The Blubber

China’s DeepSeek Disrupts U.S. AI Dominance, Says Scale AI CEO

For the past decade, the U.S. has led China in the artificial intelligence race—but according to Scale AI CEO Alexandr Wang, everything changed on Christmas Day.

Speaking at the World Economic Forum in Davos, Switzerland, Wang said that DeepSeek, China’s leading AI lab, released an "earth-shattering model" on December 25, followed by DeepSeek-R1, a powerful reasoning-focused AI model that directly competes with OpenAI’s o1 model.

DeepSeek: A Game Changer in AI?

"What we’ve found is that DeepSeek is either the top-performing model or at least on par with the best American models," Wang said in an interview with CNBC.

Wang described the U.S.-China AI race as an ‘AI war’, noting that despite U.S. export controls, China may possess far more Nvidia H100 GPUs—the industry’s most sought-after AI chips—than many believe.

Wang also predicted that the AI industry will become a trillion-dollar market, aligning with broader forecasts that generative AI could generate over $1 trillion in revenue within a decade.

U.S. Ramps Up AI Investment

Earlier this week, President Donald Trump announced a joint venture between OpenAI, Oracle, and SoftBank to invest billions in U.S. AI infrastructure.

• The project, called Stargate, was unveiled at the White House alongside SoftBank CEO Masayoshi Son, Oracle co-founder Larry Ellison, and OpenAI CEO Sam Altman.
• Key partners include Microsoft, Nvidia, Oracle, and semiconductor giant Arm.
• The initiative begins with a $100 billion investment, with up to $500 billion planned over the next four years.

"The United States needs a massive amount of computational capacity and infrastructure," Wang emphasized. "We need to unleash U.S. energy to enable this AI boom."

Are We Close to AGI?

In the interview, Wang estimated that artificial general intelligence (AGI)—AI that matches or surpasses human intelligence—could be achieved within the next two to four years.

Wang defined AGI as "powerful AI systems that can use a computer just like a human… essentially acting as a remote worker at the highest level."

The debate over AGI remains divisive, with some experts believing it’s imminent, while others argue it may never be fully realized.

Meanwhile, Amazon-backed AI startup Anthropic has accelerated progress in AI capabilities. In October, the company announced its AI agents could interact with computers like humans—navigating screens, selecting buttons, entering text, and even browsing the internet in real time.

The AI Race Is Just Getting Started

With China’s DeepSeek making rapid advan