DeepSeek rattles Wall Street. What the Chinese startup means for Nvidia, Meta and the rest of our AI stocks

How real is DeepSeek? That’s the No. 1 question investors are trying to answer Monday, as the emergence of the Chinese AI startup and its R1 model sparked a steep sell-off in many artificial intelligence-linked stocks. DeepSeek positions its model as competing with the best that America has to offer, but claims it used less-advanced Nvidia chips and spent much less on development than its U.S. tech peers did on similar models. It’s a crucial question, especially for investors in the tech and AI space. If it’s all nonsense, then Monday’s sell-off needs to be bought hard. If it’s not, then we need to change how we think about the AI trade, which has been a major tailwind for the U.S. stock market since OpenAI’s ChatGPT launched in late 2022. The reason is that if DeepSeek actually is far more efficient at using compute power — meaning it can run a cutting-edge large language model at a fraction of the cost of those underpinning ChatGPT — then one has to ask: What the heck are the American tech giants spending billions of dollars on? Are U.S. tech executives nothing more than the Man Behind the Curtain from