AIWas Supposed to Come for Our Jobs. So Far, Evidence Is Mixed.
AI搶工作？證據仍不明
Is artificial intelligence taking our jobs?
人工智慧正在取代我們的工作嗎？
It’s arguably the single biggest question facing the labor market right now, all the more so after Tuesday’s midterm elections showed widespread public animosity toward AI and data centers, in particular.
這大概是當前勞動市場最令人關切的問題；尤其期中選舉顯示，民眾普遍對AI、特別是資料中心感到反感，這個問題更顯迫切。
The evidence, though, is far from clear.
不過，目前掌握的證據，仍遠不足以下結論。
White-collar industries like accounting, legal services and finance mostly added jobs in October, but their growth has been decidedly weak over the past few years. The information sector — a broad category that includes much of tech but also media companies — has been cutting jobs steadily since late 2022. Microsoft last week announced that it was offering buyouts to tens of thousands of employees, making it the latest tech giant to cut jobs.
會計、法律服務和金融等白領產業，10月就業人數大多不減反增，但過去幾年成長明顯疲弱。資訊業則涵蓋大部分科技業和媒體公司，自2022年底以來持續裁員。微軟上周宣布，將向數萬名員工提供優離方案，成為最新一家裁員的科技巨頭。
Economists are cautious about attributing that softness to AI, however. Other forces, such as high interest rates and shifting demand patterns following the pandemic, could also be playing a role. And despite announcements like Microsoft’s, layoffs remain low, even in white-collar sectors.
不過，經濟學家並不急著把這種疲弱歸咎於AI。高利率、疫情後需求型態改變等其他因素，也可能造成影響。而且，即使有微軟這類宣布，整體裁員仍維持低檔，白領產業也不例外。
Similarly, some economists have highlighted relatively high rates of unemployment among younger workers, especially young college graduates, as evidence that AI is displacing entry-level jobs. But that, too, could be the result of other factors, including the generally low rate of hiring and turnover, which is particularly hard on people trying to break into the job market.
同樣，有些經濟學家指出，年輕勞工、尤其是剛畢業的大學生失業率相對偏高，可作為AI正取代入門職缺的證據。但這也可能是其他因素造成的，包括徵才與人員流動率偏低；這對剛要進入就業市場的人尤其不利。
To the extent there is a consensus among economists, it looks something like this: AI hasn’t had a striking effect on employment, but it may be starting to show up in some corners of the labor market.
就經濟學家之間已有的共識而言，大致是這樣：AI尚未對整體就業造成明顯衝擊，但可能已開始影響勞動市場的某些特定領域。
Even if AI is taking a toll on hiring, it isn’t quite clear what that means. Are companies using AI to replace workers or merely pressing “pause” on hiring while they figure out where these new tools will be most useful? Or are they spending so much on tokens and AI infrastructure that they are looking to cut costs elsewhere? Those explanations could all look similar in the moment but have very different implications over the longer run.
即使AI正在衝擊徵才，背後的意義也還不清楚。企業究竟是利用AI取代員工，還是只是在釐清這些新工具最適合用在哪裡時，暫緩徵才？又或者，他們在AI詞元和AI基礎建設上花費太多，因此想在其他地方削減成本？這些解釋當下看來可能很相似，但長期而言，意涵卻大不相同。
The impact of AI is showing up clearly in one place, however: blue-collar jobs. The data center build-out is helping to prop up employment in the construction sector, even as high interest rates have led to a sharp pullback in homebuilding. Demand for chips and other AI equipment is also creating jobs in manufacturing, offsetting losses elsewhere in the sector.
不過，AI的影響確實清楚反映在藍領工作上。即使高利率已使住宅興建大幅降溫，資料中心建設仍在支撐營建業的就業。晶片及其他AI設備的需求，也在製造業創造就業，抵銷該產業其他領域的職缺流失。
The future of that boom now looks uncertain following this week’s election results, however. The plunge in share prices for tech companies Thursday — partly reversed Friday — reflected investors’ concern that public backlash to data centers could lead to a slowdown in construction, which could have broader economic implications given the extent to which AI has been driving growth.
期中選舉結果出爐後這波榮景能否延續，如今看來並不確定。科技股股價周四重挫，周五已部分反彈，反映投資人擔心，民眾對資料中心的反彈可能導致營建降溫；鑑於AI對經濟成長的推動程度，這可能帶來更廣泛的影響。
文／Ben Casselman，譯／國際中心
▪高為元調查案激烈討論近5小時 清大校務會議表決流會…校方發聲了▪桃園武和宮前學生施暴影片上網 單親媽：妹妹勸阻不了還被恐嚇▪清大校長挾優勢讓校務會議流會？學生會發聲明說明、抗議：絕未被掌握▪陸生借位拍照稱「小小賴清德拿下」遭驅逐出境 台大發聲了▪教授遭控刁難學業碩班生不幸亡 台北科大認定霸凌不成立