It would not be a stretch to describe this period of hyperactive growth in the tech industry as a historic moment.

将科技行业这段异常活跃的增长期称为历史性时刻并不为过。

Nvidia, which makes computer chips that are essential to building artificial intelligence, said on Wednesday that its quarterly profit jumped to nearly $32 billion, up 65 percent from a year ago and 245 percent from the year before that.

周三，生产构建人工智能所必需芯片的英伟达表示，其季度利润跃升至近320亿美元，同比增长65%，较两年前增长245%。

Just three weeks ago, Nvidia became the first publicly traded company to be worth $5 trillion. Microsoft, Google, Apple and Amazon are also now valued in the trillions. In their most recent quarters, the four companies reported more than $110 billion in combined profits.

就在三周前，英伟达成为首家市值突破5万亿美元的上市公司。微软、谷歌、苹果和亚马逊目前市值也都达到了数万亿美元。这四家企业最近季度录得总利润超过1100亿美元。

“There’s been a lot of talk about an A.I. bubble,” said Jensen Huang, Nvidia’s chief executive, after his company’s blowout quarterly report. “From our vantage point, we see something very different.”

“关于人工智能泡沫的讨论不绝于耳，”英伟达首席执行官黄仁勋在公司交出漂亮的季报后表示。“但从我们的视角看，情况截然不同。”

But some industry insiders say there is something ominous lurking behind all this bubbly news. They look at the same eye-popping growth and the same stunning wealth creation as Mr. Huang and see a house of cards. And they say it is hard to know what the damage will be if it collapses.

但一些业内人士表示，所有这些看似美好的消息背后都隐藏着不祥之兆。他们与黄仁勋目睹了同样的增长与财富奇迹，却认为这是脆弱的纸牌屋。他们表示，如果崩塌，由此造成的破坏将难以估量。

Even Nvidia’s growth can be explained away. Demand for the company’s chips doesn’t mean people want to use A.I. It merely means that companies are building giant A.I. systems in hopes someone will pay to use them. The Nvidia-led rally on Wall Street lasted only a few hours, after all, and the company’s share price was down about 3 percent at the end of trading on Thursday. A sharp reversal in tech stocks pulled the whole market lower, with the S&P 500 dropping 1.6 percent for the day.

就连英伟达的增长也并非无懈可击。市场对其芯片的需求并不意味着人们想使用人工智能，而是仅表明企业正在斥巨资构建人工智能系统，寄望于未来能吸引付费用户。英伟达引领的华尔街涨势仅维持数小时，截至周四收盘该公司股价下跌约3%。科技股大幅回调拖累整个市场，标普500指数当日下挫1.6%。

英伟达强劲的季度业绩曾短暂缓解华尔街对人工智能巨大开支的担忧。

The heart of the pessimist’s case against the A.I. boom, however, is the money pouring into the start-up world and the billions upon billions those companies are spending on data centers.

不过，对人工智能热潮持悲观态度者持有的核心论点是，巨额资金正涌入相关初创企业，而这些公司在数据中心上的支出惊人。

OpenAI, the company that kicked off the boom three years ago, is now worth an estimated $500 billion, making it the most valuable start-up in the world. Anthropic, OpenAI’s archrival, is now worth an estimated $183 billion. And Thinking Machines Labs, which was started in February, is already believed to be valued in the tens of billions of dollars.

三年前引爆这场热潮的OpenAI目前估值约5000亿美元，是全球最有价值的初创企业。OpenAI的主要竞争对手Anthropic目前估值约1830亿美元。今年2月才创立的Thinking Machines Labs据信估值也已达到数百亿美元。

OpenAI is not profitable and doesn’t expect to be until 2030. Anthropic is also in the red. Thinking Machines just put out its first product.

OpenAI目前尚未盈利，而且预计2030年前都难以实现收支平衡。Anthropic同样处于亏损状态。Thinking Machines则刚刚推出首款产品。

That hasn’t stopped them from spending. Anthropic recently said it would invest $50 billion in new data centers. Sam Altman, OpenAI’s celebrity chief executive, said his company was committed to spending $1.4 trillion on computing power for its A.I. pursuits.

但这并没有阻止它们烧钱。Anthropic近日宣布将投资500亿美元用于新数据中心的建设。OpenAI的明星首席执行官萨姆·奥尔特曼则表示，公司将为其人工智能事业投入1.4万亿美元用于算力。

“What OpenAI is engaged in is the most dramatic case of ‘Fake It Until You Make It’ that we have ever seen,” said Gil Luria, head of technology research at D.A. Davidson. “They are making huge commitments that they literally can’t afford.”

“OpenAI现在做的事情，是我们见过最夸张的‘一直假装下去直到成真’的案例，”D.A. Davidson的科技研究负责人吉尔·卢里亚说。“他们做出的承诺巨大，实际上根本无力承担。”

OpenAI and its partners are pumping $500 billion into new data centers in the United States as part of what they call Project Stargate. In today’s dollars, that is enough to fund the Manhattan Project 15 times over. It could pay for the entire Apollo moon project. Twice.

OpenAI与其合作伙伴正斥资5000亿美元在美国兴建新一代数据中心，若按当前币值计算，这个所谓的“星门计划”足以支持15次曼哈顿计划的全部开支，甚至能支付两次完整的阿波罗登月计划。

“Stargate alone — if it does actually reach $500 billion — would be the largest infrastructure project in the world, several times over,” said Evan Conrad, chief executive of San Francisco Compute, a start-up that specializes in hardware for A.I.

“仅星门计划本身——如果其规模真的达到5000亿美元——就将是世界上最大的基础设施项目，而且超越现有纪录数倍，”San Francisco Compute的首席执行官埃文·康拉德说道，该公司是一家专注于人工智能硬件的初创企业。

Tech companies, governments and their partners around the world will spend nearly $3 trillion on data centers by 2028, according to analysts at Morgan Stanley. To make that happen, they will borrow nearly a trillion dollars from banks and other financial institutions.

据摩根士丹利的分析师预测，到2028年，全球的科技公司、各国政府及其合作伙伴将在数据中心的投入近3万亿美元。为了实现这一目标，他们需从银行和其他金融机构借贷近1万亿美元。

Over the last 12 months, Google, Microsoft, Amazon and Meta spent about $360 billion on new data centers. With their huge profits, they can afford it. Other companies have to take on debt. That includes established companies like the software maker Oracle and smaller outfits with names like CoreWeave and Nebius.

过去12个月里，谷歌、微软、亚马逊和Meta在新数据中心上投入约为3600亿美元。凭借巨额利润，它们负担得起这种开支。其他公司则不得不举债投入。这其中有甲骨文这样的老牌企业，也有CoreWeave、Nebius等较小的公司。

Because that debt is held by a wide array of financial institutions — including private credit lenders as well as traditional banks — experts are struggling to understand how much risk is in the system.

由于这些债务分散在众多金融机构——包括私人信贷机构与传统银行——专家们难以准确评估整个系统中究竟蕴含多少风险。

Adding to worries, critics say some of the deals OpenAI has made with chipmakers, cloud computing companies and others are oddly circular. OpenAI is set to receive billions from tech companies but also sends billions back to the same companies to pay for computing power and other services.

更令人担忧的是，批评者指出OpenAI与芯片制造商、云计算公司等达成的交易存在奇特的资金循环现象。OpenAI在获得科技巨头数十亿美元注资的同时，又向同一批企业支付等量资金以获取算力等服务。

Some financial analysts worry that these deals make the market look stronger than it really is. Ultimately, the health of the market will depend on whether companies like OpenAI can turn a profit before debt overwhelms them.

一些金融分析师担心，这类交易正在制造市场繁荣的假象。最终，市场的稳定性将取决于像OpenAI这样的公司能否在被债务压垮前实现盈利。

(The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The two companies have denied the suit’s claims.)

（《纽约时报》已起诉OpenAI和微软，指控其人工智能系统侵犯新闻内容版权。OpenAI和微软均否认相关指控。）

Nvidia has also made some deals that have raised questions about whether the company is paying itself. It announced that it would invest $100 billion in OpenAI. The start-up receives that money as it buys or leases Nvidia’s chips.

英伟达的某些交易引发了同样的质疑。该公司宣布将向OpenAI投入1000亿美元，而这些资金最终会通过采购或租赁英伟达芯片的方式回流至该公司。

On Tuesday, Nvidia announced a similar $10 billion deal with Anthropic, which will buy $30 billion in A.I. computing backed by Nvidia chips. That money will actually go to buy computing power from Microsoft, which also invested $5 billion in Anthropic.

周二，英伟达宣布与Anthropic达成100亿美元的类似协议，后者将采购价值300亿美元、由英伟达芯片支持的人工智能算力。这笔资金实际将用于从微软购买算力，而微软也向Anthropic注资50亿美元。

Goldman Sachs has estimated that Nvidia will make 15 percent of its sales next year from what critics also call circular deals.

高盛估计，英伟达明年的销售额中有15%将来自批评者所称的“自我循环交易”。

A.I. companies’ chatbots and image generators are already being used by hundreds of millions of people. Many of them pay monthly fees that can top $100. But it’s not so clear that business customers — the real cash cow for the tech industry — are as keen to use A.I.

人工智能公司开发的聊天机器人与图像生成器已吸引数亿用户，其中不少人每月支付超过100美元的服务费。但作为科技行业真正利润支柱的企业客户是否同样热衷采用人工智能技术仍存疑问。

Nearly eight in 10 businesses have said they have used A.I. technologies. Just as many have said these technologies had “no significant bottom-line impact,” according to recent research from McKinsey & Company.

麦肯锡的最新研究显示，近八成企业声称已应用人工智能技术，但同等比例的企业表示这些技术对其“利润没有显著影响”。

Sundar Pichai, the chief executive of Alphabet, Google’s parent company, said in an interview with the BBC this week that the spending and skyrocketing valuations were driven at least in part by “irrationality.” If the market crashes, he said, the damage will be widespread. “I think no company is going to be immune, including us,” he said.

谷歌母公司Alphabet的首席执行官桑达尔·皮查伊本周在接受BBC采访时表示，这些支出和飙升的估值至少部分源于“非理性因素”。他警告称若市场崩盘，冲击将十分广泛。“我认为没有公司能幸免，我们也不例外，”他说。

Tech industry veterans often compare the A.I. boom to the dot-com bubble of the 1990s. When that bubble burst, hundreds of start-ups disappeared and established companies that were selling technology to those young outfits experienced huge losses. But other start-ups found lasting success and did, in fact, change the world — most notably Amazon and Google.

科技行业资深人士常将当前人工智能热潮与1990年代的互联网泡沫相提并论。当年的泡沫破裂时，数百家初创公司消失，而那些向新兴企业提供技术支持的成熟公司也遭受重创。但也有部分初创公司获得了持久成功，并深刻改变了世界——最著名的例子是亚马逊和谷歌。

“When bubbles happen, smart people get over-excited about a kernel of truth,” Mr. Altman told reporters earlier this year. “Are we in a phase where investors as a whole are over-excited about A.I.? My opinion is yes. Is A.I. the most important thing to happen in a very long time? My opinion is also yes.”

“当泡沫形成时，聪明人总会因一个具有合理价值的核心而过度兴奋，”奥尔特曼今年早些时候对记者表示。“当前投资者整体是否对人工智能过度追捧？我的看法是肯定的。人工智能是否是这个时代最具变革性的技术？我的答案同样是肯定的。”