Commentary: China and U.S. Place Their Chips on an AI Windfall

On the second day of his second term, President Donald Trump signed an executive order to initiate a national strategy concerning artificial intelligence (AI), underscoring its importance for the United States' economic and national security interests [para. 1]. AI is described as a formidable asset capable of generating vast wealth but also presenting challenges to social cohesion. In February 2019, during his first term, Trump had introduced the “American AI Initiative” which focused on several areas including increasing AI research funding, making government data more accessible, fostering AI talent, and encouraging international cooperation. Yet, the plan wasn't fully realized [para. 2].

Four years later, Trump, shortly after assuming office again, disclosed a $500 billion investment venture targeted at bolstering AI data centers across the nation, in collaboration with entities like SoftBank, OpenAI, and Oracle [para. 3]. Masayoshi Son, SoftBank's CEO, emphasized AI's immense potential, claiming it could surpass human intelligence by tenfold [para. 4]. OpenAI has gained significant recognition through its development of ChatGPT, which leverages advanced AI chips and energy-intensive data centers to improve chatbot efficiency and reliability [para. 5].

In September 2024, OpenAI’s CEO, Sam Altman, appealed to the Biden administration for enhanced energy support tailored for AI infrastructure. Oracle facilitates OpenAI by enhancing its model training capabilities, with Arm Holdings, Microsoft, and Nvidia serving as pivotal technical partners [para. 6]. Robust capital influx from Wall Street and global investors has fueled an increase in the market value of these firms by 6%, with Nvidia particularly benefitting as its market value briefly surpassed $3.61 trillion, overtaking Apple [para. 7]. The AI advancement has significantly impacted the data center AI chip market, reaching $17.7 billion in sales by 2023 [para. 8].

The challenge to US supremacy in AI comes from a Chinese company introducing DeepSeek, an open-source AI model that promises cost efficiency and performance superior to top US AI technologies. Employing AI-enhanced large language models (LLMs), DeepSeek claims to run 96% cheaper than ChatGPT-4o by optimizing to operate with minimal high-end chip resources [para. 9]. While DeepSeek has not entered the US AI market, its affordability and open-source nature may become a threat to US competitors, potentially altering the competitive dynamics in AI technology [para. 10].

Further, the US seeks to maintain its technological lead by curbing the export of certain advanced technologies to China, resulting in distinct AI ecosystems developing in both countries. The US aims to strengthen domestic manufacturing amid dependence on global supply chains, whereas China aspires to a self-sufficient AI architecture, a motive that has drawn the interest of Southeast Asia, which commands significant portions of the global chip assembly landscape [para. 11].

A Federal Reserve memorandum highlighted China's chip production focus on conventional chips due to restricted access to high-end chip technologies, which nonetheless hold substantial value within Europe and America [para. 12]. Constructing an influential AI ecosystem involves more than advanced AI chips; it's reliant on an infrastructure consisting of sophisticated hardware, software, and a competent workforce of AI specialists and engineers [para. 13].

The future trajectory of AI and semiconductor industries will demand both specialization and scalability, with predictions indicating a 41% growth in the AI market between 2025 and 2026 [para. 14]. The Trump administration's renewed emphasis on AI and tech leadership likely signifies immediate growth prospects for tech companies and sustained interest in the sector, though market dynamics remain ever-shifting due to ongoing innovations as predicted by Moore’s Law [para. 15].