Unlock Editor’s Roundup for free
Roula Khalaf, editor of the FT, picks her favorite stories in this weekly newsletter.
It is difficult to talk about the economic history of the 21st century without discussing “comrade China”. That’s the term often used to describe China’s entry into the global market, a shift that brought rich countries an abundance of cheap goods but left entire industries and workforces stranded.
DeepSeek can provide a follow-up. A little-known Chinese hedge fund has thrown a grenade into the world of artificial intelligence with a large language model that, in fact, matches the market leader, Sam Altman’s OpenAI, at a fraction of the cost. And while OpenAI treats the functions of its models as proprietary, DeepSeek’s R1 keeps its technical internals on the outside, making it attractive for developers to use and build upon.
Some content could not be loaded. Check your internet connection or browser settings.
Things move faster in the age of AI; terribly so. Five of the biggest AI-ready tech stocks — chipmaker Nvidia and so-called hyperscalers Alphabet, Amazon, Microsoft and Meta Platforms — combined lost almost $750 billion in market value before U.S. markets opened on Monday. It could be particularly bleak for Nvidia if it turns out that DeepSeek won without using its flashier chips.
Investors in technology companies – including Europeans such as chipmaker ASML, and energy companies that investors had hoped would get a boost from the data center boost – have been left wondering if their investments will go up in smoke. Hyperscalers were due to make nearly $300 billion in capital spending this year, according to Visible Alpha estimates. Analysts expect Meta and Microsoft to report investments for 2024 totaling $94 billion when they report earnings on Wednesday.
Indeed, the game is not over. DeepSeek’s current potential is still unclear, and it has yet to reach “artificial general intelligence,” the human condition that Meta and OpenAI are pursuing. But the rules may have changed. At the very least, DeepSeek could get some of the American giants’ customers. At worst, it has challenged the core belief that more hardware is the key to better AI. This principle has supported the market value of Silicon Valley companies as they invest hand in hand.
What’s bad for hyperscalers can still be a windfall for everyone else. For most business users, having the absolute best model is less important than having a reliable and good enough model. Not every driver needs a Ferrari. Advances in reasoning like R1 can be a big step for “agents” who deal with customers and perform tasks in the workplace. If they are offered more cheaply, corporate profitability should increase.
In this sense, this second strike by China may resemble the first. It can bring not only destruction, but a realignment—albeit a painful one for many. Researchers have estimated that for every job lost in the China shock, the purchasing power of American households increased by more than $400,000. The race for AI supremacy is on pause; the great giveaway has begun.
john.foley@ft.com
