Kai-Fu Lee: In the AI ​​era, China and the United States will become the undisputed duo

In September, the new book "AI·Future" by Kai-Fu Lee, founder and CEO of Sinovation Ventures, was released globally. On September 2, Kai-Fu Lee held a media meeting at the Quintiles Hotel in Beijing, and Leifeng.com also attended the event.

In response to the media's question about the current status of AI venture capital in China, Kai-Fu Lee said that the concept of AI is very popular and many companies want to use AI to package themselves, which has brought a lot of bubbles to the AI ​​industry. In the past period of time, the valuations of AI companies were generally too high. "In the past three or four months, the valuations of AI companies have dropped by 20%-30%. If they drop by another 20%-30%, the valuations of AI companies will be reasonable, and Sinovation Ventures will increase its investment."

早在去年4月，李开复与创新工场CTO王咏刚合著的《人工智能》一书中，李开复就警惕了AI行业过火，将会产生泡沫。此前，在一次对李开复的直播采访中，李开复表示：“在我看来每个创业者都很想把自己的创业公司包装成一个 AI 公司，而每个 VC 也都想要标榜自己是一个 AI 投资人。但是 AI 投资其实不适合新手参与。如果对 AI 技术本身没有了解就急着参与到 AI 的早期阶段来，这样的人会赔得连裤子都不剩。到了今年年底，我觉得会有很多泡沫破掉。”

It can be seen that Kai-Fu Lee is cautious about the current development of artificial intelligence, but this does not affect his long-term optimism about artificial intelligence.

"AI·Future" is the second book written by Kai-Fu Lee on the theme of AI. Kai-Fu Lee said that the starting point of "AI·Future" is different from that of "Artificial Intelligence". "Artificial Intelligence" is positioned as a popular science book on artificial intelligence, and it wants everyone to understand what artificial intelligence is and how it is related to their lives. The book contains many descriptions of core technologies such as deep learning, as well as a large number of cases. "AI·Future" is actually not a book about AI technology, but a book on humanities, talking about the future led by AI technology. Based on his many years of industry experience, he made predictions for the next fifteen years from the perspectives of scientific research, industry, and investment.

The media present at the event got an advance copy of "AI·Future". After reading the entire book, Leifeng.com editors found that "AI·Future" focuses on two core issues: First, will China surpass the United States and take the lead in artificial intelligence? Second, what threats and challenges will artificial intelligence bring to mankind?

In Kai-Fu Lee's view, in the era of artificial intelligence, China and the United States will become the undisputed two giants. "The United States is the country with the deepest accumulation of artificial intelligence research and the most application results in the world... However, as the 'deep learning' technology promoted by pioneers such as Hinton is widely used in the Internet and business, artificial intelligence has entered the era of practical work from the era of discovery of the lone expedition.

In the transitional stage between the two eras, the development speed of artificial intelligence that exceeds Moore's Law has prompted researchers to share their results immediately. Researchers and engineers in China and the United States can benefit from instantly connected global research resources, which gives China, an artificial intelligence student, an equal opportunity to catch up with its American teacher. More importantly, the balance of competitiveness in the era of practical artificial intelligence will tilt towards commercial execution, product quality, innovation speed and big data, which are precisely the factors where China is superior to the United States. "

Kai-Fu Lee said that he believed the book "AI: The Future" would sell well in the United States because Americans do not understand China's Internet and artificial intelligence and they would be very interested in learning more.

In the first three parts of the book, Kai-Fu Lee describes the current status of China's Internet and artificial intelligence. Perhaps the biggest difference between China and the United States in terms of innovation and business environment is that the American technology circle despises imitation and plagiarism, but China does not have such a good soil for innovation as the United States. Under the fierce entrepreneurial competition, relying on imitating foreign products and technologies, China has produced a large number of world-class entrepreneurs, such as Alibaba, Baidu, Tencent, Sohu, Didi, etc. However, after imitating, Chinese entrepreneurs usually carry out localized innovation according to China's national conditions, which is the part that foreign giants cannot compete with.

Kai-Fu Lee said, "15 years ago, Chinese Internet startups that started out by 'learning' were inspired by American business models and competed fiercely with each other... When this generation of Chinese entrepreneurs learns to use artificial intelligence, it will completely subvert the rules of the game."

After the traditional Internet era, which was mainly developed by imitation, China has also produced many innovative models: mobile payment, O2O, shared bicycles, etc. The data accumulated by China based on its population base and unique market is also a major advantage. "China's mobile user base makes China's data advantage three times that of the United States, mobile food delivery is 10 times that of the United States, mobile payment is 50 times that of the United States, and shared bicycle facilities are 300 times that of the United States. By utilizing these rich data resources, China's computer vision, drones, speech recognition, speech synthesis and machine translation companies have become the world's most valuable startups."

Faced with the continuous improvement of AI intelligence and operational capabilities, people are most worried that their jobs will be replaced by AI. This is also the most commonplace issue since the development of AI. However, in 2013, Kai-Fu Lee was diagnosed with stage 4 lymphoma. The serious illness gave him new thoughts on life and living.

Kai-Fu Lee talked about how when he was in his 20s, he believed that AI would replace the human brain in the future; when he was in his 30s and immersed in research, he found that there were good papers on AI, but it was difficult to make actual products; when he was 40, he saw the limitations of AI and that it was only valuable in certain fields; and when he fell ill, he believed that AI would help people solve repetitive work problems and allow people to enjoy life better.

Kai-Fu Lee is a typical workaholic who hopes to achieve something in the field of AI and regards this as his greatest goal and value in life. But after he got sick, he found that the most important thing in life is to spend more time with the people he loves and do more meaningful and creative things.

“Artificial intelligence will replace humans and complete various repetitive tasks that are not exclusive to humans. Love is the characteristic of human beings.”

Kai-Fu Lee believes that within 15 years, artificial intelligence and automation will have the technical capabilities to replace 40-50% of jobs.

Examples include repetitive labor, especially work done in the same or very similar places (such as dishwashing, assembly line inspection, sewing); various interactions with fixed scripts and dialogue (such as customer service, telemarketing); relatively simple data classification (such as filing documents); working in a very narrow area of ​​a company (such as a telemarketer of bank financial products, an accountant for a department); and work that does not require a lot of face-to-face interaction with people (such as sorting, assembly, data entry).

Compared with artificial intelligence, humans have the advantage of creativity and compassion, and they can engage in more humane occupations and positions. In his opinion, the following are the types of jobs that will be difficult for artificial intelligence to replace (at least at the current stage):

Artificial intelligence is not good at coming up with new concepts, and creative jobs are difficult to replace (for example, medical researchers, artificial intelligence scientists, award-winning scriptwriters, public relations experts, entrepreneurs); for artificial intelligence, even understanding common sense is difficult, and complex/strategic jobs that require understanding of multiple fields and making strategic decisions are difficult to replace (for example, CEOs, negotiation experts, M&A experts); progress in robotics and mechanics is actually slower than artificial intelligence software, and sensitive jobs are difficult to replace (for example, oral surgeons, aircraft mechanics, chiropractors); robots work well in specific environments (such as assembly lines), but are not easy to adapt to new environments, and jobs that need to adapt to new and unknown environments are difficult to replace (for example, geological surveys, cleaning after gatherings); Artificial intelligence does not have human emotional intelligence, and people are unwilling to "trust" machines to let machines handle human tasks, so empathetic/humanized jobs are difficult to replace (for example, social workers, special teachers, marriage counselors).

Kai-Fu Lee shared that considering hard work as the only value of life is the value of the industrial revolution era. In the era of artificial intelligence, we should rethink the meaning of existence. After recovering from his physical condition, his working hours have been reduced a lot compared to before. In the one-month summer vacation when his daughter was on vacation, he spent less than half of his time in the company, and the rest of the time was spent with his daughter.

Leifeng.com (Official Account: Leifeng.com) learned that in addition to the employment issue, Kai-Fu Lee is also worried about the uneven distribution of wealth in the era of artificial intelligence. Because if artificial intelligence companies can replace a large number of human jobs with machines and algorithms, the wealth they create will be amazing. In addition, the artificial intelligence industry tends to be a "winner takes all" industry, which will lead to problems of profit and wealth concentration. Globally, China and the United States are already far ahead of other countries in the development of artificial intelligence, forming a new generation of bipolar world technology order.

The AI ​​revolution has brought about a new pattern and a new society. Our generation grew up with the Internet, while the next generation will grow up in a new era of artificial intelligence. The education of the next generation is also an issue that needs to be considered urgently. Kai-Fu Lee believes that the next generation will no longer be inclined to choose a stable job when choosing a job. Cultivating critical thinking and creativity is more important than career choice. Only with these, will they not face the fear of being replaced by AI.

Kai-Fu Lee predicts AI bubble will burst by the end of the year, LeCun: It is true

Full text of LeCun, Jordan, and Kai-Fu Lee’s roundtable: The current state and future of AI | GMIC 2018

Kai-Fu Lee, Hu Yu, and Zhang Yaqin: Differences between China and the United States in AI strategies and AI talent training | GMIC 2018