What you need to know
- DeepSeek, a Chinese AI startup, is shaking up the tech world by outperforming Big Tech’s models with fewer resources and lower costs.
- Its latest AI launch caused global market chaos: Nasdaq dropped 3%, S&P 500 slid 1.5%, and Nvidia lost a record $593 billion in one day.
- DeepSeek claims its AI rivals top U.S. models like ChatGPT while being more cost-effective, sending shockwaves through the AI chip industry.
A Chinese AI startup called DeepSeek seems to be outperforming Big Tech’s models with way less money and resources, and it’s totally rattling the U.S. tech sector.
DeepSeek has sent shockwaves through global markets with its latest AI launch. According to NBC News, the Nasdaq Composite dropped over 3%. The Dow managed to recover, closing up nearly 300 points after a rough start, while the S&P 500 slid about 1.5%.
Nvidia, a major player in AI, took a massive hit Monday, losing a staggering $593 billion in market value as tech stocks tanked, Reuters reports. This was due to fears that DeepSeek has jumped ahead of the U.S. in the AI race. This marks the biggest one-day loss any company has ever seen on Wall Street.
DeepSeek isn’t holding back—the company is saying its AI is just as good, if not better, than the top U.S. models, all while costing way less. That bold claim has sent Nvidia and other AI chipmakers into a tailspin, with their stocks taking a serious hit.
Launched in 2023, this Hangzhou-based startup rolled out its first large language model before that year was even over.
DeepSeek started turning heads in the AI world last month after dropping a new model that it claims rivals top U.S. players like OpenAI’s ChatGPT, while being way more efficient with pricey Nvidia chips for training on massive datasets.
The startup’s open AI model, DeepSeek V3, is crushing it on popular benchmarks. Despite its size, it’s super efficient, breezing through tasks like coding and essay writing.
Obviously, what’s got some U.S. tech watchers nervous is how this Chinese startup has managed to keep pace with top American generative AI players while spending way less cash.
If that’s true, it makes you wonder if U.S. tech companies are blowing through cash for data centers and chips when there might be a cheaper, smarter way to keep AI moving forward.