China has long been considered the global power best positioned to develop artificial intelligence. This observation comes on the basis of the funds Beijing has made available for this purpose, the push given by the country’s leadership in this field (as early as the 1990s, Jiang Zemin considered the race for AI fundamental for the country’s future, and Xi Jinping himself has mentioned the topic a number of times in his speeches), and the immense amount of data that Chinese companies can use to improve their algorithms.
The launch of ChatGPT by Open Ai, for example, is bound to become yet another “Sputnik moment” for China’s technology sector (as happened after the victory of Google’s Alpha Go over the Chinese Go champion, the first software to defeat a human at this game, which prompted the CCP to step on the gas with indigenous innovation).
Recently, Baidu, China’s leading search engine, announced the launch of its own version of ChatGPT in March, and it appears that Tencent and TikTok are also working on something similar. The Chinese tech scene, even before it became globally recognized, was described by Lee Kai-Fu in his book AI Superpowers: China, Silicon Valley, and the New World Order (2018). Lee, who is Taiwanese with experience at Google, Apple, and Microsoft and an expert on technology in China, has been working with a venture capital firm for some time to fund his technology projects in China and elsewhere.
When the book came out in 2018, few would have bet on the Chinese tech surge. His book made it possible to take note of Beijing’s advances, which had gradually become a rival to the U.S. in a field usually considered “its own” by the West. It must also be pointed out that his book led to an almost complete about-turn: from being the factory of the world, in Western narratives China started being seen as a dangerous adversary capable of using technology with extremely repressive functions, at home and not only. Another skewed perception led to overvaluing China in areas where it still has many weaknesses (semiconductors, for example) and unclear governance issues.
Now, with China once again among the “bad guys,” there is a renewed tendency to diminish the global import of Chinese technological innovation, perhaps also as a coping mechanism: some commenters on the semiconductor clash seem to be breathing a sigh of relief that the West will not be beaten by China after all.
However, Lee Kai-Fu had the merit of opening up a world in which science fiction writers also found their place. AI Superpowers came out in 2018, but three years earlier, in 2015, Liu Cixin had become the first Chinese to win the Hugo Award for science fiction with his The Three-Body Problem. In three years, China had become an interesting phenomenon in many ways, and many writers after Liu found international acclaim.
One of them was Chen Qiufan, nicknamed the “Chinese Gibson”: Chen is an author who uses so-called science fiction realism, trying to recount a future that is already here. He and Lee formed the perfect duo, teaming up to write AI 2041, Ten Visions of Our Future (Luiss University press, 2023), a book in which Chen investigates through fiction the impact of AI in our daily lives in a tomorrow not too distant from the present day, and Kai-Fu analyzes its technical features, providing a kind of “hardware” to Chen’s fictional framework.
The result is a book that gives us a glimpse of what our world might look like before long, through the exploration of different applications of AI: from deep learning to deep faking, from robotics to ChatGPT-like bots. It aims to develop a concept expressed in Lee Kai-Fu’s introduction: that AI is the fundamental technology of our present, but also represents, above all, our inevitable destiny.
The two authors have no intentions of scaring the reader – they are both fundamentally optimists. But while Kai-Fu recounts the workings of AI as applied to different situations, Chen transfigures the cold technology talk into stories that impress on us the risks of technological development without any control, thus stressing a point that has long been made in the world of AI experts: that rather than trying to predict what we will be able to do with AI, it’s high time regulate its applications already, knowing full well that its path of advancement remains a mystery to everyone at this point.