Abstract
This book by Kai-Fu Lee is dedicated to two particularly high-interest topics in today’s economic world: China and artificial intelligence. While each topic has received exhaustive coverage in the literature, Lee has combined the two and has produced a very interesting account of Internet entrepreneurship with Chinese characteristics, making comparisons with Silicon Valley by examining cultural norms, historical background, and technological capabilities. Lee’s book would equally appeal to those interested in the modern Chinese economy and to those eager to learn about the future of AI technology.
The author opens the book by drawing an analogy between the Chinese public’s attitude towards AlphaGo’s victories over Korean and Chinese champions in 2016–17 and public anxiety in the United States about Soviet technological superiority when the Soviet Union launched the first man-made satellite in 1957. According to Lee, the victory of a machine powered by artificial intelligence over humans in Go, historically known as a highly complex and strategic game, became China’s ‘sputnik moment’ in the country’s drive to dominate the emerging AI technology, a drive shared by ambitious Chinese investors, entrepreneurs, and government officials. Lee makes another important point that the age of discovery is over, and that we now live in the age of implementation. And this transition from discovery to implementation is to the advantage of China, allowing China to exploit its strengths, such as overabundance of data and an extremely competitive domestic IT sector, and to minimize its weaknesses (outside-the-box approaches to research questions).
According to Lee, an AI superpower requires four main building blocks: tenacious entrepreneurs, abundant data, well-trained AI scientists, and a supportive policy environment. The copycat approach and widespread cloning in the early days of the Chinese Internet constituted the first ingredient. However simple and primitive copying and cloning may seem, the use of these methods created a very dynamic and market-oriented Internet ecosystem in China. Hundreds of entrepreneurs and start-ups were constantly fighting for their market share, trying new approaches, techniques, and partnerships. They constantly improved and localized products to better serve local Chinese customers, and in doing so they became more engaged than their Silicon Valley peers in logistics, payments, and other offline features. Their ability to cut costs and enhance efficiency was a crucial factor in their success over myriads of copycat competitors. Deep integration of online and offline activities eventually led to massive data collection about consumer behaviour – another important ingredient for the implementation of AI by Chinese tech giants.
In terms of two other metrics proposed by Lee for gauging a country’s potential to become an AI superpower – AI expertise and support from the government – China has also done relatively well. First of all, China has benefited tremendously from openness and connectivity in the field of AI academic research. The speed of improvements in AI has provided strong incentives for researchers to share their results as quickly as possible on websites such as arXiv.org rather than in traditional academic journals. And Chinese researchers have proved to be fast learners, as can be seen from the rising percentage of papers produced by Chinese scholars. As far as government support is concerned, Lee claims that China’s interventionist approach may also lead to faster deployment of new technologies in day-to-day life and generation of larger datasets, consequently creating more opportunities for further growth.
All in all, Lee expects the balance of AI capabilities between the United States and China to shift in China’s favour. China has better prospects of becoming a leading power in the implementation of the Internet and perception AI in five years’ time, while being on a par with the United States with regard to self-driving cars and other autonomous AI technologies. The author concludes that the more important divide may actually happen not between countries such as the United States and China but within each country in general, because an increase in productivity enabled by new technology can eventually lead to higher inequality and job insecurity. Unlike in the previous stages of the industrial revolution, white-collar workers will be negatively affected by advances in intelligent automation. This disruptive force which AI poses is by no means less important than opportunities provided by the new technology. According to Lee, humanitarian and social implications of AI pose important challenges: how to reform education systems?; what skills will become invaluable for the human workplace of the future?; and how to maintain privacy in a data-centric world powered by AI?; and so forth. It is likely that countries which find solutions to these questions will become the true winners in the new world order of AI.