字号·· | 护眼
华尔街日报

人工智能正在抢走白领的工作吗?最新数据告诉了我们什么Is AI Taking White-Collar Jobs? What the Latest Numbers Tell Us

点「原文对照」整页切到原文,或双击某段只看那段的原文。

咨询公司、律师事务所和会计企业似乎是人工智能削减薪资支出的显而易见之地。然而在10月,这三个行业均增加了就业岗位。

Consulting firms, law offices and accounting businesses might seem like obvious places for artificial intelligence to shrink payrolls. In October, all three industries added jobs.

这些增幅促成了劳工部周五发布的月度就业报告好于预期。总体而言,专业和商业服务业增加了7万个就业岗位,为自2022年7月以来最大单月增幅。

Their gains contributed to Friday’s brighter-than-expected monthly jobs report from the Labor Department. All told, professional and business services added 70,000 jobs, its largest monthly increase since July 2022.

这些增幅使得人工智能将成为白领职位杀手的担忧变得更加复杂——这种担忧助推了本周选举中怀疑AI者当选。

The gains complicate fears that AI is set to become a white-collar job killer—concerns that helped sweep AI skeptics into office in this week’s elections.

管理、科学和技术咨询服务业在10月较上月增加了1.73万个就业岗位,法律服务业增加了2400个,会计及相关服务业增加了2300个。这些行业常被列为最易受AI冲击的行业之中。

Management, scientific and technical consulting services added 17,300 jobs in October over the previous month, while legal services added 2,400 and accounting and related services added 2,300. All are industries frequently ranked as among the most exposed to AI.

哈佛大学经济学家大卫·德明表示:“我看不出如何能从这份报告得出我们正步入类似白领职位末日般的结论。”

“I don’t know how you could look at this report and conclude that we’re entering anything like a white-collar job apocalypse,” said Harvard economist David Deming.

非营利宏观经济研究机构“雇佣美国”的高级经济学家普雷斯顿·缪表示,报告迄今几乎未显示高端白领职位出现大范围流失的迹象。

Preston Mui, a senior economist at Employ America, a nonprofit macroeconomic research group, said the report showed little sign so far of the sweeping losses among higher-end white-collar jobs.

缪说:“这并非许多人担心的那场知识型工作者的惨剧——至少目前还不是。”然而,这些增幅与许多工作者对就业市场的感受背道而驰——由于许多公司既不愿招聘也不愿裁员,许多人难以进入就业市场或跳槽到更好的工作。

“It’s not the bloodbath among knowledge workers that a lot of people had feared,” Mui said. “At least not yet.”

此外,并非所有白领行业都表现稳健。新发布的9月就业数据显示,文件准备服务业同比削减就业17%;电话呼叫中心和公共关系机构各流失4.7%;广告代理机构削减3.9%。

However, those gains run counter to many workers’ feelings about the job market—with many companies reluctant to either hire or lay off, many people are having a hard time breaking into the job market or moving up to a better job. What’s more, not all white-collar industries are holding up as well.

许多经济学家认为,人工智能对入门级、重复性工作构成了特定威胁——例如呼叫中心员工、数据录入员、初级文案撰稿人——这些工作很容易实现自动化。与此同时,如果技术能提高生产力,高技能白领工作可能会从人工智能中受益。例如,律师可以减少准备文件的时间,将更多精力用于为客户提供咨询,从而可能支撑对其服务的需求。

Newly released September jobs data shows document-preparation services cut employment 17% over the year; telephone call centers and public-relations agencies each lost 4.7%; and advertising agencies shed 3.9%. Many economists think AI is a particular threat to entry-level, repetitive work—call-center workers, data-entry keyers, junior copywriters—which can be easily automated. At the same time, higher-skill white-collar jobs could benefit from AI if the technology makes them more productive.

“它正在改变雇主的需求,”招聘网站Indeed的经济学家Cory Stahle表示。“他们更倾向于有经验的工人和人工智能技能。我们看到的是,雇主正在增加额外的要求。”虽然关于人工智能取代工作的传闻报告比比皆是,但经济学家一直难以找到确凿证据,证明它正在导致大规模失业。一个复杂因素是:许多科技公司在疫情结束后过度招聘,随后在ChatGPT于2022年11月首次发布前后开始裁减员工。这意味着许多被归因于人工智能的裁员,实际上可能只是管理层误判的结果。

Lawyers, for example, could spend less time preparing documents and more time advising clients, potentially supporting demand for their services. “It’s changing what employers are looking for,” said Cory Stahle, an economist at jobs site Indeed. “They’re leaning in to experienced workers and AI skills. What we’re seeing are employers adding additional requirements.”While anecdotal reports of AI supplanting jobs abound, economists have struggled to find concrete evidence that it is causing widespread job losses. One complication: Many tech companies overhired coming out of the pandemic, and then began reducing their workforces right around when ChatGPT was first released in November 2022.

Mui指出,行业总量数据也可能掩盖特定职业的变化。这些数据按雇主所属行业对工人进行分组,而非按其职业分类。例如,一家公司可能在削减行政职位的同时,为其他职位招聘人员。

That means that many layoffs that get attributed to AI could actually just be the result of management miscalculations. Industry totals can also obscure changes in particular jobs, Mui noted. The figures group workers by their employer’s industry, rather than their occupation.

“结果发现,AI 虽然应用面广,但深度不够。除非它能完成某人工作的全部内容,否则很难将其接入系统以完全取代工人,”德明说。“我们拥有的许多工作都关乎‘人’的事务——建立关系和建立信任。AI 无法为你做到这一点。”

A company could, for example, cut administrative positions while hiring for other roles. “It turns out AI is broad but not that deep. Until it can do the entirety of someone’s job, it’s hard to plug it in to totally replace a worker,” Deming said.

AI 基础设施建设为工人提供了另一个需求来源:建设数据中心的工作。雇主在非住宅建筑商和专业承包商领域合计增加了 10,800 个就业岗位——这些工人(如电工)在数据中心工地需求旺盛。(与此同时,住宅建筑一直处于低迷状态。住宅建筑领域的同类岗位减少了 16,400 个。)据商务部数据,自今年初以来,数据中心建设的月度支出增长了近 80%,而同期其他非住宅建筑支出下滑了 4%。

“A lot of the jobs we have are about people stuff—forming relationships and building trust. AI can’t do that for you.”The AI build-out offers another source of demand for workers: the job of building data centers. Employers added a combined 10,800 jobs in nonresidential builders and specialty trade contractors—workers such as electricians who are in high demand at data-center sites. (Home-building, meanwhile, has been in a slump. Counterparts in residential building shed 16,400 jobs.) Monthly spending on data-center construction increased by nearly 80% since the start of the year, according to the Commerce Department, while other nonresidential construction slid 4% over the same time.

Indeed 7 月的一项研究发现,提及数据中心的招聘广告在 2024 年 6 月至 2026 年 6 月间翻了一倍多,而整体招聘广告下降了约 12%。数据中心招聘广告中,约四分之一是面向安装和维护工人的。

A July study by Indeed found that job postings mentioning data centers more than doubled between June 2024 and June 2026, while overall postings fell about 12%. Roughly a quarter of data-center postings were for installation and maintenance workers.

最大的数据中心建设工地会雇佣数千名建筑工人,持续数年,但这些设施建成投入运营后,雇佣的人员要少得多。周二,选民支持了那些批评数据中心建设、并承诺减缓或暂停建设的民主党候选人。

The largest data-center building sites employ thousands of construction workers for years, though the facilities employ far fewer people once they are up and operating. Voters on Tuesday backed Democratic candidates who have been critical of data-center construction and pledged to slow it down or pause it.

阅读媒体原文 ↗