什么是"杰文斯悖论"，微软CEO急了！深夜发帖反驳
来源：倍可亲(backchina.com)By Anita Hamilton
作者：安妮塔·汉密尔顿
A late-night post on X from Microsoft CEO Satya Nadella on an economic concept known as Jevons Paradox sought to counter fears that the emergence of China's DeepSeek large language model and chatbot would pop the artificial-intelligence bubble.
微软首席执行官萨提亚·纳德拉在X上的深夜发帖，讨论一个名为杰文斯悖论的经济概念，试图回应人们对中国DeepSeek大型语言模型和聊天机器人的出现可能会引发人工智能泡沫破裂的担忧。
The mere fact that Nadella posted it shortly before 1 a.m. Eastern time Monday morning points to just how worried the tech industry is about DeepSeek.
纳德拉在周一早上东部时间凌晨1点前不久发帖，恰恰说明了科技行业对DeepSeek的担忧有多么严重。
Market panic set in as social media posts suggested that DeepSeek developed its model for less than $6 million, far less than the more than $1 billion it costs to train some models.
市场恐慌加剧，社交媒体发帖暗示DeepSeek以不到600万的价格开发其模型，远低于一些模型训练超过10亿的成本。
Nvidia lost nearly $600 billion in market value on Monday, the biggest one-day drop ever for a publicly traded company, as investors worried that companies are overspending as they seek to develop and train their AI models.
英伟达周一的市值损失接近6000亿，这是公开交易公司单日最大跌幅，投资者担心公司在开发和训练其人工智能模型时过度消费。
Closer reading of the technical paper where the low figure first appeared shows it doesn't include the cost of prior research. But even if DeepSeek had only cost $6 million, that might not be as big a problem for the rest of the industry as the market reaction implies.
仔细阅读技术论文，低成本首次出现的地方表明并未包括先前研究的成本。但即使DeepSeek只花费了600万，这对行业的其他部分可能也并不像市场反应所暗示的那样大问题。
Named for William Stanley Jevons, a 19th-century English economist, the Jevons Paradox posits that lowering the cost of using fuel doesn't reduce demand for it, but instead causes it to rise.
杰文斯悖论以19世纪英国经济学家威廉·斯坦利·杰文斯的名字命名，该理论认为降低燃料使用成本并不会减少对其的需求，反而会造成需求上升。
"It is wholly a confusion of ideas to suppose that the economical use of fuel is equivalent to a diminished consumption. The very contrary is the truth," he wrote in his 1865 book, The Coal Question.
他在1865年的著作《煤的问提》中写道："认为燃料的经济使用等同于减少消费是一种完全的观念混淆。事实恰恰相反。"
The idea has been applied to other commodities, from iron to electricity and energy as a whole.
这个概念已被应用于其他商品，从铁到电力以及整体能源。
Nadella believes the concept can be applied to AI models as well. As he wrote in his post, "Jevons paradox strikes again! As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can't get enough of."
纳德拉认为这个概念也可以应用于人工智能模型。正如他在发帖中所写，"杰文斯悖论再次出现！随着人工智能变得更加高效和可及，我们将看到其使用量激增，变成我们无法满足的商品。"
Microsoft was spared much of the stock market carnage on Monday, falling just 2%, versus Nvidia's 17% plunge. That could signal that Nadella was onto something when he invoked a long-dead economist to make his point.
微软在周一躲过了股市的重创，仅下跌2%，而英伟达则暴跌17%。这可能表明，当纳德拉引用一位已经去世的经济学家来支持他的观点时，他是说对了。
Write to Anita Hamilton at anita.hamilton@barrons.com
请给Anita Hamilton发邮件，邮箱是anita.hamilton@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
此内容由《巴伦周刊》制作，由道琼斯公司经营。《巴伦周刊》与道琼斯新闻社和《华尔街日报》独立发行。
