AI unicorns are no longer the golden rice bowl, and there is a mismatch of needs between “learning AI” and “engaging AI”.
Editor’s note: This article comes from WeChat public account “Ran Caijing” (ID: rancaijing) , Author Su Qi, Editor Wei Jia. Authorized release.
“Artificial intelligence has not made humans unemployed, people who engage in artificial intelligence are unemployed first.” This was a joke, but it became a reality in 2019.
This year, AI companies such as HKUST News, Bitmain, and Guangxi Technology have reported layoffs. The once-popular AI unicorn is no longer a panacea. Also in this year, Ali AI labs introduced two scientists with an annual salary of one million US dollars, and Huawei also paid a high-priced annual salary of 2 million for the new PhD.
For a large number of new computer graduates who are about to find work after the year, such “cold and hot” news also makes them feel confused and worried. When they were looking for a job around their resumes, they discovered that there was a mismatch in demand between “AI-learning” and “AI-engagement.”
“Side-laying, while grabbing people.” Behind the AI industry’s big move and lay-off, there are several key issues reflected. The AI talent cultivated by the school does not match the market, and students can’t find role models; It is very difficult for AI startups to quickly recruit AI talents that meet the needs of the job; and truly top AI talents, large companies are unable to retain them, and many of them come out of their way after mastering resources and experience.
According to the Talent Diversity Insights report released by LinkedIn, the top five emerging occupations in China are: new media operations, front-end development engineers, algorithm engineers, UI designers, and data analysts, more than half of which are related to intelligence Related. The demand for data-related skills in China has increased seven-fold in the past five years, but there are still 15% of job vacancies in the market.
What kind of talents will artificial intelligence need in the next ten years? How should the related education and training be carried out? In response to these problems, Ran Finance and AIPHAROS Moonlight Club co-sponsored a salon, Du Zidong, associate researcher at the Institute of Computing, Chinese Academy of Sciences, Yin Xucheng, deputy dean of the School of Computer and Communication Engineering, University of Science and Technology Beijing, Zhu Jupeng, co-founder of 51 headhunter, and Ma Yingying, angel investment director Rui, Gong Siying, Chief Strategy Officer of Mor.AI, Cui Yunfei, Director of Bello, and other researchers, practitioners, and investors in the field of AI conducted in-depth discussions.
Shooting people while grabbing people
The once-popular AI industry has cooled down in the last two years.
“2019 China Artificial Intelligence Industry Investment and Financing White Paper” shows that 2014In the application, this has gone hand in hand with the United States, but the gap between the AI hardware and the platform is very wide, and the AI theory is basically blank.
University can’t keep up with AI talent training. The school’s education and company’s specific talent needs are mismatched, which also leads to problems in the employment of enterprises. It can’t retain top talents, can’t quickly train new talents, and can’t find real needs. The problem of talents abounds. 51 headhunting co-founder Zhu Jupeng said that undergraduates generally receive general education, those things are Know How, and corporate employees look at To Do. Students can’t get started immediately when they arrive at the company, which makes the enterprise unavailable.
Aware of these issues, universities are already actively changing. Colleges and universities in the United States computer direction TOP4 have jointly released the “White Paper on Machine Learning Systems”, and teachers teaching computer system courses have become the next most sought-after potato.
For the domestic, no reference course, no ready-made teachers, and no mature teaching materials have become the three big mountains before universities. In 2018, the School of Computer Science, University of Chinese Academy of Sciences began to provide an artificial intelligence-oriented system course called “Intelligent Computing System”. This course is being or will be offered in Peking University, China University of Science and Technology, Tianda University, Beihang University, Nankai, Beijing Institute of Technology, and Huake. Du Zidong believes that China needs a large number of developers and designers of intelligent computing systems.
Yin Xucheng said that the basic technical system of the “artificial intelligence discipline” includes mathematical foundations (calculus, linear algebra, etc.), computer and system foundations (computer composition principles, algorithm foundations, etc.), machine learning and pattern recognition methods (machines Learning methods, deep learning methods, etc.), artificial intelligence technologies (speech recognition, text recognition, computer vision, natural language processing, etc.), artificial intelligence applications (autonomous driving, intelligent security applications, intelligent financial applications, etc.).
“At present, there are more than a dozen disciplines doing artificial intelligence in China, and computer, electronics, communication, and mathematics all offer related courses.” He believes that to cultivate system-level talents in artificial intelligence technology, we need Cultivate in the above five aspects.
What kind of AI talents will be needed in the future?
According to data from Tsinghua University, computer vision, speech, and natural language processing are the three largest application directions in the Chinese market, accounting for 34.9%, 24.8%, and 21%, respectively. However, several interviewees told Ran Cai that in addition to these application directions, talents in management, sales, and integration of various disciplines are also needed.
Yin Xucheng believes that if you look at the Gartner curve (technical maturity curve), you can see that many intelligent systems can now automatically generate algorithms, and systematic and comprehensive high-end technical personnel are increasingly scarce. “The road gets narrower as you go up. Now the more competitive positions, everyone’s skills are actually good. In this case, the overall quality and other aspects (such asCompetition) is very important. “Yin Xucheng said.
Many technical talents are proud of technology. Once involved in the mystery of capital, it is easy to lose themselves in the struggle for interests. The failure of China’s driverless unicorn Roadstar.ai is a good example. Roadstar.ai has raised US $ 128 million in funding. Eventually, due to internal disputes among the founding team, this star project suddenly stopped, and a series of internal behaviors also triggered widespread discussion of technical talents in the society.
Some investors believe that when autonomous driving is in the wind, financing is too easy and too easy for people to expand. Pure technical talents lack real management experience and social experience. They do not know how to deal with market and capital demands and within Temptation.
Mor.AI Chief Strategy Officer Gong Siying believes that in the future, AI management talents must be in strong demand. Many investors started investing in technology or investment experts when they first invested in some AI startups. They thought that if you did technology, you would be successful, but later found that this logic may not be true. AI genius is not equal to AI talent.
The management talents of AI need not only a set of management logic theory, but also a sensitivity to science and technology and a keen judgment on the industry. The industry of AI has become too fast. In the next two decades, if such AI management talents can emerge, they will be the scarce resource of the entire industry.
For future AI talents, Inno Angel Investment Director Ma Rui values the founder’s comprehensive quality. In addition to having a certain technical literacy, communication ability, product ability, ability to gather core members and ability to attract talent, Are very important. “My current bottom-level investment logic is very simple. You don’t want the founder or team to come to your house as a guest, why should you vote for him?”
Looking back in recent years, AlphaGo’s victory over Li Shishi in 2016 directly ignited the enthusiasm of venture capital in the AI field. The promulgation of the “New Generation Artificial Intelligence Development Plan” in July 2017 marked the rise of artificial intelligence to a national strategic level. The introduction of the science and technology board in 2019 has given sufficient space for financing and exit of AI companies.
Capital swarms, but the benefits you see are minimal. The “2018 China Artificial Intelligence Business Landing Research Report” shows that in 2017, over the entire industry chain, more than 90% of AI companies are still at a loss stage, and most of them have annual operating income of less than 200 million.
In a hot and cold contrast, the AI industry is showing signs of fatigue. Relevant data shows that by 2020, the market size of China’s artificial intelligence core industry will exceed 160 billion yuan, but commercialization, implementation, and scale will still be difficult problems facing AI companies.
“For the angel stage, the field of artificial intelligence that can be invested is basically gone now.” Ma Rui said that now there are very few opportunities to start a business in the AI field. It may not be an algorithm model, but Is the ability to land.
But this market is not completely without opportunities. In the next few years, the integration of artificial intelligence and other fields, especially basic sciences, can increase the calculation volume or speed by a factor of ten or one hundred, and also form a vent.
He said that there are several basic elements that can form a vent: one is that the market space must be large enough, and the other is that the timing is right. AI first landed in finance or security. The big reason is that the data of security and finance was deposited before the formation of AI. Over time, everything is interconnected, and many industries without data have begun to accumulate. These areas that did not use AI before will also use AI soon.
Mary’s investment theme set for himself this year is to reduce costs and increase efficiency. Companies that can use AI technology to help enterprises reduce costs and increase efficiency will have opportunities.
“In addition, China will definitely increase the salary of basic subject talents in the future.” He said that artificial intelligence may become as popular as computer science in the next five or ten years. Disciplines are merged. “It is possible that you not only need an artificial intelligence PHD, but also a physics PHD, or a biological PHD, in order to become a more powerful expert in this field.”
From the perspective of investment, if you really want to achieve the growth of the jumping class, the most important thing is the insight of society and humanity. Applying artificial intelligence to the right place will always be an excellent entrepreneur or entrepreneur Essential competencies.
According to data from Tsinghua University, computer vision, speech, and natural language processing are the three largest application directions in the Chinese market, accounting for 34.9%, 24.8%, and 21%, respectively. However, several interviewees told Ran Cai that in addition to these application directions, talents in management, sales, and integration of various disciplines are also needed.
Yin Xucheng believes that if you look at the Gartner curve (technical maturity curve), you can see that many intelligent systems can now automatically generate algorithms, and systematic and comprehensive high-end technical personnel are increasingly scarce. “The road gets narrower as you go up. Now the more competitive positions, everyone’s skills are actually good. In this case, the overall quality and other aspects (such asCompetition) is very important. “Yin Xucheng said.
Many technical talents are proud of technology. Once involved in the mystery of capital, it is easy to lose themselves in the struggle for interests. The failure of China’s driverless unicorn Roadstar.ai is a good example. Roadstar.ai has raised US $ 128 million in funding. Eventually, due to internal disputes among the founding team, this star project suddenly stopped, and a series of internal behaviors also triggered widespread discussion of technical talents in the society.
Some investors believe that when autonomous driving is in the wind, financing is too easy and too easy for people to expand. Pure technical talents lack real management experience and social experience. They do not know how to deal with market and capital demands and within Temptation.
Mor.AI Chief Strategy Officer Gong Siying believes that in the future, AI management talents must be in strong demand. Many investors started investing in technology or investment experts when they first invested in some AI startups. They thought that if you did technology, you would be successful, but later found that this logic may not be true. AI genius is not equal to AI talent.
The management talents of AI need not only a set of management logic theory, but also a sensitivity to science and technology and a keen judgment on the industry. The industry of AI has become too fast. In the next two decades, if such AI management talents can emerge, they will be the scarce resource of the entire industry.
For future AI talents, Inno Angel Investment Director Ma Rui values the founder’s comprehensive quality. In addition to having a certain technical literacy, communication ability, product ability, ability to gather core members and ability to attract talent, Are very important. “My current bottom-level investment logic is very simple. You don’t want the founder or team to come to your house as a guest, why should you vote for him?”
Looking back in recent years, AlphaGo’s victory over Li Shishi in 2016 directly ignited the enthusiasm of venture capital in the AI field. The promulgation of the “New Generation Artificial Intelligence Development Plan” in July 2017 marked the rise of artificial intelligence to a national strategic level. The introduction of the science and technology board in 2019 has given sufficient space for financing and exit of AI companies.
Capital swarms, but the benefits you see are minimal. The “2018 China Artificial Intelligence Business Landing Research Report” shows that in 2017, over the entire industry chain, more than 90% of AI companies are still at a loss stage, and most of them have annual operating income of less than 200 million.
In a hot and cold contrast, the AI industry is showing signs of fatigue. Relevant data shows that by 2020, the market size of China’s artificial intelligence core industry will exceed 160 billion yuan, but commercialization, implementation, and scale will still be difficult problems facing AI companies.
“For the angel stage, the field of artificial intelligence that can be invested is basically gone now.” Ma Rui said that now there are very few opportunities to start a business in the AI field. It may not be an algorithm model, but Is the ability to land.
But this market is not completely without opportunities. In the next few years, the integration of artificial intelligence and other fields, especially basic sciences, can increase the calculation volume or speed by a factor of ten or one hundred, and also form a vent.
He said that there are several basic elements that can form a vent: one is that the market space must be large enough, and the other is that the timing is right. AI first landed in finance or security. The big reason is that the data of security and finance was deposited before the formation of AI. Over time, everything is interconnected, and many industries without data have begun to accumulate. These areas that did not use AI before will also use AI soon.
Mary’s investment theme set for himself this year is to reduce costs and increase efficiency. Companies that can use AI technology to help enterprises reduce costs and increase efficiency will have opportunities.
“In addition, China will definitely increase the salary of basic subject talents in the future.” He said that artificial intelligence may become as popular as computer science in the next five or ten years. Disciplines are merged. “It is possible that you not only need an artificial intelligence PHD, but also a physics PHD, or a biological PHD, in order to become a more powerful expert in this field.”
From the perspective of investment, if you really want to achieve the growth of the jumping class, the most important thing is the insight of society and humanity. Applying artificial intelligence to the right place will always be an excellent entrepreneur or entrepreneur Essential competencies.