Key Takeaways
- Bill Ackman raised alarms about China’s aggressive data center buildout threatening American democratic institutions if the U.S. falls behind in artificial intelligence development
- Moonshot AI, a Chinese startup, unveiled Kimi K3, which climbed to the top of a prominent coding benchmark, surpassing Anthropic’s Claude Fable 5
- White House adviser David Sacks described the development as worrisome, pointing to restrictive U.S. policies on infrastructure expansion and model regulation
- Nvidia shares dropped 2% Friday; Micron entered bear market territory amid broader semiconductor sector anxieties
- Forecasting platforms continue to favor Anthropic with over 67% probability of maintaining the leading AI model through year’s end
The accelerating pace of China’s artificial intelligence infrastructure development has sparked anxiety among American investors and government officials, with prominent figures now openly questioning whether the United States can maintain its technological edge.
Hedge fund manager Bill Ackman took to social media on July 15 to highlight that China faces no restrictions on building data centers. He emphasized that achieving superintelligence first is critical for American interests, warning that failure could jeopardize the nation’s security and democratic institutions.
Ackman’s statement followed the Friday, July 18 launch of the Kimi K3 model by Moonshot, a Chinese artificial intelligence company. According to reports, the system achieved first place on the Frontend Code Arena benchmark, outperforming Anthropic’s Claude Fable 5.
David Sacks, who leads the President’s Council of Advisers on Science and Technology, characterized the benchmark performance as troubling. He attributed part of the challenge to restrictive American regulations, including efforts to halt data center construction and proposals mandating government review of advanced AI systems before deployment.
Ackman’s reply to Sacks was brief but emphatic: “Agreed.”
Regulatory Hurdles Slow U.S. Progress
This week marked a turning point when New York instituted the nation’s first statewide moratorium on new data center projects. Both Ackman and Sacks contend that as American officials grapple with energy consumption concerns and regulatory frameworks, Chinese competitors face no such obstacles.
Major Chinese technology corporations such as Alibaba, Tencent, Baidu, and ByteDance have pledged combined investments exceeding tens of billions of dollars toward AI infrastructure through 2027. This massive capital deployment forms the backbone of Beijing’s strategic initiative, supplying the computational resources essential for developing cutting-edge AI capabilities.
According to Ackman, dominance in artificial intelligence—particularly at the superintelligence threshold—will ultimately depend on access to computational power, data resources, and energy supply. Should China achieve this milestone ahead of the United States, he believes it would create fundamental strategic disadvantages affecting both national defense and democratic systems.
Semiconductor Sector Faces Headwinds
Financial markets reacted swiftly following the Kimi K3 announcement. Nvidia shares declined 2% on Friday, continuing a challenging period for chip manufacturers as market participants reassess the computational requirements of the AI arms race.
The selloff evoked memories of early 2025’s DeepSeek episode, when an economical Chinese model temporarily wiped out billions in semiconductor market capitalization before confidence stabilized.
Micron crossed into bear market territory this month due to unrelated pressures involving a competing Chinese initial public offering and potential export limitations. The stock showed modest recovery Friday.
Gavin Baker from Atreides Management provided an alternative perspective. He suggested that Kimi K3 might actually help most technology companies while pressuring AI research labs that allegedly maintain inference profit margins near 90%.
Venture capitalist Chamath Palihapitiya noted that pricing for advanced AI capabilities has plummeted dramatically, creating unsustainable conditions for model developers’ profitability.
Baker challenged the narrative of cheap computing tokens, explaining that Kimi K3 produces lengthy, reasoning-intensive outputs that actually cost 50% to 70% more to operate than equivalent American systems, despite appearing less expensive on paper.
On forecasting platform Polymarket, Anthropic maintains greater than 67% probability of retaining the top AI model position through December, while Moonshot sits in the low single-digit range.





