Rachel Cifu and Andrew DeWeese traveled to Taiwan this summer to learn more about the country’s stronghold on the semiconductor industry and its complex geopolitical position. They explain how their preconceptions of Taiwan and its people quickly changed when speaking with leaders in business, technology, and policy.
With a table full of Taiwanese cuisine and sparse American classics, the President of the Yale Club of Taipei welcomed us to the city by weaving together stories of New Haven pizza places and litigation battles in defense of the Taiwanese semiconductor industry. We discussed everything from Timothy Dwight College to TikTok to Taipei’s night markets as Nick Chen BA ’79, our gracious host, epitomized both the hospitality and genuine curiosity that characterized all the Taiwanese people throughout our trip.
With the support of the Brady-Johnson Program in Grand Strategy, we both traveled to Taipei as part of our summer research projects, where we studied very different but interconnected issues. Rachel focused on the Taiwanese AI industry, looking specifically at how Taiwan Semiconductor Manufacturing Company’s (TSMC) monopoly on next-generation semiconductors feeds into domestic AI development and the national security concerns that arise from both areas. Andrew spent the summer focusing on Taiwan’s geopolitical role in other countries’ national security strategies, especially that of India. His brief few days in Taipei capped a summer in Washington and several weeks in India interviewing government officials all across the country to determine precisely what kind of support Taiwan could expect to receive, as well as what agency the ROC government had to affect that allocation. Having already lived for several months in Taipei last summer, Andrew leveraged the return opportunity to deepen relationships with relevant contacts and assess how public opinion had changed in the intervening year and increased cross-strait tensions. We each interviewed experts in government, academia, technology, research, business, law, and think tanks to try to better understand the local perspective on issues that Western media often talks about rather than with.
While our projects were different in scope, this central question of the “Taiwanese perspective” immediately proved to be flawed. We each asked Nick Chen for his initial thoughts on our broad research questions, to which he replied: “There is no Taiwanese perspective, I don’t know what you’re talking about.” His discomfort stemmed from the spectrum of political, economic, and security beliefs held across the Taiwanese population and instantly complicated the direction of our projects. He argued that by grouping together the entire island, we were simply buying into Western ignorance and belittling the agency of individual Taiwanese people and leaders. This knee-jerk reaction to our academic interests forced us to rethink the types of questions we should be asking to whom and what insights we were looking to gather from each interview.
Instead of asking the founder of Taiwan AI Labs general questions about Taiwanese goals in certain types of AI, we focused on his efforts to use LLMs to counter mainland Chinese disinformation campaigns and to improve Taiwanese healthcare systems through his AI Federated Learning Alliance. We asked a director at the Industrial Technology Research Institute — the Taiwanese government’s R&D powerhouse — about how national strategic tech priorities drove certain pilot projects with TSMC and other parts of the chip supply chain. We questioned a member of the AI Center of Excellence and a policymaker on the National Science and Technology Council about their impending AI regulation and incentive programs. Each person we met was more eager than the next to talk about their life’s work and their individual contributions to Taiwan’s technological advancement. While many of their goals for the future of Taiwan overlapped, they all had very different perspectives on more political issues: the role of U.S. export controls on TSMC’s most advanced chips, the potential involvement of the U.S. military in a worst-case scenario, their Presidential elections in January 2024 and, most divergently, the relationship between Taiwanese chip and AI companies and manufacturers/consumers on the mainland. The openness with which everyone spoke, even on some of the most complex and controversial geopolitical problems of our age, stood out to us as a unique unifier and as something we felt comfortable deeming “the Taiwanese perspective.”
Andrew found similar trepidations when speaking with Indian officials about their assessments of “the Taiwanese perspective.” They viewed the situation in Taipei as enormously complex and difficult to read, strained further by the divergent interests of average Taiwanese people, KMT business leaders, ROC official diplomatic policy, and current DPP leaders. Rather than try to mind-read a multifaceted third party, India preferred to focus on the one “perspective” and “interest” it knew with full certainty: its own. Every diplomat spoke in terms of interest, concern, and capability — when Delhi and Taipei found overlap, they could work together. Where such failed, they did not even try — to the frustration of Taiwan’s senior diplomat to the U.S., as she referenced in Washington.
Where our projects overlapped was in the necessity for small-scale strategy implementation and outcome-driven policy, rather than the broad-based ideological discourse that characterizes the media’s treatment of Taiwan today. In the AI sphere, this meant development of an LLM trained on traditional Chinese characters instead of the simplified Chinese characters typically used in Beijing. We spoke with the director of this project, who gushed about the collaboration between the NSTC, National Taiwan University, Taiwan AI Labs, ITRI, and several private tech companies that drove his model’s success even with substantially less data than is available on the mainland. This aggregation of resources from the public, private, and research sectors in pursuit of the geopolitical goal of countering mainland influence and retaining Taiwanese control over its AI models is a prime example of successful grand strategy with tangible real-world benefits. We realized that discourse is only effective if it mobilizes strategic collaboration and outcome-driven plans. In the India sphere, all diplomacy with Taipei focuses on tangible outcomes — primarily, new trade deals, business opportunities, and specific investments. These outcome-driven plans can operate successfully while larger questions of ideology remain both contested and unanswered.
Speaking with leaders on the ground who are actively working to bolster Taiwanese industries and security humanized the issues we discuss in our Yale classrooms thousands of miles from the heart of the issues. These conversations brought us a much more nuanced understanding of different perspectives in Taiwan, of the ways that our actions in the U.S. directly harm or help the strategies employed on the ground, and of the genuine kindness that so many Taiwanese embody daily.
Rachel Cifu and Andrew DeWeese are Yale College students in the Brady-Johnson Program in Grand Strategy