The Next Couple Of Years For The Job Market Could Be Tough

Author: Sam Ro | October 27, 2025 01:04pm

History shows that the world is pretty good at adapting to disruptive technology.

We have repeatedly seen the same pattern: While many jobs may get destroyed in the technological transition, many new ones are also created. A study recently cited by Goldman Sachs found that 60% of workers are employed in jobs that weren't recognized by the Census before 1940.

Wells Fargo's Ohsung Kwon compiled a brief list of some of these newer jobs.

Of course, this process of creative destruction doesn't happen overnight.

AI's impact on jobs

As AI technology sweeps across industries, everyone's asking what it means for the labor market in the coming years.

In the three years since the launch of ChatGPT and AI quickly becoming ubiquitous, the labor market hasn't shifted all that much. As of August, employment was at a record high, and layoff activity remained depressed.

But Goldman Sachs economists warn we may experience some "transitional friction." And they caution that we may not see major job losses until the economy enters a full-blown recession.

"A leading explanation for this phenomenon is that companies use recessions to restructure and streamline their workforce by laying off workers in less productive areas," Goldman Sachs' Pierfrancesco Mei and David Mericle wrote in an Oct. 13 research note. "This is especially true when recessions follow productivity booms that give companies some pent-up ability to cut labor costs and improve efficiency without significantly hurting their productive capacity."

Unfortunately, the subsequent rebound in jobs following such cuts wouldn't necessarily begin right away. From the economists:

One notable example of this was the so-called "jobless recovery" after the 2001 recession, which followed the technology-led productivity boom of the late 1990s. As Exhibit 8 shows, total employment took a long time to recover as companies continued to shed routine jobs for several quarters after the end of the recession. During that recovery, despite a soft labor market, productivity growth remained elevated and GDP growth rebounded earlier than employment growth.

What makes AI particularly concerning for the economy is that it threatens to replace workers, which could make this transition particularly challenging.

"The type of technology is important: employment has tended to grow more quickly in occupations where technological progress has been labor-augmenting, but more slowly where it has been labor-substituting," the economists wrote.

We're having this discussion as key labor turnover metrics, including job openings and the hiring rate, have fallen sharply over the past three years. This suggests the economy is at a tipping point, which means large-scale job losses could be next.

The big picture

I'm personally cognizant of how AI threatens to disrupt industry. I work in media. I regularly hear about how AI is creating headaches for the news business, ranging from declines in search traffic to newsroom cuts to various ethical quandaries.

In recent weeks, TKer's analytics show that a small but growing number of people are coming to the website via ChatGPT referrals, which suggests there are even more people reading summaries of my writing without me ever knowing about it. That stresses me out, and I expect that trend to get worse.

But that's life these days. As soon as you settle into things, the world changes. And you have to adapt.

While many companies may struggle or even fail during this transition, many will evolve and adapt, perhaps in surprising ways. This has always been the case. Accordingly, as an investor with a broadly diversified portfolio of stocks, I remain optimistic in the long run. I'm bullish on the promise of AI, especially in the context of how productivity gains help earnings and stock prices head higher. Even if the disruption means my job as I know it is in peril.

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.