‘Stretched thin’: Alibaba’s Qwen chief explains why US export controls give America edge over China in AI

Top executives in China's artificial intelligence sector are candidly recognizing the widening gap between their AI capabilities and those of their American counterparts. Key factors contributing to this divide include restrictions on computing power and stringent US export controls that are reshaping the competitive landscape for advanced AI models. Justin Lin, the head of Alibaba’s Qwen series of large language models, has assessed the chances of any Chinese company surpassing leaders like OpenAI or Anthropic through significant breakthroughs as less than 20% within the next three to five years. This perspective reflects a shared sentiment among leaders at other major Chinese firms, such as Tencent and Zhipu AI, one of the nation’s top LLM developers. During a panel at the AGI-Next summit in Beijing, Lin emphasized that access to computing resources—heavily affected by US restrictions on advanced chips—is becoming a pivotal factor distinguishing American AI labs from their Chinese counterparts. He remarked, "A massive amount of OpenAI’s compute is dedicated to next-generation research, whereas we are stretched thin—just meeting delivery demands consumes most of our resources." Lin's comments raise critical questions about the nature of innovation itself. He posed a classic dilemma: "Does innovation happen in the hands of the rich, or the poor?" US export controls have severely limited Chinese companies' access to cutting-edge semiconductors, compelling many to focus on immediate commercial deployment rather than longer-term research initiatives. In contrast, US laboratories are able to invest substantial resources into training advanced models and exploring breakthroughs in reasoning and artificial general intelligence. These insights come during a period of significant market activity for Chinese AI firms, with Zhipu AI and MiniMax Group together raising over $1 billion through their public listings. MiniMax’s stock surged more than double on its first trading day, while Zhipu’s shares increased by 36% following its debut. Despite this market enthusiasm, Lin's statements highlight the underlying tension between rapid market growth and the limitations imposed by insufficient computing resources, which may hinder the long-term research capabilities of Chinese AI companies.