How China’s New AI Model DeepSeek Threatens US Dominance

A little-known AI lab from China has caused panic throughout Silicon Valley after releasing AI models that can outperform America’s best despite being built cheaper and with less powerful chips.

DeepSeek, as the lab is called, revealed at the end of December a free, open-source large-language model that it says took just two months and less than $6 million to build, using reduced-capacity chips from Nvidia called H800s.

The new development has raised alarms about whether America’s global lead in artificial intelligence is shrinking, and has called into question big tech’s massive spending to build AI models and data centers.

In a set of third-party benchmark tests, DeepSeek’s model fared better Meta‘s Llama 3.1, OpenAI’s GPT-4o and Anthropic’s Claude Sonnet 3.5 in accuracy ranging from complex problem solving to math and coding.

DeepSeek on Monday released r1, a reasoning model that also outcompeted OpenAI’s latest o1 in many of these third-party tests.

“Seeing the new DeepSeek model, it’s super impressive in terms of both how they’ve really effectively made an open source model that does this inference time calculation and is super computationally efficient,” Microsoft CEO Satya Nadella told World Economic Forum in Davos, Switzerland, Wednesday. “We should take developments out of China very, very seriously.”

DeepSeek also had to navigate the strict semiconductor restrictions imposed by the US government on China, cutting it off from access to the most powerful chips, such as Nvidia’s H100s. Recent progress suggests that DeepSeek either found a way around the rules or that the export controls were not the chokehold Washington intended.

“They can take a really good, big model and use a process called distillation,” said Benchmark General Partner Chetan Puttagunta. “Basically, you’re using a very large model to help your small model learn what you want it to learn. It’s actually very cost-effective.”

Little is known about the laboratory and its founder, Liang WenFeng. DeepSeek was born out of a Chinese hedge fund called High-Flyer Quant, which manages about $8 billion in assets, according to media reports.

But DeepSeek isn’t the only Chinese company making inroads.

Leading AI researcher Kai-Fu Lee has said his startup 01.ai was trained with only $3 million. TikTok parent company ByteDance on Wednesday released an update to its model that claims to outperform OpenAI’s o1 in a key benchmark test.

“Necessity is the mother of invention,” said Perplexity CEO Aravind Srinivas. “Because they had to figure out solutions, they actually ended up building something much more efficient.”

Watch this video to learn more.