When it comes to AI powers, the United States is probably the most intuitive answer. Because the development of AI requires the integration of academic, economic, and international competitiveness, and even its application in national defense, industry, people's livelihood, etc. requires funds, technology, and talents. In the past, the United States has always dominated the development of global science and technology. The big winners in the AI wave—OpenAI, NVIDIA, Meta, Google, etc. all come from the United States.
According to France's "Challenges", the development of AI in the future is likely to enter a new era of multi-polarization, including China, the Middle East and other places, which may overtake the United States. What are the hidden worries of the United States, which currently dominates the world? What capabilities do other countries have that can aspire to the technological hegemony of the United States?
If the United States follows the old path of seclusion and traditional energy, it may not be competitive enough.
According to statistics from the British independent media "Tortoise Media" in "The Global Artificial Intelligence Index 2024" (The Global Artificial Intelligence Index 2024), the United States has 27% of the world's high-level AI talents (at least full-time AI researchers with a master's degree or above), ranking first in the world. one.
But the hidden worry lies in immigration policy. A report by the National Security Commission on Artificial Intelligence (NSCAI) pointed out that nearly 70% of top machine learning researchers believe that visa and immigration policies are the biggest obstacles to attracting international talents, such as the approval of the H-1B technical work visa. The rate drops from 46.1% in 2021 to 14.6% in 2024, or those holding international student visas need to prove that they intend to leave the United States after graduation, etc., which are a major barrier to transnational talents.
NSCAI called on the authorities that "immigration reform is a national security priority" and that strengthening international cooperation can promote innovation. However, if Donald Trump implements a "lockdown strategy" in the future, the United States will fall into a dilemma of working behind closed doors.
On the other hand, AI is a power-hungry monster. According to a survey by the Institute for Progress (IFP), a non-partisan think tank in the United States, every time generative AI is used, it consumes 10 times the power of traditional web search. The United States may not be able to load the existing power infrastructure, and since Trump took office, his policy has focused on natural gas and oil extraction. Although it can currently supply domestic demand, if Trump revokes environmental protection regulations in the future, it means that the country will be even more negative about renewable energy. Whether it can meet the increasing power demand of AI will be a challenge.
Competing Candidate 1 - China: Affected by U.S. restrictions, it will accelerate technological independence
As the main competitor of the United States, China ranks second in the "2024 Global AI Index", far surpassing other countries.
Although affected by the U.S. export ban, it is difficult for China to obtain key hardware such as high-end GPUs (graphics processors). China's response is to bypass the black market to buy hardware and more actively promote the development of domestic basic science and technology.
Huawei, which was the first to bear the brunt, saw global revenue decline by 30% and profits by 70% in 2021. However, revenue rebounded in 2023, with an annual increase of 9.6% and profit growth of 2.4 times. In addition, Microsoft's withdrawal from China has allowed Huawei to recruit its original engineers, adding 10,000 to 2,000 employees. It was also revealed that it will mass-produce AI chips in 2025. This all shows that Huawei has not been annihilated by the United States. Instead, it has allowed it to fight against the odds, making its own hardware that it cannot buy, and strengthening its vertical integration capabilities.
And as the West tightens restrictions on China's access to sensitive technologies, Chinese companies are poaching top engineers in the fields of advanced semiconductors and AI through high salaries. For example, Huawei offered up to three times the salary of Zeiss SMT employees in Germany in an attempt to poach employees, which led to the German intelligence agency launching an investigation and worrying that Huawei might use this to gain access to advanced technology secrets.
The outside world has different views on China being suppressed by the United States: they believe that China's scientific and technological strength will be forced to grow rapidly, or that it is just a paper tiger that promotes national power and is superficial. But no matter which one it is, the US ban itself identifies China as a threat.
China has set AI development goals and vowed to become a global leader by 2030. The third phase of the "National Grand Fund" to encourage the development of IC is estimated to invest US$47.5 billion, ranging from the construction of national data centers to special funds for local governments. Support, China has formed a development strategy of "integration of government and enterprise". Although the United States still has the advantage in the short term, in the long run, whether China has the ability to develop technologies restricted by the United States determines whether China can catch up with or even surpass the United States. .
Competitive Candidate 2 - Middle East: Combining energy, capital, and talent advantages to become the next AI champion
Another thing worth noting is the Middle East, which has made its fortunes on oil. Saudi Arabia has a "2030 Vision Plan" (vision 2030) and is currently planning to invest US$100 billion in AI. We hope that by then AI can account for 12% of the country's GDP and get rid of oil dependence; the United Arab Emirates has established a partnership with OpenAI to integrate technology into application scenarios, such as health care, finance, public services, etc.; Kuwait and Qatar each invested 8,000 billion and US$475 billion.
Large sums of hot money will flow into academic institutions to train talents, develop data centers, support new innovations, or cooperate with technology giants. For example, Saudi Arabia has agreed to invest US$5 billion to US$10 billion with Google to develop the Arabic language. Large-scale language model to build AI infrastructure.
These Gulf countries are not only rich but also powerful. For example, Israel has always had high-quality manpower and its population is less than 0.2% of the world's population. However, it accounts for more than 20% of the world's Nobel Prize winners. It is also known as the "country of entrepreneurship" and the Israeli government and private sector have different national conditions. Special, special data can be obtained, including communication records, public data, and other sensitive information related to homeland security. These highly sensitive and real-time data streams can be directly used to develop and test the real-time decision-making and prediction capabilities of AI, such as In security threat early warning systems.
Middle Eastern countries maintain a delicate balance with China and the United States. Although they cooperate with American technology giants, they also accept telecom infrastructure support, cloud and big data solutions from Huawei and Alibaba. There is an atmosphere between the three countries like a snipe and a clam at odds with each other. , and Saudi Arabia publicly stated that the Middle East has geographical advantages, abundant energy, and the ability to provide large sums of money for aid, and its goal is to be the next AI champion.
Source: Challenges 2024.11.21, WSJ, Tortoise, IFP
