President Donald Trump’s artificial intelligence (AI) czar, David Sacks, is pointing to evidence that China’s DeepSeek AI startup spent a lot more money developing its models than has been reported.
DeepSeek sent the U.S tech sector into turmoil on Monday after reporting that it had spent only $5.567 million to train its DeepSeek-V3 AI model, which is purportedly competitive with some AI models developed in the United States that cost billions.
“New report by leading semiconductor analyst Dylan Patel shows that DeepSeek spent over $1 billion on its compute cluster,” Sacks wrote on X on Friday. “The widely reported $6M number is highly misleading, as it excludes capex and R&D, and at best describes the cost of the final training run only.”
THE DEEPSEEK AI CHATBOT BURST ONTO THE SCENE: ARE FEARS ABOUT IT OVERBLOWN?
After revealing the $5.5 million figure in its report, DeepSeek had added, “Note that the aforementioned costs only include the official training of DeepSeek-V3, excluding the costs associated with prior research and ablation experiments on architecture, algorithms, or data.”
Earlier this week, tech mogul Palmer Luckey slammed the U.S. media for widely reporting the $5 million figure from DeepSeek, accusing the press of ignoring that a significant portion of the Chinese AI company’s infrastructure costs were still unknown.
US REPORTEDLY INVESTIGATING WHETHER CHINA’S DEEPSEEK USED RESTRICTED AI CHIPS
“I think the problem is they put out that number specifically to harm U.S. companies,” Luckey told FOX Business’ “That Claman Countdown.” “You had a lot of useful idiots in U.S. media kind of just mindlessly reporting that that’s the case, and neither China nor the media nor DeepSeek has any kind of incentive to correct the record as a lot of U.S. companies like Nvidia crashed to the tunes of hundreds of billions of dollars.”
DeepSeek’s model appears able to match the capability of chatbots like OpenAI’s ChatGPT and Meta’s Llama but at a fraction of the development cost. It also rose to No. 1 on the Apple App Store over the weekend and reportedly can use reduced-capability chips from Nvidia.
Those revelations slammed the U.S. tech sector on Monday.
“There’s a reason they put out the news that way, and if the stock market is any indication, it’s accomplishing exactly what they hoped to,” Luckey added. “So, look: We can recognize that Chinese AI is a real competitive threat without losing our minds over it and falling for CCP [Chinese Communist Party] propaganda.”
