The CEO of Scale AI said that China has caught up with the US in DeepSeek

The U.S. may have led China in the artificial intelligence race for decades, said Alexander Wang, CEO of Scale AI, but everything changed on Christmas Day.

Wang said the company provides training data to key AI players, including OpenAI. google And MetaAt the World Economic Forum in Davos, Switzerland, DeepSeek, a leading Chinese AI lab, released a “ground-breaking model” on Christmas Day, followed by DeepSeek-R1, a strong reasoning-focused AI model, Thursday at the World Economic Forum in Davos, Switzerland. Comparable to the recently released o1 model from OpenAI.

“What we found is that DeepSeek … has high performance or is on par with the best US models,” Wang said.

In an interview with CNBC, Wang described the artificial intelligence race between the U.S. and China as an “AI war,” and said he believes China has more. Nivea H100 GPUs – the AI ​​chips widely used to build leading powerful AI models – are more powerful than people might think, especially given US export controls.

Wang also said that he believes that the AI ​​sector will reach one trillion dollars, which is equivalent to the assumption that the generative AI market is ready. Top 1 trillion dollars in income in ten years.

“The United States needs a large amount of computing power, a large amount of infrastructure,” Wang said, later adding, “We need to unleash American power to enable this AI development.”

Earlier this week, President Donald Trump announced a partnership with OpenAI, Oracle and SoftBank to invest billions of dollars in US AI infrastructure. Project Stargate was announced at the White House by Trump, SoftBank CEO Masayoshi Son, Oracle co-founder Larry Ellison and OpenAI CEO Sam Altman. Key early technology partners include Microsoft, Nvidia and Oracle, as well as semiconductor company ARM. We will invest $100 billion to begin with and up to $500 billion in the next four years, he said.

In an interview Thursday, Wang said he believes it will take two to four years to reach Artificial General Intelligence, or AGI, a widely-mentioned but clearly-observed branch of AI technology used in the AI ​​sector, or equivalent. It surpasses human intelligence in various tasks. AGI is a hotly debated topic, with some leaders saying we’re almost there and some saying it’s never possible. Wang’s own definition of AGI is “powerful AI systems that can use a computer just like you or I can…and basically become a remote worker in a great way.”

Anthroponic, the AmazonThe AI ​​startup, founded by former OpenAI research executives, ramped up its technology development last year and said in October that its AI agents could use computers as much as humans can to complete complex tasks. Anthropics’ computing capabilities allow the technology to interpret what’s on a computer screen, select keys, enter text, navigate web pages and perform tasks with any software and real-time Internet browsing, the startup said.

The device “can basically use computers the way we do,” Jared Kaplan, executive director of Anthropic Science, told CNBC in an interview at the time. He said that he works “in tens or even hundreds of steps”.

Open AI It is reported that the plan Soon to introduce the same feature.

Asked which US artificial intelligence startups are leading the AI ​​race, Wang said the models each have their own strengths — for example, OpenAI models are good at reasoning, while Anthropics is good at coding.

“The space is becoming competitive rather than competitive,” he said.

Correction: This article has been updated to correct the name of DeepSeek’s reasoning-oriented AI model, DeepSeek-R1.