Is Silicon Valley Defecting?

Several recent articles have addressed this issue.

Guancha provides an interesting look at recent articles in US media that ask several questions: Why does Silicon Valley admire China? Why are so many Silicon Valley AI start-ups and other established corporations favoring Chinese AI over American systems? And then we have the biggest recent declaration by Nvidia’s CEO Jensen Huang’s saying, “China will win the artificial intelligence (AI) race.” The Chinese author, Liu Chenghui, uses several American media sources to supplement his narrative, which is becoming a bigger newsworthy development after the theft of Nexperia. There’s also increasing chatter about OpenAI needing a government bailout and other financial problems of the many overvalued American AI companies. There’s another important geopolitical development that needs to be reported I’ll provide on Monday. For now, the Tech War clearly shows the Outlaw US Empire falling further behind China and Russia. I suggest Warwick Powell’s recent article on the AI race for additional reading after finishing this:

Nvidia CEO Jensen Huang’s sentence “China will win the artificial intelligence (AI) race” has once again aroused public opinion’s attention to the competition between China and the United States in this field.

“How much technology in Silicon Valley is built on China’s AI?” A Bloomberg article published on this topic on November 9 cited a number of examples to illustrate that China’s low-cost open source AI models are not only attracting global users but are also quietly gaining the favor of Silicon Valley. The author believes that if the U.S. government wants to stay ahead for a long time, it might as well think about the reasons why Silicon Valley has defected to China’s AI.

Huang recently bluntly told the Financial Times: “China will win the AI race. He then urgently made up for it, saying that China was only “nanoseconds” behind the United States, and emphasized that the United States must stay ahead “by accelerating progress and winning global developers.”

According to the report, although Huang’s initial statement was oversimplified, his motivation for his pessimistic view of the future of the United States is also obvious. Over the past year, he has argued that the United States should allow his company to continue selling chips to the Chinese market, despite Washington’s tightening of export controls. But his fear that the “developer war” may have gradually lost is not unfounded.

In recent weeks, a subtle shift has become more apparent. There has long been speculation that China’s low-cost open-source AI models could attract global users from American products to Chinese products. And now there are signs that this trend is quietly seeping into Silicon Valley.

Venture capitalist Chamath Palihapitiya recently revealed on the “All-In” podcast he co-hosted with White House AI adviser David Sacks that a company he works with has migrated major computing work to the Kimi K2 model developed by Chinese company Moonshot AI. He said the open-source model is “actually much cheaper than OpenAI and Anthropic.”

Shortly after, Airbnb CEO Brian Chesky also admitted that the reason he didn’t connect his travel app with OpenAI’s ChatGPT was because the relevant tools were “not mature enough.” He said Airbnb’s new customer service system relies on more than a dozen different AI models, “which largely relies on Alibaba’s Qwen series: it’s very good, fast and cheap.” ”

Because Chesky has a close personal relationship with OpenAI CEO Altman, his remarks are particularly noteworthy.

The number of cases of public recognition of Chinese AI models is increasing.

Thinking Machines Lab, a startup founded by former OpenAI CTO Mira Murati, said in a blog post that its latest research was “inspired by and built on the research results of Alibaba’s Tongyi Qianwen 3 team.”

But what’s even more interesting is that this shift is taking root in a more stealthy way.

Cursor, a high-profile AI programming leader with a valuation of about $10 billion, launched a new version of the assistant last month. Since then, there has been a lot of speculation on the Internet that it is based on a certain Chinese AI tool such as DeepSeek. One investor even posted that when he used it, the assistant’s internal monologue was switched to Chinese.

Another U.S.-based company, Cognition AI, also valued at about $10 billion, appears to have used the foundation model Z.ai of Zhipu AI in its new programming agent. After the social media user came to light, Zhipu appeared to confirm this with a post, saying it “highlights the positive impact and value of open-source contributions to the ecosystem.”

Neither U.S. company responded to a request for comment by email. Because Chinese AI models use a relaxed open source license, companies are completely free to build products on top of them.

Data compiled by the Hugging Face platform, a subsidiary of the American alliance Atomic Project, which supports open source, confirms this trend: Chinese models have surpassed the United States in terms of cumulative downloads by developers.

The change was slow at first, but then exploded rapidly: at the beginning of 2024, Meta’s Llama had a cumulative download of 10.6 million, while Alibaba’s Tongyi Qianwen was only 500,000; By last month, Tongyi Qianwen had reached 385.3 million times, surpassing Llama’s 346.2 million times.

In addition, systems derived from Tongyi Qianwen now account for more than 40% of the newly released language models on Hugging Face, while Meta’s share has dropped to 15%.

Geopolitical concerns have been expressed in response to the influx of international users to Chinese AI tools, but for developers eager to launch products, especially in the fields of coding and software development, those so-called risks are no longer important compared to price and performance advantages. What’s more, open-source models can be downloaded, fine-tuned, and run locally, alleviating concerns about content and data privacy.

The author believes that it is too early for Huang to announce the winner of AI. The United States still has a clear advantage in obtaining cutting-edge chips and computing power, which are key elements in the development of advanced systems.

But China, driven by low-cost and open source, has indeed attracted more and more developers, and they are the core force of AI innovation.

The article concludes by reminding the U.S. government: If you really want to stay ahead of the long-term competition, you might as well ask: Why has Silicon Valley begun to defect?

Silicon Valley’s “defection” to China’s AI has long been a sign. In May this year, Huang Renxun said on the earnings call that Alibaba’s Tongyi Qianwen model is the best among open source AI models. Twitter founder Jack Dorsey also posted a post praising the code model Qwen3-Coder of the Qwen series. Even after Musk saw the pictures generated based on Tongyi Wan2.2 training, he said that the effect was “difficult to distinguish between true and false”.

A long article published in the New York Times on October 22 bluntly said: “Silicon Valley has fallen into obsession and envy of China.”

The article reports that a complex emotion intertwined with anxiety, envy and introspection is sweeping through Silicon Valley, the center of American technology. In the face of China’s rapid progress in infrastructure construction, artificial intelligence applications, and manufacturing, many Silicon Valley elites have begun to re-examine and admire China’s efficiency and execution, while reflecting on the United States’ own difficulties in infrastructure, regulation and manufacturing.

The article argues that American companies are scrambling to develop machines that are smarter than humans. However, if Silicon Valley delves into China, it will be found that China’s artificial intelligence industry is not obsessed with general artificial intelligence, and Chinese entrepreneurs are more focused on applying artificial intelligence to services, equipment and manufacturing.

Former Google chairman Eric Schmidt and colleague Selina Xu have called on Silicon Valley to reduce its obsession with artificial general intelligence, learn from its Chinese counterparts, and integrate artificial intelligence into daily life.

The author points out that Silicon Valley’s envy of China has not only sparked discussions about work models and industrial policies but also reflected the state of the United States itself, reflecting the country’s struggle after losing self-confidence.

“We may be trapped in the past when China moves at full speed, moving goods, people and information at machine-like speed.” Andreessen Horowitz of venture capital firm said. {My Emphasis]

There’s nothing the Outlaw US Empire can do about China’s tech advantage. The problem lies in how Silicon Valley approached AI, which was likely influenced by the Military Industrial Complex that favored the general AI application and presumed ability to dominate and control that aspect of AI via closed-source systems and thus keep China and the world down. As we are seeing, that was the wrong bet. The Empire can review its policy choices all it wants but it won’t be able to discover why it’s now in the pickle it’s in. It’s the hubris inherent in the drive to retain primacy, to not allow any nation to surpass it in its abilities. The damage was done long ago and that same rake continues to be stepped on. It really is quite amazing how hubris continues to blind elites from that rake—Trump stepped on it several times in his first term, then Biden followed suit many times, and Trump continued in term two. The real reason why the Outlaw US Empire’s declining is due to its slavish devotion to Neoliberal and Neocon dogma, which is why we see no difference between either party when it comes to almost all policy. What’s key is the fact that all these overvalued AI companies have no way to raise revenues because they followed the MIC’s diktat instead of producing practical products capable of generating revenues, which is precisely what China’s done. Clearly, Chinese Capitalists are more adept than American Capitalists. And China’s additional socialist system with its planning regime gives it a big advantage, one the Outlaw US Empire can’t and won’t replicate.
*
*
*
Like what you’ve been reading at Karlof1’s Substack? Then please consider subscribing and choosing to make a monthly/yearly pledge to enable my efforts in this challenging realm. Thank You!