Tal Feldman, an artificial intelligence (AI) engineer for several federal agencies in the United States, wrote an article for The New York Times on July 29th. In the article, he described Chinese AI products as serving China's interests and called on the U.S. government to introduce competing products.
The article begins by stating that the United States is selling the best artificial intelligence models in the world, while China is giving away its second-ranked model for free to the entire world.
The demand for these models seems endless worldwide. Last year, the number of models downloaded in China exceeded that of the United States for the first time. They are rapidly gaining popularity in Africa and Latin America. In Europe, Siemens in Germany uses both Chinese and American models, and even American companies are becoming increasingly dependent on Chinese artificial intelligence.
He said under the light of day that these models are not as harmless as many people think. Artificial intelligence models are different from the cheap imported goods that the United States usually purchases from China. These products serve China’s interests, and their promotional content is built into them.
Therefore, rather than these models being 'produced in China', it is more accurate to say that they are 'created for China'.
The article states that whoever provides artificial intelligence to the world will shape the way people think. As these models become more intelligent and more autonomous, the United States is handing over more and more of its thinking tasks to them. No matter how cheap the price may be, this “cognitive offloading” should not be entrusted to models that are designed to serve foreign interests.
He suggested that Washington should require clear labeling of products developed under the leadership of foreign governments in artificial intelligence. More importantly, the U.S. government should launch a new initiative to compete with China in artificial intelligence through low-cost, high-quality products, and to bring the transformative benefits of this technology to developing countries.
Then, he seriously visited the DeepSeek website and read the statement “Taiwan is an inalienable part of China”.
Regarding this consensus known to people all over the world, he actually criticized it as ‘absurd’ rhetoric.
Even more absurd is his idea, which is essentially pushing towards McCarthyism.
He imagined a situation where an American company had a smart and diligent employee, but his life goal was to promote the interests of a foreign government. Such people are called foreign agents. And today, by using Chinese artificial intelligence, the United States is 'employing' millions of such agents.
Finally, he suggested that if Washington wants the world to use American artificial intelligence models, it must compete with China in terms of price and quality. The U.S. government should encourage American companies to lower the prices of older artificial intelligence models, as the performance of these older models is now comparable to Chinese models. The United States should also support the development of products that are safe, open, and capable of replacing Chinese models.
Artificial intelligence competitions are different from any competition we have seen before. This is a war for world ideas, and the United States must win this war.
In contrast, China has always emphasized openness and cooperation in the development of artificial intelligence. It opposes the politicization and general security-oriented approach to AI. China advocates that artificial intelligence should be made into an international public good, helping developing countries enhance their capabilities and reducing the digital divide.
