[The Editors' Verdict] The Tariff War and DeepSeek

The ‘tariff war’ among the United States, Mexico, and Canada was merely a prelude. The curtain has now risen on the ‘tariff showdown’ between the two countries with the greatest impact on the global economy, the G2 (United States and China). Amid an inflation of aggressive terms like ‘tariff bombs’ and ‘retaliatory tariffs,’ countries are rushing to raise trade barriers.

The future is uncertain. However, history shows that the outcomes of tariff wars have been disastrous. A representative case is the ‘Smoot-Hawley’ Tariff Act introduced in the United States in 1929. During the Great Depression, Senators Reed Smoot and Willis Hawley proposed this law to protect the U.S. economy. As protectionists, they imposed high tariffs of over 40% on about 20,000 imported goods. Major countries responded with retaliatory tariffs, further shrinking global trade. This law deepened the Great Depression.

Free trade has expanded the global economic pie. Free trade between countries increased productivity and promoted the spread of technology and knowledge. Specialization and division of labor were achieved in each country, such as Korea’s semiconductor industry, Germany’s automobile industry, and Vietnam’s textile industry. This led to overall efficiency improvements. It also increased interdependence among countries, contributing to the promotion of peace.

An image depicting open source contrasted with its opposite concept, closed source, as illustrated by AI. ChatGPT (DALL·E 3)

The rise of Chinese artificial intelligence (AI) startups amid the tariff war is therefore remarkable. DeepSeek demonstrated the potential of ‘affordable AI models.’ It cast doubt on the existing AI evolution model, which required more hardware, more data, more power, and more investment to evolve. When prices drop, demand increases. Various companies, startups, and individuals can adopt and utilize AI. AI prices must be kept low to achieve high penetration rates.

DeepSeek was able to offer ‘cost-reducing AI’ thanks to the concept of ‘open source’ in the IT industry. This means transparently disclosing the principles and code of software so that anyone can use, modify, change, and redistribute it as they wish. In other words, it is a way to create and enhance wealth through open collaboration, inclusiveness, and transparency.

Yan LeCun, Meta’s Chief AI Scientist and a professor at New York University, said, “DeepSeek’s success should be interpreted not as the success of Chinese AI but as the success of open-source models.” On the 25th of last month, he stated on his social media, “DeepSeek developed its ideas by utilizing existing open-source models (LLaMA),” adding, “Others can also gain benefits from the open-source DeepSeek.” He emphasized, “This is the power of open source.”

Open-source frameworks like Google’s TensorFlow, Facebook’s PyTorch, and Hugging Face have revolutionized AI research and development. Because there were no barriers to accessing AI-related materials, anyone could jump into AI development. The transparency of public code enabled rapid detection and correction of errors, and collaboration among developers worldwide has produced unprecedented advancements. Amid the tariff war, the AI industry has demonstrated how openness and transparency can be powerful drivers of innovation.

This content was produced with the assistance of AI translation services.