热门趋势
1
“末日博士”努里埃尔·鲁比尼表示，随着人工智能彻底改变工作方式，我们将走向全民基本收入或“某种形式的社会主义”——他称这已是乐观前景
2
电力公司正利用征用权为数据中心强征土地，而70%的美国人表示“别建在我家后院”
3
红龙虾餐厅37岁的CEO在进入高管层前曾当过服务员——他说这是管理“难缠的人和棘手情况”的速成课
山姆·奥尔特曼和达里奥·阿莫迪都在收回关于人工智能导致就业末日预言，因为他们正瞄准大规模IPO
萨莎·罗格尔伯格
萨莎·罗格尔伯格
记者
萨莎·罗格尔伯格
萨莎·罗格尔伯格
记者
2026年5月26日，美国东部时间下午4:34
OpenAI首席执行官山姆·奥尔特曼表示，他此前关于人工智能对白领工作影响的预测是错误的。
丹尼尔·霍伊尔/彭博社—盖蒂图片社
科技界最具影响力的两位CEO在过去一年里一直警告说，人工智能将摧毁白领就业。现在他们承认自己错了，加入了高盛CEO大卫·所罗门等其他领导人的行列，对人工智能导致就业末日的说法表示质疑。
OpenAI首席执行官山姆·奥尔特曼周二在接受澳大利亚联邦银行CEO马特·科明采访时表示，他关于人工智能经济影响的判断“相当错误”——这与他在2025年6月发出的入门级岗位面临严重风险的警告截然相反。Anthropic首席执行官达里奥·阿莫迪曾声称人工智能可能消除50%的白领岗位，如今他表示自动化实际上可能会扩大人们所做的工作。与此同时，所罗门至少从2025年底开始就一直认为这种恐慌被夸大了——现在他引用美国一个世纪的经济史来证明自己是对的。
“我很高兴自己在这件事上错了，”奥尔特曼告诉科明。“我原以为到现在为止，入门级白领岗位被淘汰的影响会比实际发生的更大。”
奥尔特曼补充说，他因为自己的炒作言论受到了很多批评，但宁可谨慎一些。“人们会说，‘哦，你本可以避免给世界带来那么多恐慌和悲观情绪’，但在当时，我觉得‘我看到这是一个真实的风险，我们或许应该谈谈它。’而且这个风险仍然可能存在。”
据报道，OpenAI 和 Anthropic 均计划于今年各自进行首次公开募股（IPO），两家公司的估值预计均将达到 1 万亿美元。
两次反转与一次证实
对于 OpenAI 首席执行官而言，他的言论收回了此前关于人工智能对劳动力影响的预言。一年前，奥特曼在 Uncapped 播客节目中对其兄弟杰克表示：“很多工作岗位将会消失……我们一直非常擅长找到新的事情去做……我不相信这种情况会有尽头。”
如今他表示，自己曾担忧的岗位流失并未成为现实，而一次个人实验更强化了这一看法。他曾尝试将 Slack 和电子邮件回复工作交给 AI，但后来又重新开始手动回复。
“我们确实很看重人与人之间的互动，”他说，“这件事……让我无法想象自己会在短期内把这项工作外包给 AI。这确实让我更新了想法，认为就业前景很可能与我们之前所想的大不相同。”
阿莫迪的转变同样显著。虽然他此前曾声称 AI 可能消灭 50% 的白领岗位，但本月早些时候，他将自动化重新定义为岗位的倍增器而非毁灭者：“如果你自动化了 90% 的工作，那么每个人就做剩下的 10% 的工作，”他说道，这一预测与经济学家亚历克斯·伊马斯和泰勒·考恩的观点类似。“而这 10% 的工作会扩展成人们所做的 100%，并使他们的生产力提高大约 10 倍。”
与此同时，所罗门并不需要改变立场，因为他从未持有过那种末日论调。在近期《纽约时报》的一篇专栏文章中，他提出了至少自 2025 年底以来就一直坚持的观点：美国历史为 AI 引发的就业恐慌提供了明确的驳斥，他将 1900 年代的电气化、1990 年代的数字革命与今天直接联系起来——“美国在应对变革、创造新就业方面有着悠久的历史……我看不出有任何理由认为这种动力会在现在停止。”
尽管存在行业性转变，所罗门指出，自1962年以来，美国民用就业人口增长了145%。他援引高盛的研究称，仅数据中心建设一项，自2022年以来就已新增20万个就业岗位。诺贝尔奖得主达龙·阿西莫格鲁2018年的一项研究支持了他的观点，该研究发现，AI带来的岗位替代效应通常会被生产率提升所驱动的劳动力需求所抵消。
“尽管有了Excel、电子邮件或Zoom带来的便利，我们中有谁觉得自己现在要做的事情变少了吗？”所罗门说道。
数据所揭示的与数据所未能揭示的
数据呈现出一幅喜忧参半的图景。截至2026年5月，科技行业裁员人数已超过11.5万人，已接近2025年全年录得的12.4万人，其中Meta、亚马逊和Snap等公司都将AI列为裁员的原因之一。然而，耶鲁大学预算实验室发现，自2022年底ChatGPT推出以来，在AI高暴露度岗位中，职业构成或失业持续时间并未发生显著变化。
多年来，科技领袖们一直在发布自己对未来工作的预测，从微软AI首席执行官穆斯塔法·苏莱曼认为AI能在18个月内自动化大部分白领工作，到英伟达首席执行官黄仁勋相信AI不会对就业岗位数量产生影响，反而会创造效率提升的机会，使那些积极拥抱这项技术的员工受益。
商界领袖和经济学家们开始就AI为何确实能促进劳动力市场达成共识。在回应所罗门专栏文章的一篇领英帖子中，Box首席执行官亚伦·列维表示，他打赌所罗门的观点将被证明是正确的。“如果你看看几十年前的工作是什么样的，再看看今天一切生产得多快、多容易——甚至在AI出现之前——你肯定会相信工作岗位会不复存在。然而，情况恰恰相反。为什么？”他给出的答案是，自动化不会减少对某一特定岗位的需求，反而会增加这种需求，因为自动化将提供“相同的价值主张，但成本更低”。
这本质上就是杰文斯悖论的理论，Anthropic 的 Amodei 以及 Apollo 的 Torsten Slok 等经济学家也曾引用该理论来解释未来的劳动力格局。该悖论以英国经济学家威廉·斯坦利·杰文斯的名字命名，指的是瓦特蒸汽机发明后的那段时期：更高效的燃煤技术非但没有导致煤炭消耗量减少，反而使这种商品变得更便宜、更普及。Slok 指出，像呼叫中心员工和放射科医生这类岗位，虽然其工作内容易受自动化影响，但在 AI 被广泛采用的情况下，从业人数却保持稳定甚至有所增长。
Slok 在最近的一篇博文中写道：“每次交互成本的降低并不意味着交互次数的减少。它意味着可以服务更多客户、开辟更多渠道、触及更多有价值的市场。原本被认为会缩减行业规模的技术，反而在推动其扩张。”
人工智能领域最新动态
Trendingnow
1
'Dr. Doom' Nouriel Roubini says we're headed for universal basic income or 'some form of socialism' as AI revolutionizes work—He calls that optimistic
2
Power companies are using eminent domain to seize land for data centers as 70% of Americans say not in my backyard
3
Red Lobster’s 37-year-old CEO waited tables before joining the C-suite—he says it was a crash course in managing ‘difficult people and situations’
Sam Altman and Dario Amodei are both walking back their AI jobs apocalypse prophecies as they eye blockbuster IPOs
Sasha Rogelberg
Sasha Rogelberg
Reporter
Sasha Rogelberg
Sasha Rogelberg
Reporter
May 26, 2026, 4:34 PM ET
OpenAI CEO Sam Altman said he was wrong about his prediction on AI's impact on white-collar jobs.
Daniel Heuer/Bloomberg—Getty Images
Two of the most influential CEOs in tech spent the last year warning that AI would gut white-collar employment. Now they’re admitting they were wrong, joining other leaders like Goldman Sachs CEO David Solomon in casting doubt on an AI job apocalypse.
OpenAI CEO Sam Altman, in an interview with Commonwealth Bank of Australia CEO Matt Comyn on Tuesday, said he was “pretty wrong” about AI’s economic impact—a reversal from his June 2025 warnings that entry-level roles were at serious risk. Anthropic CEO Dario Amodei, who once claimed AI could eliminate 50% of white-collar jobs, now says automation may actually expand the work people do. Solomon, meanwhile, has argued consistently since at least late 2025 that the panic was overblown—and is now pointing to a century of American economic history to say he was right.
“I’m delighted to be wrong about this,” Altman told Comyn. “I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened.”
Altman added that he’s taken a lot of flack for his hype, but better safe than sorry.”People are like, ‘Oh you could have saved the world a lot of fear mongering and a lot of doom and gloom’ but at the time I was like, ‘I see this is a real risk we should probably talk about it.’ and it still may.”
Both OpenAI and Anthropic are reportedly preparing to launch their respective IPOs this year, each company with an estimated valuation of $1 trillion.
Two reversals and a vindication
For the OpenAI CEO, his comments walk back his prophecy on AI’s impact on labor. A year ago, Altman told his brother Jack on the Uncapped podcast: “A lot of jobs will go away…we have always been really good at figuring out new things to do…I’m not a believer that that ever runs out.”
Now he says the displacement he feared simply hasn’t materialized, and a personal experiment reinforced it. He tried delegating his Slack and email responses to AI, then began responding to come again manually.
“We really do care about our interactions with people,” he said. “This thing…is not something that I can imagine myself outsourcing to an AI anytime soon. It really updated me to thinking that the jobs picture is likely to be very different than we thought.”
Amodei’s evolution has been similarly dramatic. While he previously claimed AI could wipe out 50% of white-collar jobs, he reframed automation earlier this month not as a destroyer of jobs but a multiplier of output: “If you automate 90% of the job, then everyone does the 10% of the job,” he said, offering a prediction similar to those made by economists Alex Imas and Tyler Cowen. “And the 10% kind of expands to be 100% of what people do and kind of 10-times their productivity.”
Solomon, meanwhile, didn’t need to change his position because he never held the apocalyptic one. In a recent New York Times op-ed, he offered the same argument he has made since at least late 2025: that American history offers a clear rebuttal to AI job panic, drawing a straight line from the electrification of the 1900s to the digital revolution of the 1990s to today: “The United States has a long track record of creating new jobs in response to disruption … I don’t see any reason to think this dynamic will stop now.”
Despite sectoral shifts, Solomon noted, civilian U.S. employment has grown 145% since 1962. He cited Goldman Sachs research showing data center construction alone has added 200,000 jobs since 2022. A 2018 study by Nobel laureate Daron Acemoglu backs his claim, finding that AI’s displacement effect is typically offset by productivity-driven demand for labor.
“Do any of us feel like we have less to do these days despite the convenience of Excel, email or Zoom?” Solomon said.
What the data shows and what it doesn’t
The data offers a mixed picture. Tech layoffs through May 2026 have passed 115,000, already approaching the 124,000 logged in all of 2025, with Meta, Amazon, and Snap among those citing AI as a driver of cuts. Yet the Yale Budget Lab has found no significant changes in occupational mix or unemployment duration in high-AI-exposure jobs since ChatGPT launched in late 2022.
Tech leaders have been issuing their own predictions on the future of work for years, ranging from AI being able to automate most white-collar work within 18 months, according to Microsoft AI CEO Mustafa Suleyman to Nvidia CEO Jensen Huang’s belief that AI will not have an impact on the number of jobs, but will instead create opportunities for efficiency that will benefit employees leaning into the technology.
Business leaders and economists have started to come to a consensus on why AI could indeed be a boost for labor. In a LinkedIn post in response to Solomon’s op-ed, Box CEO Aaron Levie said he’s betting that Solomon will be proved correct. “If you looked at what work looked like a few decades ago and saw how much faster everything is or easier it is to produce today — even before AI — you’d certainly have been convinced there’d be no jobs left. Yet the opposite has happened. Why?” The answer, he offered, is that automation will not decrease demand for a certain role, but rather increase it, as automation will deliver “the same value proposition, but cheaper.”
It’s essentially the theory of Jevons paradox that Anthropic’s Amodei and economists like Apollo’s Torsten Slok have also called up in explaining the future of labor. Named for English economist William Stanley Jevons, the paradox refers to the period following the invention of the Watt steam engine, when instead of more efficient coal burning resulting in less coal being burned, the commodity instead became cheaper and more popular. Slok has noted professions like call center employees and radiologists, both with roles vulnerable to automation, have remained steady or increased despite wider AI adoption.
“Lower cost per interaction does not mean fewer interactions,” Slok said in a recent blog post. “It means more customers served, more channels opened and more markets worth reaching. The technology that was supposed to shrink the industry is fueling its expansion.”
Latest in AI