Artificial intelligence (Al) is regarded as a key to drive the world's future development. Although Europe and the United States are dominated by private enterprises, while the Mainland China is led by national entities, the core of AI formation, in fact, is rested on available top-notch talent. The 2017 Global AI Talent White Paper released by the Tencent Research Institute in December last year stated that there are approximately 300,000 AI researchers and practitioners in the world, while the market demand for AI talent is in millions. In the first 10 months of 2017, the demand for AI talent was twice of that in 2016. The report suggests that the bottleneck is education - though there are 20,000 graduates from related disciplines each year, the number is far from adequate to meet the demand.
University is a battlefield for acquiring talent. For example, Mark Zuckerberg, CEO of Facebook, personally recruited Yann LeCun, a French professor from New York University. LeCun studied computer deep learning starting from 1980s. Google recruited Li Feifeii, a Chinese-born expert of computer vision at the Stanford University in 2016; Yoshua Bengio, an expert in artificial neural network in Canada, promised to assist Microsoft last year. Enterprises have also set up laboratories around the world to attract talent. For example, Tencent set up an AI lab in Seattle, U.S. and Baidu has a lab in Silicon Valley as well.
A mountain of demand has caused salaries and bonuses of AI talent to rise. DeepMind, a software development company that beat the world GO champion, has an average annual salary of US$340,000 (approx. HK$2.67 million) each for its 400 employees; Baidu's "Young Leader Program” screens young people below the age of 30. The outstanding scientists even on training are offered an annual salary of at least RMB 1 million (approx. HK$ 1.22 million). For inexperienced Al-related positions in the market, the average monthly salary in the Mainland has reached RMB 25,800 (approximately HK$ 31,700) in the past three years, which is much higher than general technical positions and that does not even count stock option and other benefits.
Al jobs are well-paid and have a promising future. Is Hong Kong also having a position in the race?
In fact, the education provided by local universities is excellent. Graduates have long been the target of recruitment. For example, Tencent established an AI joint lab with the Hong Kong University of Science and Technology (HKUST) in 2015, it was led by Professor Yang Qiang, an internationally renowned Al expert; Baidu successively recruited interns from the Chinese University of Hong Kong (CUHK), the University of Hong Kong (HKU) and HKUST at the beginning of this year; HKU and Alibaba jointly established an research institute to develop AI in Hangzhou four years ago.
However, how many Al talents are there in Hong Kong? It is roughly estimated that there are less than 1,000 in the local tertiary institutions - according to Professor Francis Lau, Deputy Dean of the School of Engineering at HKU, there are only 100 AI talent in the whole of the university. Although CUHK has already applied deep learning study as early as 2001, it only opened the Hong Kong's first deep learning elective course and the bachelor's degree in finance and technology in the recent two years. No wonder the financial institutions such as JP Morgan Chase and HSBC have complained about the shortage of technology talent in Hong Kong. Last year, HKUST's Professor Yang further expressed that he failed to recruit 50 AI staff for Huawei.
HKU, HKUST, and CUHK are among the world's top 50 in several international university rankings. But why are there so few AI talent and such a weak innotech atmosphere? I believe the source of problem are lack of an ecosystem of government, business, academia and research; lack of open data and market size.
An ecosystem of government, business, academia and research enables the commercialization of scientific research and technology, improvement and marketing of the product. We have only a few successful cases, let alone to be certain that researchers are likely to share the financial returns of their efforts. Basically, the government as one of the largest users can provide the much needed initial momentum to establish an ecosystem. However, the implementation of new measures by the government is complicated and time-consuming. For example, it takes three years for the trial of the first 50 smart lamposts from conception to commissioning. I wonder if the technology tested would be outdated by then.
At the same time, the success of AI depends on the huge amount of data. However, the government's one-stop open data platform or common spatial data infrastructure (CSDl) will only be ready until 2023. This kind of information sharing platform has long been available in Europe, the U.S., and Singapore. While it is not new, we still have to wait for several years to enjoy it, which definitely weakens the potential and retards Hong Kong's development of AI and innotech, I feel very much helpless.
In addition, the recent German venture capital fund Asgard proposed to European policy makers to promote AI, saying that Europe must be united to build a market comparable to that of China and the U.S. The proposal is similar to Hong Kong's strategy to integrate with the Greater Bay Area, which is yet to be accepted by the Hong Kong people.
Apart from rule of law, Hong Kong's long-standing leading edges, as mentioned by Alibaba's Jack Ma at the HKU recently, are: accommodating, innovation and young people. Mrs Carrie Lam, the Chief Executive, also pointed out earlier that in today's economic development, it seemed that those who had recruited and nurtured the most numerous talent would be successful. Therefore, I sincerely hope that all parties can embrace innovative thinking and provide more opportunities for our young people. Hong Kong can then develop into a leading international innotech hub.
Dr. Winnie Tang, Honorary Professor, Department of Computer Science, University of Hong Kong