Editor: Editorial Office
Liepin Big Data Research Institute released the "AIGC Employment Trend Big Data Report 2023", the average annual salary of recruitment has reached 400,000, and the demand for doctors has increased by more than 100% year-on-year.
Needless to say, everyone knows how popular ChatGPT has been in the past six months.
With the global popularity of ChatGPT, AIGC has also successfully broken the circle from the field of science and technology to become a national topic across all circles.
AIGC's subversion of graphics, text and video creation has had a profound impact on related industries, and various predictions emerge one after another.
In the discussion of different opinions, computer/AI and other related majors have also become a hot spot for college entrance examination volunteers.
Just recently, Liepin Big Data Research Institute released the latest AIGC Employment Trend Big Data Report 2023.
The report provides a comprehensive analysis of employment opportunities, salaries, talent reserves and delivery in the AIGC field, and combines relevant expert interviews to reveal the skills and qualities necessary to enter the AIGC field.
1. AIGC talent demand analysis
1. The demand for AIGC talent in the first quarter of 2023 is 5.6 times that of the same period three years ago, while AI is less than twice that of AI
Liepin Big Data Research Institute took the number of new jobs in AIGC and AI in the first quarter of 2020 as the base, and divided the number of new jobs in the two fields from the quarter to the first quarter of 2023 from their respective bases to observe the growth trend of their talent demand.
The comparison found that from the first quarter of 2020 to the first quarter of 2021, the job growth of AIGC and AI was comparable.
Since then, AIGC has generally continued to grow, and the number of new jobs in the first quarter of 2023 is 5.63 times that of the first quarter of 2020.
In contrast, the growth rate of AI new jobs has slowed down relatively slowly, 1.95 times that of the first quarter of 2020.
2. In the past year, AIGC's new positions have increased by more than 43% year-on-year, and the average annual salary of recruitment has exceeded 400,000
In the past year (June 2022-May 2023), the number of new jobs in AIGC has increased significantly, with an increase of 43.66% compared with the previous year (June 2021-May 2022).
In the past year, the average annual salary of AIGC new job recruitment was 401,200, which was 80,900 higher than the average annual salary of AI recruitment in the same period (320,300).
3. AIGC Job Opportunity Direction Analysis
1) Nearly 50% of AIGC positions are distributed in the Internet industry, with an average annual salary of 430,000
In the past year, the three industries with the largest distribution of AIGC positions were IT/Internet/gaming, automotive, electronics/communications/semiconductors, accounting for 49.13%, 17.59% and 6.63%; The corresponding average annual salary of recruitment is 432,300, 346,500 and 428,300.
2) AIGC jobs in the research technology/business services industry grew the fastest year-on-year, with a growth rate of more than 200%
From the three industries with the fastest year-on-year growth in AIGC new jobs in the past year, AIGC jobs in the scientific research technology/business services industry ranked first, with 211.86%.
The foundation of scientific research technology/business services lies in the quality of R&D and services, and AIGC provides a more effective and convenient path for them.
AIGC's job growth in the energy/chemical/environmental protection, IT/Internet/game industries ranked second and third year-on-year, with 120.99% and 61.88%.
The average annual salary of AIGC job recruitment in these three industries is higher, 503,500, 337,600 and 432,300 respectively.
3) Algorithm engineers in the AIGC field are in the greatest demand, with an average annual salary of 450,000
From the top 10 distribution of AIGC new positions in the past year, the top three are algorithm engineers, natural language processing (NLP), and product managers, accounting for 14.67%, 7.37%, and 5.40%.
Among the TOP10 functions, the highest average annual salary is image algorithm, with 556,200; the average annual salary of deep learning, natural language processing (NLP), machine vision, and machine learning is more than 500,000; Algorithm engineers came in sixth with 450,500.
The recruitment salary of these ten functions is generally higher, especially the technical functions are more advantageous, which is closely related to AIGC's high entry threshold and talent scarcity.
4) Beijing and Shanghai AIGC have the most positions, and the average annual salary of Beijing AIGC job recruitment exceeds 470,000 yuan
Among the top 10 cities with new AIGC jobs in the past year, Beijing and Shanghai have the most jobs, accounting for the first and second proportions, accounting for 22.21% and 20.37%.
Shenzhen and Hangzhou ranked third and fourth, accounting for 11.75% and 10.15%. Guangzhou and Suzhou ranked fifth and sixth with 3.90% and 3.84% respectively.
In terms of the average annual salary of recruitment in these top 10 cities, Beijing, Shenzhen, Nanjing and Shanghai ranked in the top four, with 471,900, 463,500, 430,600 and 427,400 respectively.
Hangzhou and Guangzhou ranked fifth and sixth with average annual salaries of 394,400 and 375,000.
Li Zhifei, founder and CEO of Mobvoi and former Google headquarters scientist, said that AIGC tool attributes reconstruct the workflow of knowledge creation content and bring new possibilities and business models to the AI industry, while reducing costs and improving efficiency of content production is to build market increments for large-scale production.
Second, AIGC recruitment threshold analysis
1. The proportion of positions requiring 3-5 years of experience exceeds 36%, and the demand for talents with 5-10 years of experience exceeds 70% year-on-year
From the distribution of work experience requirements for AIGC new positions in the past year, 3-5 years of work experience positions are the most, accounting for 36.23%; Followed by 5-10 years of work experience, accounting for 23.29%.
AIGC's demand for talents in 5-10 years has grown the fastest, with a year-on-year increase of 70.75% in the past year; AIGC talent demand with 3-5 years of experience ranked second in terms of growth, with a year-on-year increase of 57.61%.
It can be seen that AIGC recruiters prefer practitioners with certain work experience.
2. AIGC is more eager for highly educated talents, and the demand for doctoral talents has increased by more than 100% year-on-year
In the past year, the demand for undergraduate talents in AIGC and AI new positions accounted for 70.80% and 70.43%.
In terms of demand for higher education, AIGC is significantly higher than AI, and the former's demand for masters and doctoral degrees accounts for a total of 21.56%; The latter is 14.24%.
AIGC is more eager for doctoral talents, and its demand has grown the fastest among all academic qualifications, with a year-on-year increase of 108.11% in the past year.
3. Analysis of talents of interest to AIGC companies
1) The most favorable function of AIGC enterprises: algorithm engineers
In the past year, among the top 10 talents in the three-level functional distribution of AIGC-related enterprises who actively communicate, algorithm engineers lead with 10.83%.
Product managers ranked second, accounting for 3.37%; Natural language processing (NLP) and intelligent network engineers ranked third and fourth, accounting for 2.38% and 2.16%.
It can be seen that algorithm engineers are most favored by AIGC enterprises.
2) AIGC companies are most interested in talents from technology companies, and the number of people reached is the first with Huawei's background
Judging from the top 5 distribution of talent source companies actively communicated by AIGC enterprises in the past year, Huawei ranks first.
Ranking second to fifth are Baidu, Tencent, ByteDance, and Meituan.
4. Hard technology and soft capabilities required to enter AIGC
According to Li Zhifei, the hard skills required to enter AIGC include:
- Familiar with the basic principles and algorithms of machine learning and deep learning;
- Understand the basic concepts and techniques of natural language processing, including text analysis, text generation, etc.;
- Good data processing and data analysis capabilities, including data cleaning, feature engineering, etc.;
- Good programming skills, proficient in using Python, Java, C++ and other programming languages, as well as good software engineering skills, including version control, code specification, testing and debugging, etc.;
- Innovative thinking that can constantly explore new technologies and applications;
- and problem-solving, communication and teamwork skills.
3. AIGC talent reserve analysis
1. In the past year, the number of talents under the age of 25 in the field of AIGC has increased the most year-on-year, exceeding 60%
In the past year, the overall talent growth of AIGC was 19.53% year-on-year.
In terms of age groups, the proportion of talents under the age of 30 increased from 31.64% to 35.61%, an increase of 31.70% over the previous year.
Among them, the number of talents under the age of 25 increased by 61.90% year-on-year, the largest increase among all age groups; The number of talents aged 25-30 increased by 27.62% over the previous year. It can be seen that there is a significant increase in young people among AIGC's talents.
Despite this, talents aged 30-35 still account for the largest proportion, at 35.77%.
2. AIGC talents have better educational backgrounds than AI, accounting for nearly 45% of masters and PhDs
In the distribution of AIGC talents in the past year, undergraduate accounted for the largest proportion, 48.49%. Master's and doctoral degrees accounted for 42.20% and 2.79% respectively, totaling 44.99%.
Among AI talents, 54.09% are undergraduates, and 23.90% are masters and PhDs (22.32% for masters and 1.58% for doctors), which is far lower than AIGC.
3. Beijing and Shanghai have the most AIGC talents, with a total of nearly 50%
In the TOP10 distribution of AIGC talent cities in the past year, Beijing and Shanghai ranked the top two, accounting for 26.01% and 23.34%, a total of 49.35%.
Shenzhen ranked third, accounting for 11.90%. Hangzhou and Guangzhou ranked fourth and fifth with 5.78% and 5.27% of the total.
Hangzhou is an important Internet town, the headquarters of several Internet manufacturers, and there are many AI-related companies, so the AIGC talent reserve is relatively large.
4. AIGC talents have higher education and senior practitioners have higher salaries
1) The average annual salary of AIGC talents exceeds 400,000, and the average annual salary of doctors exceeds 750,000
In the past year, the average annual salary of AIGC talents was 401,200 yuan, which was 121,900 yuan higher than that of AI (279,300).
The salary of AIGC talents is directly proportional to the level of their education, and the average annual salary of college, bachelor's, master's and doctoral degrees has gradually increased, to 251,100, 358,200, 443,300 and 758,600 respectively.
2) After AIGC talents reach 15 years of experience, the salary ushered in a big explosion, with an average annual salary of more than 670,000
AIGC talent salaries and experience are on the rise. The average annual salary of AIGC talents with less than 5 years is less than 260,000.
In 5-8 years, it exceeded the 300,000 mark, reaching 327,700; the average annual salary of AIGC talents with 10-15 years of experience approached 500,000, which was 490,300.
The average annual salary of AIGC talents with more than 15 years is as high as 674,100. In this sense, experience is wealth.
5. AIGC talent source analysis
1) The proportion of talents from the Internet industry is the first, and the proportion of talents of Internet giants is high
Judging from the top 10 secondary industry distribution where AIGC talents last worked in the past year, talents from the Internet industry accounted for the most, accounting for 12.67%; Computer software and vehicle manufacturing ranked second and third, accounting for 9.78% and 9.47%.
The top 5 companies where AIGC talents last worked were ByteDance, Huawei, Alibaba, Baidu, and Tencent, and the distribution of talent companies that actively communicated with AIGC enterprises overlapped with the top 5, which further confirmed that the experience of high-tech manufacturers is easier to enter the AIGC field.
2) The most people worked as product managers and algorithm engineers in their previous jobs
From the top 10 three-level functional distribution of AIGC talents' last job in the past year, product managers and algorithm engineers ranked in the top two, accounting for 6.91% and 5.35%. Other functions account for less than 5 per cent.
In this regard, Li Zhifei said that due to AIGC's increasingly diversified demand for talents. In addition to traditional computer science and data science talents, AIGC also needs professionals with knowledge and experience in AI model optimization, natural language processing, robot operating systems, and more.
4. Analysis of talent job search behavior of AIGC
1. In the past year, the number of talents delivering AIGC has increased by more than 270% year-on-year, which is 13 times the growth rate of the number of AI deliveries
As AIGC heats up globally, the number of people delivering in this field is soaring, and the number of deliveries is growing much faster than AI.
Liepin big data shows that the number of talents who delivered AIGC in the past year increased by 274.73% year-on-year, which is 13 times the growth rate of the number of AI deliveries in the same period (21.09%).
2. The largest number of natural language processing and product managers in the AIGC field, accounting for nearly 20%
According to the distribution of the top 10 third-level functions with the largest number of deliveries from AIGC in the past year, the top three functions are natural language processing (NLP), product managers, and algorithm engineers, accounting for 9.35%, 9.30%, and 8.15%.
It is worth noting that content operations and new media operations are also among the top 10.
In terms of explicit impact, AIGC is more helpful for content and new media, so related positions are more attractive to job seekers.
For example, Xin Zhiyuan's editorial post is hiring.
3. Companies submitted by AIGC talents: Internet and artificial intelligence companies account for the majority of the country
In the top 10 distribution of companies posting AIGC jobs in the past year, Internet giants occupy four seats, of which Baidu, Ali, ByteDance, and Tencent ranked second, fifth, sixth, and eighth respectively.
Among them, the computer hardware company Hitevision Technology ranked first; AI companies SenseTime, Lingxin Intelligent and AI institutions Guangdong-Hong Kong-Macao Greater Bay Area Digital Economy Research Institute (Futian) ranked third, seventh and tenth; Internet company Kunlun Wanwei ranked fourth.
Manufacturing company Sany Group came in ninth. It can be seen that the number of job seekers applying for AIGC positions is concentrated in Internet and artificial intelligence companies.
In the face of AIGC's powerful technology, many professionals are deeply worried about how to keep their "jobs".
Li Zhifei pointed out that there are two types of occupations that are easiest to replace AIGC, one is a profession that can complete the closed-loop work on the computer and has a large number of repetitive links in the work content, such as basic art/design; The other category is occupations that are easy to standardize, such as customer service and marketing copywriting using fixed phone techniques.
The least likely to be replaced are occupations that require face-to-face interaction and physical skills, such as craftsmen such as plasterers, electricians, and mechanics, and service workers such as hairdressers, cooks, doctors and nurses.
Li Zhifei suggested that professionals should consciously cultivate skills that are difficult to replace with AIGC, such as innovative thinking, the ability to solve complex problems, and interpersonal skills.
ChatGPT is on fire, but what is taught in school has little to do with it
Coincidentally, some time ago, a question on the hot search of Zhihu also reflects the further improvement of academic requirements in the field of AIGC - what are the big models all on, and those who learn AI can not find a job?
In this regard, a friend pointed out the reason: I learned to weave cloth by hand, and now the Jenny loom is on fire...
ChatGPT's core technology "Transformer" was published in 2017, and the undergraduate students majoring in artificial intelligence who graduated this year entered the school in 2019.
The courses of artificial intelligence majors in domestic universities and computer majors are very similar.
In the first year, almost all artificial intelligence students will learn basic programming languages, and by the sophomore year, they will be exposed to web design, front-end, data structures and other courses.
It was not until the junior year that he began to contact neural networks, deep learning, intelligent speech, image recognition, and other courses with artificial intelligence majors.
Most likely, content related to "Transformer" may not have the opportunity to be learned at all.
Therefore, the biggest problem affecting the employment of AI undergraduates is that the content of school learning and the actual ability required by the industry are too seriously disconnected.
Moreover, this disconnect is also reflected in the fact that AI companies also fail to recruit useful employees.
The lag in education has been amplified, leading to a serious decoupling of industry and academia.
Many positions now require proficiency in programming, data structures and algorithms, high-number line algebra probability theory, compilation principles, machine learning, deep learning... The comprehensive quality requirements of talents are extremely high.
On the other hand, the explosive development of the artificial intelligence industry this year is brought about by a group of high-end technical talents.
The current stage of development of the industry itself urgently lacks high-end technical talents.
The knowledge reserve and practical experience of undergraduates are difficult to meet the industry's requirements for talents.
The development of industry technology is a thousand miles a day, and the content that undergraduates learn most of the time may not be understood yet, and it is already useless.
And artificial intelligence is different from the Internet, and the winner-take-all model requires higher technology and talents, but does not require a lot of labor in basic positions.
This has led to undergraduate graduates who do not have the advantage of academic qualifications in the embarrassing position of "high is not low".
How to become a qualified AI engineer?
On Zhihu, a senior HR of an Internet listed company, Sister Huohuo, gave some suggestions to those who are interested in entering the artificial intelligence industry.
The editor finds it very pertinent and paraphrases it as follows:
In this way, whether you are studying or preparing to enter the artificial intelligence industry, you still have to go to a graduate school.