‘Not your parents’ automation’: How generative AI will impact jobs in major cities

In previous generations, automation resulted in small communities’ economies and populations being decimated as mines, steel mills and shipyards closed, along with other heavy industries. Those areas, which typically were not in the large cities, saw their blue-collar workforce hollowed out, with many still struggling to recover.

Now on the horizon is another wave of automation, this time driven by artificial intelligence and generative AI. But according to a recent research brief from the Brookings Institution, this latest wave of automation will have a heavier impact on larger cities and metropolitan areas that rely on white-collar industries.

Metro areas with the most jobs exposed to generative AI include those located in traditional tech powerhouses like Silicon Valley, Seattle, San Francisco and New York. It also includes places like Boulder, Colorado, and Salt Lake City, Utah, which have both seen an influx of tech jobs in recent years.

“It's not your parents’ or your grandparents’ automation,” said Mark Muro, a senior fellow at Brookings Metro and a co-author of the report. “Your grandparents probably were on the wrong end of certain industrial automations, and this is a very different animal.”

Researchers used data from OpenAI — the company behind ChatGPT — on specific occupations and how “exposed” they are to either being changed or eliminated by generative AI. And they found that the technology is well suited to tasks like coding, writing, financial analysis, engineering and certain aspects of the law. Brookings also warned that customer service and clerical work is most at risk of complete disruption.

Those findings are largely in keeping with the public statements from state and local leaders over the last couple of years, who see generative AI as a way to augment various aspects of their employees’ jobs, like helping them find information, draft communications or plan for various scenarios. Government leaders have largely shied away from suggesting that generative AI will lead directly to job losses, however. Muro cautioned that it is too early to determine what the extent of any losses might be.

“These places, a lot of big, fortunate coastal metropolises, could well benefit from this technology, but may also see a lot of disruption,” he said. “Both things can be true; both things can happen, including in the same place.”

It is incumbent, then, on state and local governments to prepare their residents for the onset of generative AI. Many have already established training programs for their employees, but, for the broader populace, workforce development efforts may need to be expanded as people see their jobs either disrupted or disappear. Muro noted that many of the places set to be most disrupted should have the money to cope.

“The cities presumably will have more resources to support, potentially, those who are disrupted,” he said. “But I think we're going to have to think about ways for better training and retraining, and also for some support while they transition into something. I'm not a long-term bailout person, but I do think that there will be a need to sustain workers as they make changes. We also are going to need to better fix on what are the durable skills to actually teach.”

Researchers also said they were concerned about generative AI exacerbating the divide between rural and urban America, something that Brookings has written extensively about. Jobs that are less exposed to generative AI tend to be located in less urban areas of the country and include sectors like manufacturing, construction and food preparation.

“The locus of AI involvement is much more urban than with other previous automation forms,” Muro said. “To the extent it is beneficial and a productivity driver, it will further add to the productivity and prosperity of the most productive and prosperous places in America. Meanwhile, to the extent that smaller towns or rural spaces are not involved and are insulated, they may also lose out on those benefits. It could very much exacerbate a regional divide.”

It all represents an intriguing future, as governments of all sizes reckon with the impact of generative AI and how to best prepare their residents for its growth in use. Muro said leaders should be prepared for the technology to be a “two-edged sword,” as it carries great benefits but could also result in great disruption that is as-yet unseen.

“We first should conclude that there may well be very significant change,” he said. “Some of it could be quite disruptive, and some of it could be beneficial. If you think about all these major changes, like ship building in the past, or working in an Amazon fulfillment center amidst hundreds of robots, those are the things they have in common: big change, transformations that can benefit some, but are harmful to others. And both are true.”