Semiconductor chips have quickly become strategic resources for China and the U.S. to leverage in their race for digital dominance. This is in large part due to the close association between semiconductor chips and Artificial Intelligence (AI). Global shortages are coinciding with increasing demand and geopolitical turmoil, forcing states to redraw their supply chains and rethink their AI development strategies. A recent report issued by the National Security Commission on Artificial Intelligence argues the U.S. needs to secure domestic semiconductor chip manufacturing in order to stay ahead of AI competitors. While the Trump administration was adamant about protecting the American semiconductor industry, alarms are being raised in Washington about possible economic and security threats that may emerge if the U.S. loses its technological edge in the future. Efforts to secure U.S. semiconductor dominance are gaining momentum, as demonstrated by the Biden administration’s recent executive order on supply chains.
Semiconductor Chips & Artificial Intelligence
The emerging field of AI technologies helps underscore the importance of global semiconductor supply chains. AI mainly refers to algorithms developed through machine learning, a process which entails receiving and interpreting data inputs to make decisions and imitate human actions. However, much of the future promises of AI lie in deep learning. Deep learning is much more sophisticated than machine learning because it bases decision making on artificial neural networks modeled on the human brain. Deep learning algorithms use vast amounts of data to teach themselves and perform tasks. For example, advances in facial recognition systems are modeled around neural networks learning to identify visual patterns by comparing them to stored data. The same technologies are used to train self-driving cars and medical evaluation scanners. For deep learning to work and evolve, fast and effective semiconductor chips, as well as massive amounts of data, are crucial. Higher-end processors provide faster learning capabilities, which becomes important for states wanting to stay ahead in AI.
With deep learning, AI has the ability to lower costs and increasing efficiency in fields such as manufacturing, social welfare, finance, and education. This makes AI an attractive field of investment for countries like China that face slowing economic growth. China’s aging population, strained social welfare system, and urban-rural divide are all barriers to the country’s future economic development that could be alleviated by implementing AI technologies. The automation of manufacturing could lead to increased productivity in industries that suffer from labor shortages, while increased access to automated education and healthcare could lower costs and expand access to social welfare. Estimates from McKinsey classify China as the quintessential beneficiary of AI developments, stating half of all manufacturing industries in China could be automated. Another advantage for China is access to a vast domestic and foreign consumer market, increasing the speed of data collection and AI learning. In sum, AI could help support China’s transition from a labor-intensive export-led economy to one experts consider to be capital-intensive and consumption driven.
Debunking Myths of AI and Economic Growth
China’s drive to adopt AI is deeply intertwined with the promises and benefits the technology is projected to offer society, however, the widespread impacts AI can potentially have on automation have been greatly exaggerated by AI consultancies and advocates. In a study of productivity increases through automation, Jeffrey Funk observed 40 notable AI startups with various goals of productivity and concluded that the majority of automation would affect white-collar office work. The direct results of automation in healthcare were limited, while the scale of change in advanced manufacturing and engineering projects was even slimmer. Funk’s skepticism is shared by Lee Vinsel, and both authors believe AI is still in an early stage of infancy and that policymakers and leading tech companies are drumming up the “hype” around AI’s potential to spark the 4th Industrial Revolution. It is more realistic to view the global AI competition as a goal for manufacturers to create international standards and achieve dominance in foreign markets. The leading AI exporter can leverage their position to sell cheaper existing AI infrastructure and technologies to foreign markets instead of helping them build their own. China has already laid bare its plan for continuous AI exports with the Digital Silk Road, offering advanced systems of communications, FinTech, sensors, and underwater cables to collect data with the ultimate goal of exporting AI systems and securing developing countries’ future integration. With U.S.-China relations growing increasingly hostile, whichever country dominates digital infrastructure development could eclipse the other.
The belief that Beijing’s inclusion in the global economy would lead to liberalization in China was all but abandoned under President Trump. However, it is only now, at a time when the U.S. is reaffirming its global presence by rallying allies against China, that the world is seeing a return to Cold War-style containment tactics. As an increasingly protectionist Washington is attempting to maintain its technological advantage, it is becoming harder for China to invest abroad and acquire intellectual property. In addition to its lead in semiconductor chip technology, the U.S. is also leading the world in top AI research. The country attracts the most foreign AI researchers due to its high ranking academic and research institutions. China may have a talent pool increasingly viewed as the world’s largest, however, a majority of global AI researchers go on to work, live, or study in the U.S. Washington’s future goals will be to hold on to their leading position long enough for China’s future export advantages to diminish, or perhaps implode, as U.S. providers become more in demand.
The world needs to think critically about how it views and understands the phenomenon of AI. It should not be considered a world-altering technology that will fundamentally transform societies in the short timeframe outlined by policymakers. The U.S. and China are both pursuing strategic and economic opportunities by creating universal standards in AI. Advanced semiconductor chips will be essential for any country trying to corner the market. Washington is hedging its bets while it has a clear advantage over Beijing and has learned from past mistakes in the deployment of 5G. Beijing knows its AI presence is key for the country’s future economic growth and global ambitions, and is hoping steps it took during the past four years of American isolationism will provide enough time to keep an increasingly concerned Washington away from emerging markets. Ultimately, this situation could lead to AI reinforcing an emerging digital divide, with societies and individuals caught up in geopolitical interests.