Summarize this content to 2000 words in 6 paragraphs in Arabic Stay informed with free updatesSimply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.A frantic sell-off of energy and infrastructure shares this week after advances in artificial intelligence by Chinese start-up DeepSeek shows how little is understood about the power demands of AI, the International Energy Agency has warned.\u00a0Investors heavily marked down shares in companies ranging from European gas turbine makers to US power generators and infrastructure providers as DeepSeek\u2019s surprise breakthrough suggested the development and use of AI may require less power than previously thought.Companies hit included Siemens Energy, GE Vernova, Mitsubishi Heavy Industries, Constellation Energy, Vistra Energy, Schneider Electric, ABB and Legrand.Schneider, which overtook the valuation of TotalEnergies last September as investors rushed to buy companies that would benefit from the seemingly insatiable power demands of AI, has now fallen back behind the French oil major in market capitalisation. \u201cThis abrupt reaction highlights that the market currently does not yet have adequate tools and information to assess the outlook for AI-driven electricity demand,\u201d said Thomas Spencer, one of the analysts leading energy and AI work at the IEA. He added that more volatility in the energy sector was an \u201cinevitable risk\u201d given the current opacity over the operational and energy performance of data centres. DeepSeek\u2019s model appears to have been trained using much less electricity than its western peers, requiring less than 10 per cent of the computational resources of Meta\u2019s Llama model to develop, according to Nature.\u00a0\u201cThe electricity consumption of data centres is proportional to their computational power,\u201d said analysts at Rystad, an energy consultancy. \u201cIf what they claim is true, total energy consumption could be significantly lower.\u201dAnalysts at Citi said that if the numbers proved to be accurate, it would have a significant impact on companies providing energy infrastructure. Companies that supply data centres have benefited from robust growth expectations. \u201cMore computationally efficient AI could bring these trends into question,\u201d they wrote. The sell-off caused alarm across the energy sector, with one senior executive at Siemens Energy exclaiming that the company\u2019s 21 per cent fall in its share price on Monday was simply \u201cnot reasonable\u201d. The share price of the German company, a four-year-old spin-off of the former gas and power division of Siemens, had risen 336 per cent in the past year, before Monday\u2019s crash. It has a bulging order book for its gas turbines from utilities building gas-fired power plants to service AI data centres. After the market closed on Monday, it said it had made \u20ac9bn of revenues in the first quarter, well ahead of expectations, and was growing organically by 18 per cent year on year. The company said the sell-off was overdone, since data centres will only account for a small slice of its future growth. The chief executive, chief financial officer and head of investor relations at one electricity infrastructure company held a call on Monday to try to understand the impact of the DeepSeek news. They concluded that the situation would stabilise \u201clittle by little\u201d, if financial results were positive. Others insisted that while DeepSeek\u2019s model might be more efficient, the breakthrough would speed up the adoption of AI, creating more power demand overall. This effect is known as the Jevons paradox after the English economist William Jevons, who observed in 1865 that technological breakthroughs that led to more efficient use of coal increased the consumption of coal.\u00a0\u201cJevons paradox strikes again!\u201d wrote Satya Nadella, the chief executive of Microsoft on LinkedIn. \u201cAs AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can\u2019t get enough of.\u201dRystad noted that the wider energy transition could be affected by a less power-intensive AI, since some of the largest buyers of zero-carbon electricity had been Big Tech companies. Amazon, for example, is the largest corporate buyer of renewable energy in the world, having signed more than 500 power purchase agreements across 27 countries, for a total of 77 terawatt hours of electricity each year, equivalent to the entire electricity demand of Belgium.\u00a0With less renewable energy going to AI, it might become available for other sectors \u201chelping displace faster the use of fossil fuels\u201d, said Rystad.\u00a0Video: AI power demand could stifle industry’s growth | FT Energy Source <\/p>\n","protected":false},"excerpt":{"rendered":"
Summarize this content to 2000 words in 6 paragraphs in Arabic Stay informed with free updatesSimply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.A frantic sell-off of energy and infrastructure shares this week after advances in artificial intelligence by Chinese start-up DeepSeek shows how little is understood about the<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[63],"tags":[],"class_list":{"0":"post-185025","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-tech"},"_links":{"self":[{"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/posts\/185025","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/comments?post=185025"}],"version-history":[{"count":0,"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/posts\/185025\/revisions"}],"wp:attachment":[{"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/media?parent=185025"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/categories?post=185025"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/globetimeline.com\/ar\/wp-json\/wp\/v2\/tags?post=185025"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}