Another week has passed, and uncertainty remains over the export of Nvidia’s advanced artificial intelligence chips to China. Those who support continued export controls believe the chips will help build Chinese military systems that threaten the United States and its allies. They believe that artificial intelligence chip control is also needed to maintain and expand the United States’ leading position in the artificial intelligence service market.
But they were wrong. These arguments assume that China cannot succeed in the field of artificial intelligence without these advanced artificial intelligence chips, but this is not the case.
Advanced AI chips simply reduce the cost of AI. Today’s most advanced AI models require massive amounts of AI chips to build and run. Advanced chips, with higher performance; therefore, fewer are needed to achieve the same AI performance.
But AI costs can be reduced in other ways. As DeepSeek shows, clever software and algorithm design can drastically reduce the number of AI chips needed. China’s decision to open source its AI models will, among other things, enable it to leverage the best software and algorithms to reduce the cost of AI. Second, AI chips only account for a portion of the overall cost. AI-based systems also incur other costs—engineering, data, software and licensing, regulations, energy, and infrastructure—where China has a considerable cost advantage. Third, AI hardware performance depends heavily on packaging and interconnection—how AI chips are assembled and connected. China can leverage its world-class advantages in both areas to achieve high performance. recently announced Huawei Super star clusters are more powerful than any star cluster NVIDIA system, despite not using state-of-the-art artificial intelligence chips.
Advanced chips also reduce the power costs of AI. These chips are manufactured using the latest technology from TSMC (and sometimes Samsung) – each new technology is more energy efficient than the last. The high power consumption of AI systems worsens both the monetary cost and the speed of deployment, as obtaining large amounts of power quickly is challenging, especially in the United States. However, China’s electricity supply is growing much faster than the United States and is more likely to successfully meet the power needs of its AI data centers, even if they consume more power due to lack of access to advanced AI chips. High power also results in a larger carbon footprint, but it shouldn’t limit China’s ambitions in any technology it deems important.
In addition, many AI applications do not require advanced chips. A variety of applications in cybersecurity, facial recognition, medical image analysis, advanced driver assistance systems (ADAS), logistics, and robotics can all be handled using AI models that are much simpler than state-of-the-art models. These models can be built and run on chips produced in China. China aims to dominate these applications. Even for more complex applications, recent research shows that state-of-the-art models can be replaced by a set of simpler models. The collection does not require advanced artificial intelligence chips to build and run. Therefore, it is unclear whether China will fall behind in these applications.
It’s unclear whether the development and use of future state-of-the-art models will require advanced chips. There are signs that the advantage of the most advanced models is stabilizing. Considering these models require significant investment, future models may be different and use fewer resources, including chips. Even if the use of advanced AI chips is controlled, this will further level the playing field. It is also possible that China will learn how to produce advanced AI chips on its own – it has certainly invested in several technologies that have the potential to surpass the state-of-the-art.
Overall, China can significantly mitigate the disadvantage of not having access to advanced AI chips. Furthermore, China is willing to bear any higher upfront costs, especially for AI-based military and strategic technologies, because they know they can reduce downstream costs through scale and manufacturing advantages. Not surprisingly, despite the implementation of AI chip controls over the past few years, China continues to produce competitive, state-of-the-art models and dominate AI-based applications such as robots and self-driving cars.
There may still be some truth to the debate over AI chip control – if there are no costs, why not take advantage of China’s increased costs for AI development, no matter how small. But the cost is huge. China could have become one of the largest markets for U.S. advanced artificial intelligence chip companies. The United States has lost its market. Second, the control of AI chips makes this a matter of national pride and triggers a wave of investment in China’s domestic AI chip ecosystem. Even if chip controls are reversed, it’s unclear whether the United States will regain market share. China has also retaliated in a variety of ways, measures that have further harmed the U.S. economy and geopolitics.
If the United States wants to lead in artificial intelligence, chip control is not the answer. Instead, it should focus on improving the innovation, investment, energy and regulatory ecosystems. It should make it easier for the world’s best AI scientists to live and work here. It should diversify, strengthen and secure the AI supply chain. It should work with allies to lead the development of international AI standards and practices. It should reduce the cost of AI (e.g., through selective open source or public-private partnerships) to ensure that American AI (and its values) are most ubiquitous. It should prioritize high-end and enterprise applications, which have wider moats, to compete with talent and resource-rich fast followers who have cost and speed advantages.
The value of AI chip control has been overstated. These controls did little to slow down China’s growth and caused significant economic and geopolitical damage to the United States. Now is the time to abandon these controls and focus entirely on sustaining and growing AI leadership through innovation.
The views expressed in Fortune opinion pieces are solely those of the author and do not necessarily reflect the views and beliefs of: wealth.