"The Python talent gap is 300,000, the salary is high and there are no friends," and "there is zero basic entry, 90 days entry to high salary." With the hot artificial intelligence, artificial intelligence short-term training courses are also alive and well, frequently 20,000 to 30,000 training. Fees, consultations are endless, and high salaries become the most attractive slogans, these agencies without exception, the salary will be placed in the most prominent position.
In the past year, due to the rapid outbreak of the artificial intelligence industry, a huge talent shortage has pushed the salary of the industry personnel to an embarrassing level. How to retain and manage talents behind grabbing talent has also become a new challenge.
From fresh graduates to skilled cattle
“The annual doctoral salary in the field of artificial intelligence is about 500,000, and the master's is from 250,000 to 350,000. This salary level is equivalent to the salary of an ordinary post engineer with 2 to 3 years of work experience.†Co-founder of fluency, artificial intelligence company Dr. Lin Hui, Chief Scientist, revealed to First Financial that in October this year, the company launched the school's first move and succeeded in recruiting more than a dozen graduates from Tsinghua University, Peking University and Fudan University. These are from the Department of Computer Science and Mathematics. Students majoring in statistics and statistics often hold multiple offers before they graduate.
IDG Capital's “2017 Internet Unilateral Unicorn Remuneration Report†shows that the senior position compensation in the artificial intelligence industry is 55% higher than the overall level, 90% higher for mid-level positions, and 110% for junior positions, and one is artificial. The war for talents triggered by intelligence is imminent, and competition for talents in the existing market has turned white.
“If you know that one person is a Ph.D. student and he is reading a neural network, he is basically destined to do research for a lifetime because he will never be able to do it.†Dr. Fang Mu, the robot CTO of the future, mentioned that five years ago, people often talked about it. A joke, Fang Mu is an undergraduate at Zhejiang University, a master at Shanghai Jiaotong University, and a Ph.D. student at the Chinese University of Hong Kong. She is engaged in research and development in the field of machine vision and robot navigation. She has been expressing her emotion for only five years. Now no one knows about deep learning. , And the salary of students is not comparable.
Apart from fresh graduates, the competition for AI senior talent is even more fierce. It includes the excavation of company talents and the snatching of academic talents from colleges and universities. At this year’s Yunqi conference, Alibaba Cloud’s chief scientist Qian Wanli disclosed that he had received more than 700 emails from headhunters in one year.
In order to rob senior talents, fluency said that in September this year, the AI ​​Lab was set up in the United States. The goal was to tap "influential talents in the academic and industrial fields." According to Lin Hui, the number of Chinese AI talents does not detract from the United States, and the number of papers published is particularly evident. However, from a qualitative point of view, leading figures are more in the United States, and most of these big cattle have already settled in the United States. The cost to attract both to China is relatively high, and establishing a laboratory in the local area is the best choice.
In May of this year, Tencent dubbed the top expert in speech recognition technology Dr. Yu Dong as the deputy director of AILab, and set up the Seattle AI laboratory in the United States. In November this year, Zhang Jian, director of QQ, AILab, disclosed to reporters that the US laboratories are hiring. There may be 20 to 30 people next year, focusing on cutting-edge technology research. This also means that these Internet companies are competing with global competitors for AI cutting-edge talent.
Go to Silicon Valley to dig people
The AI ​​talents snatched the turmoil of the war and the headhunting industry. In the past year, many headhunting companies launched AI headhunting services, among which Silicon Valley talents were particularly sought-after.
“People in Silicon Valley are very expensive. Facebook's doctoral graduates for two to three years are about 400,000 to 500,000 US dollars. If they go back to BAT companies, they generally hope to have a 30% increase.†Riton Consulting Strategic Recruiting Director Zhou Wei told CBN that during this time he was helping a Chinese unicorn company to find a Chinese person with extensive AI experience as the AILab dean. The request was the chief scientific technology or AI expert of the Internet company. Strong academic offensive and commercial application ability, published in the world's top journal papers, quite industry appeal.
AlexRen, Managing Partner of Capital Capital Silicon Valley, founded TalentSeer last year and is mainly responsible for the high-end AI search service between China and the United States. They are looking for talents in AI chips, driverless robots, dialogue robots, cloud service networking, and deep learning platforms. In the year they helped more than 50 companies recruit nearly 100 people. The team also expanded from a few people to more than 20 in just one year.
According to his experience, the technical backbone of master's or doctoral work for three to seven years is the most sought-after talent. A typical AI team needs 5 to 10 people, and the average salary is between 250,000 and 350,000 US dollars, including 20% ​​bonuses. In addition to stock options, the cost of the entire team is about two to three million. The talents in the unmanned area are particularly sought-after. They often receive four offers in a week. The headhunter must give job opportunities at a faster rate. they.
In the era of artificial intelligence, the effect of “talent attracting talents†has become more apparent. Selecting a bully with great appeal in the industry will greatly help subsequent attracting other talents into the company.
However, the work of excavating top talents is not a good job. "The inherent salary differences between China and the United States, especially the so-called 'FLAG' (Facebook, linkedIn, Apple, Google) represent the company's stock and options are all in place. It is not easy to attract talents,†said Zhou Wei. “The cycle of excavating talents often takes six months or even one year.â€
This also puts forward higher requirements for headhunter work. In contrast to most previous headhunters that communicate via linkedIn channels and target audiences, headhunters must understand the academic background of each individual, which Labs they have worked on, which papers they publish, and the focus of academic research, and the corresponding companies and companies. Business differences, product direction, and industry coverage differences also need to be well-known, so that they really enter the "circle."
In order to understand that the unmanned vehicle industry AlexRen has found nearly 40 industry experts in two months, each time more than one hour of exchange learning, mutually confirm the technical points and the ecological relationship between enterprises. In addition, every month, within the team, there is a headhunter sharing learning results and looking for industry experts to conduct internal lectures.
2018 Silicon Valley Return Year
China's huge market opportunities and vast amounts of data are not attractive for overseas talents. After graduating from the University of Washington’s Department of Electronic Engineering, Lin Hui entered the US headquarters of Google as a research scientist. In 2012, he decided to return to China to start a business. In his view, “China has a large number of C-end data that can help product iterations and put existing The technology is even more cattle. At the same time, the demographic dividend also has obvious advantages in finding people to label data. These are very attractive to R&D personnel of scientific research and they are also not available in the United States."
Unparalleled CTO Technology co-founder Matt Scott has served as senior research and development director at Microsoft Research Asia. As an American, he eventually chose to start a business in China. “Besides the data, China’s funding for artificial intelligence ventures and government support are incomparable. From the national level to the local level, there are very clear plans for artificial intelligence. In his view, the future of artificial intelligence is in China.†Dr. Huang Dinglong of Science and Technology told China Finance.
In China, artificial intelligence has been put on the national agenda, and it has become the world's second largest area for artificial intelligence, second only to the United States. According to the Wuzhen Index “Global Artificial Intelligence Development Report 2017†data, from 2000 to 2016, China’s artificial intelligence financing has reached a total of US$2.76 billion. Among them, the three years from 2014 to 2016 are the most rapid development periods of artificial intelligence in China. In the three years, artificial intelligence financing accounted for 93.59% of the total, and the investment frequency accounted for 87.22% of the total, much higher than Israel and India.
Due to the scramble for advanced technology talents by artificial intelligence technology and the demonstration effect of overseas entrepreneurs such as Facebook and Google to return to China for successful entrepreneurship, together with the expiration of four-year options and stock grants for the first batch of Chinese employees of Silicon Valley companies including Pinterest and Uber, Silicon Valley once again rumors that the two companies will be listed next year. IDG Capital believes that 2018 will be the return year for Silicon Valley technology professionals.
Chinese AI companies have higher valuations
It is difficult to recruit people, but how to retain talents and transform academic research and engineering capabilities into commercial values ​​requires a more in-depth process. For these academic big cows, what is more important than money is recognition of what the company does. "Many AI talents will have obvious tendency to consider their own research direction and technology application subdivision areas. For example, a US group candidate clearly indicated that he hopes to enter the education field." Zhou Wei mentioned.
AlexRen also said that there are two misunderstandings in the recruitment of Silicon Valley talents by many Chinese companies. One is that they are not well-informed and arrogant about the talent structure and culture of the United States. The other is the blind belief in Silicon Valley's talent, especially the blind belief in the talents of the traditional large companies in Silicon Valley. Their senior management identities are due to individual capabilities or platform causes. They do not understand their areas of expertise and their advantages.
The running-in of talents and companies is also a long-term concern for headhunters in the later period. In Zhou Wei's impressions, there are indeed cases where one beats and two falls after the break-in. Some people enter the company as professional managers, but the direction that the boss wants to develop is not recognized by him. He has never entered the core project, and there are also many CEOs who are impetuous. After recruiting talents, they are eager to see the output immediately and eventually part ways. .
Attracted a large number of outstanding talents into the enterprise and put forward higher requirements for talent management. Lin Hui admitted that the cost of talent is relatively conservative. “As a company, it is more desirable that technology will generate commercial value in products and iterates faster, but it is necessary to retain talent. The key is to make it clear that the company's management must think clearly about the technical planning and direction, and then dismantle it into concrete targets and tasks to each individual."
For investors, talent is the core factor for evaluating an AI company, even the primary factor. Wei Feng, a founding partner of Coco Capital, who has invested in a number of investments in the field of machine intelligence, told China Business News, “Investment is investment, and the role of people in the AI ​​era is even more prominent, but from the perspective of industrial capital investment, it is more important There needs to be a balance between technology and commercialization, the spirit of scientists and engineering capabilities."
In the investment, Cocoa Capital made a 6 to 4 reference standard for technical talents and commercialization capabilities. "We cannot completely separate from the market demand and consider purely scientific research. We need to implement the technology to the stage where the company can use it. Understanding the needs of customers, partnering with supply chain partners, and team management are extremely difficult processes," said Coco Capital Partners Li Yukai.
In his view, this also requires that scientists themselves have a good ability to work, or understand the risks of technological productization, have the commercialization of the face to face this matter. In the process of investing in Wing Fei Automation to help companies find talents, Li Yukai found a professor and doctoral supervisor Liu Xinjun of the Department of Mechanical Engineering at Tsinghua University. Professor Liu Xinjun clearly stated to him: "The results of scientific research are connected with entrepreneurs through the transformation of achievements. I can promise to Industrialization, I am very sorry, because I am not an entrepreneur."
Li Xiaokai, who is studying for a bachelor of engineering and master of engineering at Tsinghua University, understands that there is often a gap between scientific research and business needs. There are some constraints on doing papers in colleges, but in the actual operation of the company, the constraints will not necessarily be conceived. To create a constraint, adjustments and balances need to be made according to industry conditions.
The development of artificial intelligence industry blowout, imbalance between supply and demand is the main factor of AI talent shortage. According to the "Global AI Field Talent Report" published by linkedIn, as of the first quarter of 2017, the United States had the largest pool of talents, with more than 850,000, while China's figure only exceeded 50,000, ranking seventh in the world.
AlexRen estimates that the ratio of supply and demand for artificial intelligence talent in the United States is about 1:3, while China is about 1:6 or even 1:10. He has invested in 12 artificial intelligence companies in the United States. Overall, Chinese companies are valued two to four times higher than the United States. Times.
Comcn Electronics Limited , https://www.comencnspeaker.com