Winner Takes AI

Gunner Lee • May 23, 2025

Artificial intelligence (AI) has become a central issue in the new Cold War between the United States and China. OpenAI, the American company responsible for developing ChatGPT, urged the U.S. to prioritize AI development in a 15-page “economic blueprint” document released on Jan. 13. It outlined their vision for the future of American AI, describing the new technology to be as revolutionary as the invention of the automobile (CNBC). They highlighted the potential benefits of investment into the industry: the creation of data centers would in turn lead to the creation of a new job market, the modernization of electricity infrastructure and the expansion of local economies (OpenAI). They suggested developing “common-sense” regulations, such as using AI to improve education and creating AI that is shaped by and promotes American ideals. OpenAI warned that unless the U.S. steps up, AI will be dominated by China and expand the Chinese Communist Party's global influence (New York Times).

Generative AI has existed since the mid-20th century, developing alongside but slower than computer technology. In the 2010s, deep learning, a type of AI trained on a large dataset which is capable of recognizing complex patterns, was revolutionized with superior graphics processing unit (GPU) power and the emergence of a stronger algorithm. As the vast majority of generative AI utilizes the deep learning system, the 2010s marked a period of rapid development for the emerging technology. It was in 2020 that the AI boom began in earnest, with deepfake and image-generating programs becoming mainstream (The Atlantic). The release of ChatGPT in Nov. 2022 prompted major companies to invest in AI; Google, Meta and Apple are some well-known examples. Chinese companies also began racing to create their own AI, efforts aided by a supportive government (UN).

OpenAI’s economic blueprint was published days before the inauguration of President Donald Trump. OpenAI CEO Sam Altman was formerly critical of Trump, posting on his personal blog how he considered him dangerous (ABC News). However, with Trump’s reelection, Altman ingratiated himself with the president, even making a personal donation of $1 million to his inaugural fund (Reuters). The new administration has been open to AI development: as early as 2018, Trump supported U.S. dominance in the then-nascent industry (White House). Following reelection, he undid Biden’s executive order regulating AI development and is collaborating with OpenAI on the Stargate Infrastructure Project.

In December, Trump struck a deal with SoftBank, a Japanese investment management company, to invest hundreds of billions of dollars in U.S. projects. An immediate $100 billion investment is being used to construct AI data centers in Texas and is expected to create 100,000 jobs (CBS News). Elon Musk, who had been a key figure in OpenAI’s early days, soon fell out with Altman over disagreements regarding the leadership and future direction of OpenAI and later founded a rival company, xAI. Compared to ChatGPT, xAI’s model Grok 3 can hold more complex conversations, produce real-time information, generate uncensored images, and provide sources for its research. It is also directly connected to X (formerly Twitter) and trained to be anti-“woke” (Business Insider). Despite the two now serving Trump in intimate positions, Musk continues being hostile towards Altman, posting on his social media platform X that the government has nowhere near the amount of money promised for the project, although this claim was disputed by Altman (AP News).

In a bid to get ahead in the race for AI, the Biden Administration made several efforts to restrict Chinese access to Nvidia’s powerful AI chips, or computer chips used to process the large amount of data required for AI (Reuters). Additionally, OpenAI blocked their tools and services in China on July 9, which prompted a surge in support for domestic AI companies. Of the several Chinese AI companies, DeepSeek, a relatively unknown startup, emerged as OpenAI’s true rival. Despite being built with inferior technology and a significantly smaller budget, built with $6 million opposed to $100 million for OpenAI, their DeepSeek-R1 model released in mid-January outperformed all major U.S. AI models in accuracy in complex problem-solving, garnering massive attention in Silicon Valley (NBC News). If the Stargate Project succeeds, OpenAI will reclaim and keep their previously unopposed hegemony.