The New Chinese Revolution — Part II: AI 2030 and Belt and Road Initiative
Everybody is talking about Artificial Intelligence. From giant corporations to tiny startups to governments, everyone seems interested in investing in it. It’s understandable. PwC projects that, in 2030, this technology will contribute to a 14% increase in the global GDP, which corresponds to something like U$15.7 trillion! Currently, companies are deploying AI on an array of sectors, focusing on things like procedures improvement, customer service, industrial automation, productivity increase, and costs reduction. As technology evolves, it’s going to become more and more omniscient and omnipresent, expanding its applicability to other areas. That makes mastering AI important to companies but especially crucial for nations that want to have a leadership role in the next decades.
Not all roses, though. Throughout the world, debates are arising about how the expansion of AI’s usage is going to result in massive unemployment and that the fast pace at which it’s growing won’t give enough time to a whole generation of workers to prepare themselves to all the changes that are going to happen in jobs requirement. Figures and opinions vary: optimists believe that only repetitive and bureaucratic tasks are going to be automated, releasing human beings to work with what they love; pessimists talk about an inequality soar and demand solutions, like the universal basic income, for example. One thing, however, is a fact: artificial intelligence is here and it’s not going away.
But what actually is AI?
A simple explanation is that artificial intelligence is a simulation by machines of the human intelligence, that is, systems capable of reasoning, discovering, learning and getting conclusions on their own, different from the traditional software, which only performs functions that were programmed for.
With all the economic potential involved, AI became a priority for the most important countries of the planet. And China, of course, couldn’t be left out. On the first article of “The New Chinese Revolution”, I talked about Made in China 2025, a huge plan by the Chinese government to invest heavily in 10 breakthrough technologies that are changing the global order. Now, the focus is on another two even more audacious plans: AI 2030 and the Belt and Road Initiative.
Launched in 2017, AI 2030 (the official name is “New Generation AI Development Plan”) intends to mastery and make China a world reference in artificial intelligence. The plan has three stages:
1) By 2020, the Chinese want to match the world’s major power in the technology, the US. They estimate that AI will directly generate U$22.5 billion and have an impact of U$150 billion on other sectors, all of this in the domestic market;
2) By 2025, the country wants to be a leader in some AI applications and estimates that the technology will directly generate U$60 billion, besides having an impact of U$750 billion on other sectors in the domestic market;
3) Finally, in 2030, China wants to be the world’s biggest player in artificial intelligence. They estimate direct revenues of U$150 billion and an amazing impact of U$1.5 trillion on other sectors of their domestic market.
The numbers are really impressive and, although were set by the government (which could mean some kind of overestimation), there are no reasons to doubt it — just remember about the PwC’s research commented above.
AI had been cited as a breakthrough technology in other projects of the Chinese government before, but it only gained prominence (and an exclusive plan) after an event that, for a lot of “normal” people, wasn’t that important: in 2016, AlphaGo, a super algorithm, has defeated Lee Sedol, one of the best Go players in history. Go is a board game created in China and very popular across Asia. It is considered one of the most complicated in the world and, until that, seen as an unlike challenge to be surpassed by a machine. The Chinese became worried about the fact that AlphaGo was developed by a British company (Deep Mind) controlled by an American one (Alphabet, Google’s parent). The concern was so significant, that even symposiums were organized to discuss the subject. Watching a Western organization mastering such a millenarian game made China project that mastery in military affairs in the future, so it decided to start AI 2030.
Warfare is, without a doubt, one of the main reasons for all the interest in artificial intelligence. After all, it’s a matter of time (unfortunately) before the development of smart and autonomous weapons and, of course, whoever is ahead in that race will earn the status of superpower. Although military usage is not written in the program and China talks about ethical standards in AI deployment, it’s hard to believe that it’s not going to happen, especially when we know that the director of the Chinese Association for Artificial Intelligence, a government agency for the matter, is also a high patent general of the Chinese Army (to be fair, DARPA, the research and innovation agency of the US Department of Defense, is one of the biggest investors in the technology too). AI 2030 shows that the Asians are really aware of everything that is happening in this field around the world. The plan even cites American, European, British and Japanese strategies for artificial intelligence as well.
The Chinese program is divided into four main areas:
- Hardware, with the development of the chip industry and supercomputers;
- Database, responsible for feeding the algorithms;
- Attracting AI professionals and the formation of new ones, besides investments in R&D
- Creation of a strong AI commercial ecosystem.
Regarding the chip industry, development comes from massive domestic investment (as I wrote in the first article, president Xi Jinping plans to launch an exclusive U$47 billion fund in order to invest in chips for AI application, as well as 5G technology) and from buying foreign companies, which is a major concern for some countries — notably US and the ones in EU — that see the Chinese as a threat that steals technology at the same time that closes their internal market to avoid that the same happens to them. Although China is still way behind in this industry, some startups are getting promising results. When it comes to supercomputers, Chinese are already prominent: 202 of the 500 fastest supercomputers in the world are in the Asian country (against 143 in America).
Having the biggest population on Earth makes China database gigantic and diverse, which is crucial for developing and improving algorithms. The number of inhabitants also facilitates recruiting people to test new applications even because the Chinese are technology enthusiasts. Just to get an idea, Tencent’s WeChat, similar to Whatsapp, but with way more functionalities, has almost 900 million active users. Toutiao, a news app from Bytedance, is used for almost 700 million people monthly. Foreign companies can’t have access to all of this data. Government protectionism is heavy and even tech giants (like Google or Facebook) have little or none relevance in the country at all, which guarantees a domestic market domain for local enterprises. One example of that protectionism is a law passed in 2017 that prohibits the storage of Chinese customer data abroad.
But what’s the point of pouring billions into a plan when having a human capital scarcity? To surpass it, China has two fronts: it wants to attract high level AI professionals, Chinese or not, offering salaries compared to the ones in the US market (sometimes even bigger) at the same time that develops AI study centers across the country. There are a significant number of Chinese studying the technology at the American best universities and it’s normal for some companies to send emissaries in order to recruit those students with generous offers. Renowned professionals like Andrew Ng (Google’s Brain ex-chief) and Qi Lu (Microsoft’s ex-VP) did the same path, joining Baidu, one the China’s giant tech firms. They both, though, are back in the US after a few years. Sensetime, which I’ll talk better later, for example, has 150 PhDs in their staff and have published more articles on the subject than Google and Facebook since 2015.
Domestic initiatives aiming to improve the formation and research on AI are also taking place in China. Recently was announced the construction of a technological park at Beijing, with U$2.1 billion in investments. Tencent, Baidu, and Alibaba also have built their own research centers; the latter even committed a U$15 billion investment in high technology R&D, including artificial intelligence.
Government’s effort to improve local education institutions seems to be working: a ranking developed by Time High Foundation shows that seven of the 10 best universities in emergent countries are from China. The gap for the US, however, is still huge and hardly will be closed one day, but it’s likely to have at least one Chinese institution among the best of the world in the next years.
The last front of the strategy is the development of a strong AI ecosystem internally. It comes from massive government and private investments in startups and, especially, from the so-called “four big winners”: Alibaba, Tencent, Baidu, and iFlytek. The two first have impressive numbers: Alibaba Group, which comprehends from e-commerce to cloud computing to logistics, has a market cap of more than U$500 billion and annual revenues of U$250 billion. Tencent also has a market cap of around half a trillion dollars and annual revenues of U$240 billion. Baidu and iFlytek are more “modest”: they worth, respectively, U$93 billion and U$77 billion. The latter, by the way, drew a lot of attention when it developed a system capable of being approved with a Top 5 score in the national medicine exam.
China’s artificial intelligence ecosystem is already the second biggest in the world: 23% of all AI companies are there; only the US has more, with 42%. However, when it comes to AI startup funding, the scenario changes: in 2017, of the U$15.2 billion invested in those startups, 48% were in Chinese ones, against 38% in Americans. It was the first time it has happened. So it’s not a surprise to say that Sensetime became the most valuable AI startup in the world, with a valuation of U$4.5 billion. Its main application is a face recognition algorithm that’s been used, among others, by the local government for surveillance.
Running after the US
Despite the clear advance in the AI development, China is still way behind the US, according to an index that measures both countries capabilities, created by Jeffrey Ding, a researcher from Oxford University (by the way, this article was based on an excellent report wrote by Ding about AI 2030, which I strongly recommend). The index states that China is ahead only in database and startup investments; the other indicators are led by the US, as the box below shows.
All the evidence, however, indicate that the gap between the nations has been closing considerably. So much that, in May, the White House gathered more than 100 leaders of technology companies and VC firms to discuss the present and the future of AI in the country, besides China’s threat. After the meeting, a committee was created, involving government and academia members, and organizations representatives. Although financial and academic collaboration between companies of both countries exist — Alibaba Group is listed on the New York Stock Exchange, for example -, it has become more and more clear that there are two separated projects and that being at the top of such a prominence technology not only matters but will be crucial in the plan to expand influence through other parts of the planet.
China’s urgency to develop and master AI has to do with the creation of standards that are going to rule the sector’s future. It wants to follow what Americans did with the internet, when they developed the technology and established protocols for its use, facilitating their companies to become world’s references. The Chinese want to do the same now and, besides AI 2030, count on another audacious plan: the Belt and Road Initiative.
Belt and Road Initiative
The Belt and Road Initiative is vital in order for Made in China 2025 e AI 2030 to achieve their goals. Launched in 2017 by Xi Jinping, the program aims to integrate near 70 nations from Asia, Africa, and Europe economically, creating commercial routes among them through heavy investments in infrastructure, notably railways, highways, and seaports. Those countries correspond to approximately 65% of the world’s population and 40% of the global GDP. However, BRI has been expanding and Latin America became a Chinese target too.
Comparisons with the Marshall Plan (developed by the US to help reconstruct Europe after the World War II) are inevitable but unfounded. Starting with the amount of capital involved: while the American program has cost something like U$130 billion in today’s currency, figures around BRI are estimated in U$1 trillion, although it cannot be confirmed. Another difference is that 90% of the money in the Marshall Plan was donated, while Belt and Road’s capital are loans. Finally, MP was all financed by the American government and BRI has both public and private investments.
However, China’s real objectives go way beyond infrastructure projects in other countries. With the initiative, the Asian Giant intends to expand its influence globally and open new markets to scale its industry. According to a report from the Lowy Institute, an Australian think tank, the Chinese government has three main interests with BRI:
1) Stimulate the development of less well-off regions of the country through an integration with neighbor nations (that is the case, for example, with the China-Pakistan Economic Corridor, which links the underdeveloped region of Xinjiang to a seaport in the Pakistani city of Gwadar);
2) Promote national industries, especially tech, helping to implement Chinese standards in order to end the American standards hegemony (there is a popular domestic saying that “Third-tier companies make products, second-tier companies make technology and first-tier companies make standards”);
3) Dealing with the excess capacity of production of materials like steel and cement, not only through exportation but especially migrating whole production facilities to countries that need to build their infrastructure. This way China can focus on high technology sectors.
The BRI, however, is being criticized constantly throughout the world. Some say that the plan is confused and lacks transparency. Others allege that the majority of projects never left the drawing board. In a lot of countries concerns are being raised about their capacity to pay for the loans (in Sri Lanka, for example, Hambantota Port, financed by Xi Jinping government for U$1.1 billion, had to be sold to China in order to pay for the debt; ex-president Mahinda Rajapaksa lost his reelection because of it). In Malaysia, the excess of Chinese’s influence is also causing a lot of discomfort among inhabitants. They claim that it’s a one-way relationship, meaning that, instead of pushing the local economy, they import employees and equipment from their own country, making benefits move around themselves.
Yet, who thinks that China’s influence is limited to modest Asian and African nations is completely wrong. A Bloomberg report shows that, in the last 10 years, Chinese invested approximately U$318 billion in all kinds of European assets, 45% more than Americans did. There were around 360 companies acquired from different sectors, besides airports, seaports, wind farms, and even soccer clubs, without counting equities in big corporations, like Daimler AG, owner of Mercedes-Benz.
For everything presented in these two articles, it’s clear that we can’t underestimate China’s intentions and capabilities to become the world’s superpower in the next decade or two. But it’s also obvious that other nations are well-aware of the Chinese aims and are moving to avoid losing ground. Heavy accusations and possible tariff barriers from Donald Trump intends to brake Xin Jinping’s plans, much more than equalizing commercial balance between countries. We just have to wait and see if those attitudes are going to be enough or if more strict ones are coming.