Guide
For start-up companies and big companies in AI field, it is very important to acquire the top-level AI talents and have the continuous transmission ability of the talents, while the start-up companies are facing a lot of pressure at the grab level.
Due to algorithmic breakthroughs, policy support, capital swarming, artificial intelligence industry ushered in the stride in 2017, AI start-ups continue to emerge, but the scarcity of professionals now has become a bottleneck impeding the further development of AI start-ups.
At the beginning of December 2017, the "White Paper on Global AI Talents 2017" released by Tencent Research Institute in collaboration with BOSS directly showed that there are about 300,000 global AI professionals in the world with a market demand in the order of one million. There are more than 300 all AIs in the world The direction of colleges and universities, the annual supply of AI field of about 20000 graduates, far from meeting the market demand for talent.
In this extremely unbalanced supply and demand circumstances, start-ups, giant talent scraping between the increasingly fierce.Using the employer to recruit talent, and even pay the annual salary of millions.For start-up companies, how to develop their own advantages, Attract more AI talent?
Training system is imperfect talent shortage
"Compared with O2O, a pioneering company in the shared economy, AI focuses on talent considerations differently," said Gao Tianpi, a director at Legend Capital, on 21st Century Business Herald.
Consumption Internet entrepreneurship is 'heroes do not ask the source', as long as people have insight into the product, the business model of thinking, you can start a business.However, higher AI technical barriers to entrepreneurship, talent sources will be relatively concentrated, mostly from artificial intelligence , The top academy in computer science.
Globally, there are not many institutions specializing in the systematic training of AI professionals in the world, and artificial intelligence is even a very popular specialty ten or twenty years ago. Even if someone learns the direction of AI segmentation, after graduation Most also choose to switch.
'This is because AI technology is not up to the practical target until 2013, so many AI related majors will switch to searching, recommending etc. after graduation, and there are very few people who stay in the AI related industries such as vision and voice. Also led to the present, it is difficult to find Already have the work experience of AI talent. "Shen Jian Technology CEO Yao Chung told 21st Century Business Herald reporter.
He added that colleges and universities have begun to attach importance to the cultivation of AI talents since 2013. It will take several years from training to output and the gradual improvement in the number of graduates in the future will surely bring about a gradual improvement.
In addition to inadequate training system for colleges and universities, Yao Song believes that another major reason for the shortage of AI talents is the relative weakness of China's related industries, for example, there are only a few architects who have completed a complete processor chip in China, In the United States, such as Nvidia, Intel and other companies, you can find a lot of talent in this area.China into the industry is relatively late, can dig the pool of talent is relatively shallow.
'Scarcity of talent is a common problem for AI startups, but I believe in the next two to three years, will pour in a large number of AI talent, people will get better and better.' Huiyin Hui co-founder and COO Guo Na on the current Talents shortage with optimistic attitude.
She said just as when Apple's iOS ecosystem first came out, only very few engineers could do App applications on the iOS platform, and maybe in less than a year or two, iOS engineers would be lacking, but as Apple's ecosystem Department of development, there will be a large number of people quickly learn this technology to achieve the balance of supply and demand of qualified personnel.
What kind of AI talent is the most scarce? Shun Capital Executive Director Meng Xing told 21st Century Business Herald reporter, AI's talent echelon divided into three batches. The first batch is the strongest algorithm scientists, their own framework and cutting-edge research, which People in the world have not much.
The second group of talented people may not be able to create original frameworks. However, they can adapt, improve and customize the projects in a more popular framework. Such people gradually increase their numbers as a result of continuous training.
The third installment is based entirely on the existing framework for the adjustment of the parameters of talent, such people a lot of many people who are not previously AI industry, through open classes or training can learn these.
Meng Xing believes that the current lack of AI field is an innovative algorithm scientists who are at the top to solve the fundamental problems of people, academia, industry are a lot of competition.Also, with the AI business to The direction of landing applications, the ability to understand both application requirements and technological capabilities of the boundaries of the AI product manager is also more and more attention.
Venture company talent battle
For start-up companies and big companies in AI field, it is very important to acquire the top-level AI talents and have the ability of continuous transmission of talents, while start-up companies are facing a lot of pressure from the aspect of grabbing people.
"It is not only BAT, fast Internet, big headlines today, Internet companies that have a lot of money in hand, but also at any cost to dig in. In this competition, some of the companies we vote for Silicon Valley back It will be hard to persuade others to join. '
In this case, the role of the founder of a start-up company is even more important. "It is precisely because you find it hard to dig up so many cattle from the outset, so you need the founder is a cow .For example, Shang Tang Technology, Kuang Shi Technology The founders themselves will be technically unique. "He said.
Meng Xing believes that the start-up companies to grab the main or 'look at food dishes' .For example, some people only pay attention to money, the company needs to be as far as reasonable within the scope of the transfer of wages. Large companies pay a fixed level, start-up companies will be more flexible ; Some people do not value short-term wages, pay more attention to the long-term huge returns, start-up companies in equity, options more adequate incentives; Some people pay more attention to their own voice, the degree of respect, start-up companies can Give them more scope; some scientists pay more attention to the meaning of what they do, start-up companies can use the company's features, ideas to attract talent.
In another dimension, talent comes with people, Meng said that if venture capital companies have the resources of well-known scientists, they can also bring in talents based on alumni relations, of course, the most important way to attract talent, or to keep the company It is in the business, valuation and other aspects to maintain efficient growth, so that people think this is the opportunity.
Huiying Huiying is an entrepreneurial team with well-known scientists. It is understood that Dr. Hui Xing, the chief consultant of Huiyinghuiying is the director of Stanford University Medical Physics Center and also the top medical imaging expert in the world. Chai Xiang Fei is Stanford University postdoctoral, Guo Na Ben Shuo graduated from Tsinghua University.
'The competition between AI companies in the final analysis is the top talent competition, this talent is scarce resources in the world.' Guo Na on 21st Century Business Herald reporter said the field of medical imaging AI itself has a typical interdisciplinary characteristics, whether it is technology Team or market team, all need diversified and interdisciplinary combinations.With different backgrounds of knowledge and experience of talent, the wisdom of different areas, and ultimately the formation of closed-loop productivity, breaking the limitations of a single subject.
In terms of talent acquisition, Huihui Huiying utilizes all kinds of alumni resources and launched the 'Excellence Program' in conjunction with Stanford University in August last year to send AI talents and medical professionals to Stanford for continuous interdisciplinary training. Meanwhile, Huihui Hui Ying is also with Tsinghua University and other domestic colleges and universities to establish joint laboratories, continuing to solve the problem of talent supply and training.
'There is a long way to go from laboratory to product and clinical application, and the company is the place where industrialization and industrialization are the strongest, so the linkage between the company and universities is more conducive to the cultivation of qualified personnel.' "Ms. Guo also introduced that at present AI graduates in the industry colleges and universities, annual salary levels in the hundreds of thousands to several million RMB range.China's AI talent strength and international standards, so the salary level is also relatively close to Huiying Huiying, the company in the human monthly Costs account for more than 70% of total expenditure.
Chief scientist can not be begged
The endorsement of such top AI talent as chief scientist seems to have become the 'standard' for celebrity AI companies, after all, the chief scientist's industry position largely determines the height that startups can achieve.
'Top scientists such as top AI talent can be met with demand, is also one of the few in the world.' Xinghan capital founding partner Yang Song of 21st Century Business Herald reporter said.
For example, Xinghan Capital recently invested in Kun Yun Technology, a company that makes AI chips. One of its founders, Lu Yongqing, a member of the Royal Institute of Engineering and a professor at Imperial College London, is also the leading chief scientist in the industry. Nearly 30 years of artificial intelligence This has indeed brought a great bonus to the project. Under the leadership of Dr. Lu Xinyu and his disciple, Dr. Niu Xinyu, the company's technological capabilities can maintain its top level in the industry.
However, the chief scientist is not a necessary condition for judging the project. Yang Ge said that Xinghan Capital is still looking at the project from the nature of business. The purpose is to find truly valuable enterprises and systematically evaluate the company from the perspective of technology, products and markets Consideration.
According to Meng Xing, whether a chief scientist becomes an important indicator of a start-up depends mainly on what the start-up company does: If a company does the bottom-line technological start-up, the basic human ability determines the company's technology research and development capability, technology research and development capability The advanced nature decides whether a company can survive or not, and in such a company it is important for the founding team to equip its chief scientist with AI talent.
If companies do vertical industry AI products, such as face monitoring in the field of security, and identification of certain plant features in the field of agriculture, the company will not compete with groundbreaking academic research and framework theory. Instead, Things product, engineering, such as optimizing power consumption, optimizing user experience.
For such companies, top scientists are less important, and even less-than-top scientists have even greater benefits because they are not tangled in whether this is a matter of course but rather a more practical focus on this Things will not solve the industry problem, 'he said.