- DeepSeek’s rise sent US tech stocks tumbling on Monday morning, especially Nvidia.
- Intel’s former CEO said DeepSeek would expand the AI market instead of diminishing it.
- Meta promised a new “leading state of the art” AI model and pledged more investment.
Tech leaders and their companies have reacted with admiration and insights after AI company DeepSeek launched its flagship large language model, R1.
Just days after DeepSeek launched, the app dethroned ChatGPT with the most downloads on Apple’s Top Free Apps chart, rivaling systems by OpenAI, Google, and Meta despite being developed at a reported fraction of their costs.
The rise of the Chinese AI startup founded by quant hedge fund manager Liang Wenfeng was followed by a sharp sell-off of major AI and chip companies in the US tech markets on Monday.
Nvidia, a leader in AI hardware, saw its stock plunge by over 17% amid concern about DeepSeek’s ability to achieve competitive results with less advanced and significantly cheaper hardware.
Shares of other tech giants, including Microsoft and Alphabet, also declined Monday morning.
Here’s how Silicon Valley leaders have responded to DeepSeek so far.
Satya Nadella
Nadella, Microsoft’s CEO, posted on LinkedIn on Monday that “Jevons paradox is at play again,” referencing the concept that greater efficiency in production often fuels higher demand. “As AI becomes more efficient and accessible, its adoption will soar, transforming it into an indispensable commodity,” he added.
Earlier last week at the World Economic Forum in Davos, Nadella also said that other tech companies “should take the developments out of China very, very seriously.”
Marc Andreessen
Andreessen, cofounder of Andreessen Horowitz, praised DeepSeek’s R1 model and called it “one of the most amazing and impressive breakthroughs” and “a profound gift to the world” in an X post on Friday. On Sunday, the Silicon Valley venture capitalist — who has been advising President Trump on tech policy — went on to call Deepseek R1 “AI’s Sputnik moment.”
Deepseek R1 is AI’s Sputnik moment.
— Marc Andreessen 🇺🇸 (@pmarca) January 26, 2025
Pat Gelsinger
Gelsinger, the former CEO of Intel, challenged the market’s reaction to DeepSeek’s advancements, particularly the sell-off of AI chip stocks. He said the market is “getting it wrong” and suggested that the company’s “dramatically cheaper” AI models could expand the market for AI applications rather than diminish it.
Gelsinger also credited DeepSeek’s Chinese engineers, who “had limited resources, and they had to find creative solutions.”
Wisdom is learning the lessons we thought we already knew. DeepSeek reminds us of three important learnings from computing history:
1) Computing obeys the gas law. Making it dramatically cheaper will expand the market for it. The markets are getting it wrong, this will make AI…— Pat Gelsinger (@PGelsinger) January 27, 2025
Yann LeCun
LeCun, chief AI scientist for Meta’s Fundamental AI Research division, challenged the perception that China is surpassing the US in AI in a LinkedIn post, arguing that the correct reading is that “open source models are surpassing proprietary ones.”
He commented that DeepSeek “came up with new ideas and built them on top of other people’s work.”
Mark Zuckerberg
Though Zuckerberg did not directly respond to DeepSeek’s rise, the Meta CEO posted on Facebook on Friday promising that a new version of Facebook’s open-source AI model family Llama would become “the leading state of the art model” upon release.
Llama is an AI model designed for natural language processing tasks like text generation, translation, and summarization, which is promoted as open-source like DeepSeek.
Pledging more than 1.3 million GPUs of computing power by the end of the year, he wrote that Meta is “planning to invest $60-65B in capex this year while also growing our AI teams significantly” and that the company has additional capital to continue investing over the next few years.
Meta did not immediately respond to a request for comment.
Nvidia
In a statement, a spokesperson for Nvidia told Business Insider that DeepSeek is an “excellent AI advancement and a perfect example of Test Time Scaling,” illustrating how to leverage “widely available models and compute that is fully export control compliant.” The spokesperson added that to make inference work, it “requires significant numbers of NVIDIA GPUs and high-performance networking.”
Jensen Huang, Nvidia’s CEO, has not directly responded to DeepSeek thus far.
