It's easy to set up an AI major, but not easy to cultivate talents

Written by Chen Dengxin

Edited by Li Ji

Typeset by Annalee

The 2024 autumn recruitment season is drawing to a close.

In recent years, the number of college graduates has been increasing year by year, and it is expected that there will be up to 11.86 million fresh graduates in 2025, setting a new record high.

This means that job competition is becoming increasingly fierce.

It should be noted that although many places have relaxed the criteria for identifying fresh graduates, including those who graduated within the past 2 to 3 years, most companies still prefer to recruit 2025 fresh graduates who are still on campus.

More importantly, as large models are gradually implemented, AI empowering various industries has become a consensus across society, making AI talents highly sought after, and AI has become the "protagonist" of this autumn recruitment season.

Under the demand for AI, some are happy while others are sad.

Unable to secure offers from top companies

"When people say they have no offers, they mean they have no offers from top companies!" said Li Zhongyan.

Li Zhongyan, a fresh graduate of electronic information engineering from a 985 university studying in Nanjing, pays special attention to the autumn recruitment season. If he can secure an offer earlier, he will be able to face society more calmly.

After all, nothing is easy once you leave the university gates.

Li Zhongyan told Xin Ke Du that he started submitting resumes as early as the end of August, targeting internet companies and aspiring to go where the money is. However, the autumn recruitment season in 2024 focuses more on technical and R&D positions, especially those related to AI, which are in high demand. "This year's requirements are so high that nothing seems to matter in the face of technological development," he said.

Xin Ke Du's investigation found that companies like Baidu, JD.com, Tencent, Alibaba, and ByteDance are eagerly seeking AI talents.

For example, Baidu's 2025 campus recruitment offers over 3,000 positions, about 80% of which are related to AI technology. The most in-demand positions are machine learning, data mining, natural language processing, etc. Notably, there is a new campus recruitment position for large model algorithm engineers, who can participate in the development of large-scale pre-trained models after joining.

Image source: Baidu Campus Recruitment

Another example is ByteDance's 2025 campus recruitment, which offers over 4,000 positions across eight categories including R&D, operations, and products. The demand for R&D positions has increased by 60% compared to last year, with algorithm positions being in high demand.

It is not difficult to see that algorithms are in the spotlight.

The "2023 Internet Industry Talent Mobility Report" shows that algorithm researchers rank first in talent scarcity with a talent supply-demand ratio of 0.47, with an average of two companies competing for one talent.

According to Zhaopin data, the average monthly salary for enterprise recruitment in 38 cities was 10,313 yuan in the second quarter of 2024. Among them, the AI industry topped the salary list with an average monthly salary of 13,594 yuan, a year-on-year increase of 5%.

Image source: Zhaopin Research Institute

Against this backdrop, Li Zhongyan was unable to fulfill his wishes.

"I applied for over a hundred positions, but basically all were rejected." Li Zhongyan eventually received an offer from a technology company in Xi'an focusing on hardware R&D. "I've made it, but not completely," he said.

The quant industry does not have a "996" work schedule

Unlike Li Zhongyan, Meng Yijun is more confident.

Majoring in computer science and having accumulated internship experience at internet giants during his undergraduate studies, Meng Yijun is somewhat confident about securing an offer as an AI heterogeneous computing engineer or a deep learning high-performance computing R&D engineer.

The problem is that Meng Yijun wants to work in the quant industry.

Quantification refers to digitizing signals in financial investment behavior, analyzing them through one or more algorithms, and making judgments and decisions, rather than relying solely on human feelings or intuition.

In a word, AI is the foundation of quantification.

Since the strength of AI capabilities is closely related to quantification performance, top quant funds are willing to offer high salaries to attract talents, even fresh graduates, in order to maintain their leading position and technological advantage.

For example, Jiukun Investment launched the "Wutong Plan" in 2022 to recruit outstanding undergraduate and graduate students globally, selecting "quant stars" who can work independently.

"Don't go where there are a lot of people; you might end up doing menial tasks like data cleaning," Meng Yijun said, believing that quantitative research is more challenging. "There's no 996 in quant; you can have a work-life balance."

Despite this, Meng Yijun is still undecided.

On the one hand, the requirements of quant funds are not lower than those of internet companies, and are even higher to some extent, demanding visible contributions to excess returns.

Zhang Chenying, the investment director of Bailu Asset Management, said in an interview with "An Yong Waves" that "in the current Chinese quant private equity industry, having a large number of winners of international Olympiad medals, such as IMO (Mathematics Olympiad), IPHO (International Physics Olympiad), ICHO (International Chemistry Olympiad), and IOI (International Olympiad in Informatics), has long been the industry standard."

On the other hand, the essence of quantification is a zero-sum game, which has caused dissatisfaction in the market and encountered "constraints." Coupled with signs that the A-share market is transitioning from a bear market to a bull market, the strategies of quant funds are facing severe challenges.

Image source: Tiantian Fund Network

On October 18, 2024, Huanfang Quantification sent a message to investors: "Due to changes in the market environment, it is difficult for hedge fund products to achieve both returns and reduce risk exposure simultaneously, resulting in a significant decline in the potential return-to-risk ratio. Future returns will be significantly lower than investors' expectations. Considering the risk attributes of hedge products, the company will gradually reduce the investment positions of the entire hedge fund product line to zero."

Meng Yijun further told Xin Ke Du that the new energy vehicle industry is another popular employment direction for AI.

The "Intelligent and Connected Vehicle Industry Talent Demand Forecast Report" shows that the total demand for intelligent and connected vehicle talents in China will reach 116,000 by 2025, with a net talent gap of up to 37,000.

This can be seen from the campus recruitment of automakers.

Taking Lixiang Auto as an example, it has launched the "Lixiang+" campus recruitment plan targeting excellent AI talents worldwide, covering areas such as algorithm deployment, algorithm performance optimization, inference acceleration, large language models, visual large models, multimodal perception, AIGC, intelligent driving large models, and AI chip R&D.

Lixiang Auto 2025 campus recruitment algorithm R&D position process

Li Xiang, CEO of Lixiang Auto, said at the 2024 spring conference, "We are willing to pay higher salaries so that every employee at Lixiang Auto can grow, achieve, and be rewarded."

In fact, AI talents have become the focus and highlight of automaker recruitment.

In this regard, Musk said in April 2024 that Tesla is increasing salaries for its AI team, "The AI talent war is the craziest talent war I've ever seen!"

Those who make good use of AI will win

In addition to campus recruitment, how to cultivate AI talents has become another focus.

"AI large models are the best tool. Parents should expose their children to AI from a young age. Children who do not use AI will lose to those who do," said Zhou Hongyi. He added that if someone recruited by 360 cannot proficiently use AI, they will definitely be eliminated. "Whether students can use AI and their ability to use it will determine their future development potential and growth limit."

As a result, establishing an AI major has become a consensus among universities.

Public information shows that in 2018, 35 universities in China offered AI majors, and by 2024, the number exceeded 500, ranging from prestigious universities to vocational colleges.

Image source: Zhangshang College Entrance Examination

Zhang Yu, a teacher at a university, told Xin Ke Du, "Our school has strengthened the construction of AI-related majors and disciplines, added AI-related courses, and established an AI laboratory to promote the transformation and application of scientific research results."

Teacher Zhang Yu further stated that AI requires highly interdisciplinary and compound talents with high demands on knowledge accumulation, technical proficiency, and experience. Simply adjusting parameters and calling models is not enough for high-paying positions. "In actual work, stacking model parameters and adjusting parameters are two completely different things."

In other words, it's easy to set up an AI major, but not easy to cultivate AI talents.

China News Weekly reported that after searching the training programs of several second- and third-tier universities, it was found that the new AI major training programs were extremely similar to the existing computer science, electronic information, data science, and other majors, with only one or two course differences for most of them.

In fact, Teacher Zhang Xuefeng previously believed that students studying AI must become T-shaped talents with one specialty and multiple strengths. The "one" represents the breadth of knowledge, and the "|" represents the depth of knowledge.

As a result, there is a gap between universities and enterprises in terms of transformation.

Qi Yuan, Dean of the Institute of Artificial Intelligence Innovation and Industry at Fudan University, said in an interview with the 21st Century Business Herald, "In the context of rapid technological development, education is also facing a certain transformation. There are two main issues here. First, how to deepen students' foundational theories. Second, how to enhance students' practical and analytical abilities and align them with the needs of enterprises."

From this perspective, universities need to more precisely cultivate AI talents and improve the quality of AI talent cultivation, so that fresh graduates can embark on a more promising future.

Undoubtedly, AI will become a keyword in future campus recruitment.

(Upon request of the interviewees, the names Li Zhongyan, Meng Yijun, and Zhang Yu are pseudonyms in this article)