghai recently held its first spring comprehensive large-scale job fair to promote employment in 2025. More than 1000 companies participated, providing more than 26,000 recruitment positions. Among them, 500 participants at the main venue provided more than 14,000 recruitment positions, focusing on talents in industries such as integrated circuits, biomedicine, and artificial intelligence. Many technology companies participating in job fairs said that they are in urgent need of high-level talents in technology and algorithms.
GuShiio.comAGI learned that with DeepSeek’s rapid popularity around the world, the domestic AI field has ushered in a new turning point.The demand for talents in the industry has increased instead of decreasing. Internet companies such as ByteDance and Ali have increased their efforts to find top AI talents, and even from“The six tigers (Baichuan Intelligence, Wisdom Spectrum, Dark Side of the Moon, MiniMax, Step Stars) poached people, and the latter began to divide and turnedTo B application and commercialization, the other party turns to technology iteration and initiates organizational change.
In February this year, I exclusively learned that“Wei Wei, partner, vice president and head of open platform of MiniMax, one of the six big model tigers, resigned. In response, the company responded that the news was true and MiniMax’s domestic B-end business has ushered in a new stage of development and will be led by other leaders. team.
GuShiio.comAGI exclusively learned that Wei Wei’s next step may consider his entrepreneurial direction, and many investment institutions are in contact with him. Previously, MiniMax confirmed that Zhang Qianchuan, its product leader and former product leader for Today’s Headline Users, has currently faded out of company affairs due to personal reasons.
At the same time, according to late reports, Cao Dapeng, the person in charge of PopAI products from Zero Everything, recently joined ByteDance to serve as the person in charge of the AI video generation product iDream mobile terminal, reporting to Zhang Nan, the person in charge of editing and screening. In addition, senior executives such as Huang Wenhao, vice president and head of model pre-training of the Zero Everything Algorithm, and Li Xiangang, co-founder of Zero Everything, have also left.
However, while the executives left, the big model six tigers began to transform their strategies and re-recruit corresponding talents.
On the recruitment platform, we strive to find positions such as senior product manager of large models and digital human direction algorithm engineer in the AI application layer; we strive to launch MiniMax for underlying models and C-side applications to release positions such as video generation, multimodal large models, large language models, sound generation algorithm engineers, as well as training frameworks, algorithm service research and development, marketing managers, and data analysis.
Yan Junjie, founder and CEO of MiniMax, recently said that DeepSeek inspired him: First, their brand is very good, and its reputation and reputation are one of the best in the quantitative industry. Another inspiration is that DeepSeek did not have a product at the beginning, so it is more focused.
Therefore, Yan Junjie believes that the model capability classification should be clearly defined first, the algorithms, data and reasoning processes needed to improve each generation of models should be clarified, and technical means should be used to approach the established indicators. At the same time, technology and products need to be distinguished. The core of technology lies in continuously raising the upper limit, which requires the ability to define the next generation model.“Don’t think that the model will get better after you have a product. The purpose of the product is not to make the model better, it is just a commercial product. What you really need to think about is how to better satisfy users.& rdquo;
Founded in 2021, MiniMax has completed four rounds of financing totaling more than US$1 billion, and the company’s latest valuation is US$3 billion (CB Insights data is US$1.2 billion). Liu Hua, vice president of MiniMax, said in a speech in February that DeepSeek made the industry realize that the optimization effect of a major iteration of the underlying model on the product far exceeds the carving of the product at the engineering realization level. At the same time, the big model will quickly reach the level of professionals, Al Agents will reshape the global economy, and the big model will transform from the current pure Internet product into a new service industry.
Liu Hua emphasized that in the next 2-3 years, technology upgrades similar to GPT3.5 GPT4 will occur twice and are highly predictable.
GuShiio.comAGI exclusively learned that at the end of February this year, MiniMax launched the Conch Video stand-alone mobile app in the domestic Apple App Store, and launched a payment model. The price starts at 68 yuan per month and reaches a maximum of 2648 yuan/year.At the same time, MiniMax has also developed the AI search product Everything Tracking, and has launched its official website.
In addition to AI startups, Xiaomi, which is in the limelight, is also increasing its investment in AI and cutting-edge technology research and development.
Lei Jun, founder and CEO of Xiaomi Group, revealed that from 2021 to 2025, Xiaomi invested 105 billion yuan in core technology research and development.This year, Xiaomi will invest RMB 30 billion in technology research and development.
“Therefore, no matter which product we recently released, it will give the market great competitiveness and be very popular among users. If Zhongguancun enterprises want to survive in the market, the core thing is to strengthen their competitiveness and increase investment in R & D. rdquo; Lei Jun said frankly that in times of difficulty, you must have confidence. We must believe in China’s scientific and technological innovation, we must believe in the China market, and we must believe that continuous investment in core technologies will definitely pay off.
Talking about AI technology,Lei Jun recently admitted that Xiaomi has been focusing on the fields of mobile phones, automobiles, smart homes, and smart manufacturing in the past 15 years since it started its business. We must make up our minds to implement the latest technologies, such as how AI technology can be applied to all of our terminal products, allowing consumers to experience the better life brought by technology, how to expand the domestic market while looking global and improve the global influence of China products.
Talent has not been replaced by AI, and recruiters should urgently improve their soft skills”
At present, the external environment is becoming increasingly complex, the employment situation is facing severe challenges, some industries and enterprises still have many difficulties, and the number of college graduates continues to expand. Moreover, AI-induced technical unemployment may appear in various industries, and the market is in panic and complex emotions about the topic of AI replacing humans.
According to the latest unemployment statistics released by the National Bureau of Statistics,In December 2024, the unemployment rate among the 16-24-year-old labor force excluding school students in urban areas across the country was 15.7%, a decrease of 0.4 percentage points from November. Throughout 2024, the average national urban surveyed unemployment rate is 5.1%. Simply put, about 5 out of every 100 urban people face unemployment.
At the same time, humanoid robots at this year’s Spring Festival Gala started singing yangko, and more and more people began to immerse themselves in conversations with DeepSeek. AI is gradually permeating into all walks of life.
According to the World Economic Forum’s “2025 Future Work Report”, among the more than 2800 fine skills, the proportion of AI currently able to reach human capabilities to a certain extent is about 28.5%, and nearly 40% of work skills will change.As technology continues to advance, this percentage will continue to rise. For example, in some repetitive and logical tasks, the entry of AI can improve overall efficiency and is expected to partially replace human work.
Li Lixing, a young scholar at Peking University and professor of economics at Peking University’s National Development Research Institute, once said that overall, the current employment trend is becoming increasingly severe. The prominent phenomenon is the high youth unemployment rate. At the same time, another trend in employment is short-term and flexible. Globally, more and more workers are moving away from full-time jobs in companies and looking to the gig economy. Regarding AI, it was previously mentioned that machine replacements mainly affect blue-collar jobs. However, the biggest impact of AI on the workplace today is that it can make logical thinking decisions and then replace white-collar jobs. This is an important reason that causes people to be anxious.
“The essence behind this problem is how to deal with it. For individuals, they should consider how to improve their skills and coexist with AI, so that AI can play an auxiliary rather than a substitute; for enterprises, when faced with promoting machines to replace people, they should still discover business opportunities and hire more people to use AI to create more value choices. In fact, the opposite side of substitution is complementarity. rdquo; Li Lixing said.
The “2025 Human Resources Development Trend Report” jointly released by Zhaopin Recruitment and the National Development Research Institute of Peking University shows thatIn the first half of 2024, recruitment demand for new quality industries accounted for 21.8% of the total. More than half of the companies surveyed throughout the year deployed AI technology research and development and application. Among them, 27.2% of the companies surveyed expect that AI technology R & D investment will account for the total revenue. The proportion is more than 3%. In addition, the survey also found that 59.1% of the companies surveyed said that super individuals are scarce and talent introduction is difficult, and companies are increasingly paying attention to the multi-dimensional development potential of talents.
On February 28, at the artificial intelligence and future workplace event held by Peking University National Development Institute and Zhaopin Recruitment, Li Lixing further analyzed the impact of AI technology progress on the labor market.
Li Lixing pointed out that technological progress is divided into general and special types, which produce three effects: direct substitution, indirect promotion and overall creation at the task, occupation and industry levels. It also has three effects: substitution, complementarity and creation on individuals. If technology can be used to improve efficiency and create new positions, human-machine collaboration can have a positive effect. heCiting a report, before 2030, 8 million and 12 million jobs in Europe and the United States will need to be transformed (old jobs will disappear) respectively, accounting for 4.6% and 7.5% of the total number of jobs.
Li Lixing said that in the process of technological progress, benefiting from new technologies, the employment field will flourish, the production efficiency of enterprises will be improved, the products will be richer, and the prices will be cheaper, and the market demand will increase. Even if other traditional industries are not directly affected, due to their relatively backward efficiency, they will be gradually squeezed. Therefore, for talents, it will have effects such as substitution, complementarity, and creation, and everyone will be affected by new technologies such as AI.
As AI becomes capable of more basic tasks, companies are beginning to rely on the soft skills of human employees.
Human resources development trends report shows thatThis year, 51.6% of companies surveyed have increased their soft skills requirements in recruitment, such as good self-management skills, communication skills, etc. Zhu Li, an associate researcher at the National Development Research Institute of Peking University, said that people with soft skills can alleviate the impact of AI. In the AI era, professional depth, management ability, initiative and self-management will become the core differentiated advantages of career development.
On the other hand, AI technology is reshaping the entire human resource management (HR) process, with more than 28% of companies applying AI or large model technology in corporate human resource management. Currently, the DeepSeek-R1 version of Zhaopin Recruitment AI Recruitment Assistant has been launched on February 27. In the financial field, more than 80% of banks are using Zhaopin Recruitment’s AI Easy-Face products.
Zhang Yuejia said that HR needs to take the initiative to change, change from a transactional to a strategic one, have digital thinking and AI capabilities, use digital tools to arm itself, and achieve intelligent upgrades in recruitment, talent evaluation and other aspects.
“We can see that the entire enterprise’s talent needs and structure are undergoing some changes, and the needs of enterprises when identifying and recruiting talents are also undergoing some changes. rdquo; Zhang Yuejia said.
According to the latest McKinsey report,By 2030, China’s demand for AI professionals is expected to reach 6 million, and the talent gap may reach 4 million。
Huang Yiping, dean of the National Development Research Institute of Peking University, dean of the School of South-South Cooperation and Development, and distinguished professor of Peking University Boya, said that AI is a new opportunity for economic growth and will also bring challenges to the restructuring of the labor market.
IDC statistics show thatBy 2030, AI-related industries and services will have a cumulative global economic impact of US$19.9 trillion.It will also boost global GDP by 3.5%.
