U.S.-China AI Race Intensifies as DeepSeek Challenges OpenAI

DAVOS, Switzerland — The U.S.-China artificial intelligence race took a dramatic turn on Christmas Day 2024, according to Alexandr Wang, CEO of Scale AI, who revealed that China’s DeepSeek lab released a groundbreaking AI model, DeepSeek-R1, rivaling OpenAI‘s latest advancements.

Speaking at the World Economic Forum in Davos on Thursday, Wang described DeepSeek’s model as “earth-shattering,” noting its performance is on par with top U.S. models. “What we’ve found is that DeepSeek … is the top performing, or roughly on par with the best American models,” Wang said in an interview with CNBC.

Wang, whose company provides training data to major AI players like OpenAI, emphasized the intensifying competition between the U.S. and China. He characterized the rivalry as an “AI war,” highlighting China’s significant advancements despite U.S. export controls on critical AI hardware, such as H100 GPUs. “China has significantly more H100 GPUs than people may think,” Wang added.

The U.S. is responding with massive investments in AI infrastructure. Earlier this week, President Trump announced the Stargate project, a collaboration with SoftBank, Oracle, and OpenAI, aiming to invest up to $500 billion over the next four years. The initiative, unveiled at the White House, includes key partners like Oracle and semiconductor companies, with an initial $100 billion commitment.

Wang also weighed in on the timeline for achieving artificial general intelligence (AGI), a concept referring to AI systems that match or surpass human intelligence across diverse tasks. He predicted AGI could be realized within two to four years, defining it as “powerful AI systems that are able to use a computer just like you or I could … and basically be a remote worker in the most capable way.”

Meanwhile, Anthropic, an AI startup backed by ex-OpenAI executives, has made strides in developing AI agents capable of performing complex computer tasks. Jared Kaplan, Anthropic’s chief science officer, told CNBC in October that their technology can “use computers in basically the same way that we do,” handling tasks with “tens or even hundreds of steps.” OpenAI is also reportedly working on a similar feature.

Wang noted that the AI landscape is becoming increasingly competitive, with U.S. startups like OpenAI and Anthropic excelling in different areas. “OpenAI’s models are great at reasoning, while Anthropic’s are great at coding,” he said.

As the AI sector is projected to reach a trillion dollars in revenue within a decade, Wang stressed the need for the U.S. to ramp up its computational capacity. “The United States is going to need a huge amount of computational capacity, a huge amount of infrastructure,” he said. “We need to unleash U.S. energy to enable this AI boom.”