BREAKINGVIEWS-China advance creates more AI winners than losers

Reuters
2025/01/27
BREAKINGVIEWS-China advance creates more AI winners than losers

The author is a Reuters Breakingviews columnist. The opinions expressed are her own. Updates to add graphic.

By Karen Kwok

LONDON, Jan 27 (Reuters Breakingviews) - Investors are rerunning their artificial intelligence financial models. Increasing signs that Chinese company DeepSeek’s new AI offering is as good as American peers like OpenAI’s, but at a cheaper development cost, caused related stocks like $3.5 trillion Nvidia NVDA.O to slump over 10% on Monday. An overdue shake-up could prove painful, but it could still create more winners than losers.

According to Artificial Analysis data, one of DeepSeek's models is 13 times cheaper to run. While that only includes computer and power expenses rather than research and data costs, it’s been enough to freak out the market. Those that make data centre chips like Nvidia and Broadcom fell 11% and 14% respectively in pre-market trading on Monday. Shares of ASML, which makes machines that design chips, fell as much as 12%. Siemens Energy, which supplies power for data centres that train models, slumped nearly 20%.

In one sense, the panic among those who invest in AI’s “picks and shovels” could wind up being overblown: DeepSeek could conceivably spur a surge of demand for cheaper AI. But it’s hardly surprising that investors are looking at companies that have traded on epic multiples and wondering whether certain big western AI players will now instead rein in their capex spending, hitting demand for chips and power.

That capex spree has been considerable. Meta Platforms last week substantially increased its AI investment spending to $65 billion in 2025, while Microsoft anticipates $80 billion this year in building data centres to power AI. U.S. President Donald Trump also announced a $500 billion Stargate AI data centre project only last week, with OpenAI's Sam Altman and SoftBank's Masayoshi Son.

The partial solace for the likes of Microsoft boss Satya Nadella is that plummeting costs in AI make it more likely he can make a return on all this investment. But that’s immediately outweighed by a bigger negative. If financial heft is no longer such an advantage, hyperscalers will not have to be so hyper-large.

That factor also makes Chinese advances in AI a good thing for investors and companies in general. If cheaper models lead to AI becoming more efficient, its use will spread more widely and quickly. That may see some of the sector’s current titans dethroned – for the benefit of just about everyone else.

Follow @karenkkwok on X

CONTEXT NEWS

Investors hammered technology stocks on Jan. 27, sending the likes of Nvidia and Oracle plummeting, as the emergence of a low-cost Chinese artificial intelligence model cast doubts on Western companies' dominance in this sector.

Startup DeepSeek last week launched a free assistant it says uses less data at a fraction of the cost of incumbent players' models.

Futures on the Nasdaq-100 slid almost 4%. Those on the S&P 500 Index dropped 2%. Shares in AI chip darling Nvidia fell 10% and Oracle dropped 8% in pre-market trading.

Dutch computer chip equipment maker ASML and its peer ASM International slumped between 9% and 10%. Germany's Siemens Energy plummeted around 20% while ABB was down 6%. Schneider Electric, a major supplier of electrical equipment and other infrastructure behind data centres, was down around 8%.

European chip and energy stocks sell off after China AI advances https://reut.rs/4jw7hEC

(Editing by George Hay and Streisand Neto)

((For previous columns by the author, Reuters customers can click on KWOK/karen.kwok@thomsonreuters.com))

应版权方要求,你需要登录查看该内容

免责声明:投资有风险,本文并非投资建议,以上内容不应被视为任何金融产品的购买或出售要约、建议或邀请,作者或其他用户的任何相关讨论、评论或帖子也不应被视为此类内容。本文仅供一般参考,不考虑您的个人投资目标、财务状况或需求。TTM对信息的准确性和完整性不承担任何责任或保证,投资者应自行研究并在投资前寻求专业建议。

热议股票

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10