AI doesn't have to mean layoffs

LE VAUDREUIL, France - For many CEOs, success in adopting artificial intelligence is measured by the number of jobs they can eliminate. In just the past few weeks, companies have announced tens of thousands of layoffs they attributed to AI, a wave that one global bank boss undiplomatically described as replacing “lower-value human capital” with technology.

But such views reflect “a very narrow understanding” of AI’s potential, said Erik Brynjolfsson, who directs the Digital Economy Lab at Stanford University. “A lot of people are under the mistaken idea that the only way that you get productivity from AI is by removing labor costs.”

Brynjolfsson is one of a group of economists who argue that businesses can reap bigger gains by using AI to make workers more productive rather than replace them.

It’s a message that Schneider Electric, a global energy technology company based in France, has taken to heart. Schneider, which has a workforce of nearly 160,000 worldwide, is embracing AI across the company.

It started by identifying “where our people are either losing time doing repetitive tasks, doing tedious tasks, doing things which fundamentally are not the right ones to do,” said Philippe Rambach, the company’s chief AI officer.

In other words, the work that gets in the way of work.

Using AI in call centers

Call centers were an obvious case for Schneider. Anyone who has languished in the labyrinth of automated phone assistance may groan at the prospect of additional technological tinkering. But Rambach said the goal was to use the technology to get answers to customers faster.

Before the company started using AI, customer service agents received thousands of questions from callers and engaged in a grand hunt through millions of pages of information to track down the answer, Rambach said. “Guess what?” he said. “Our customers were not super happy with the quality of the answer and not super happy with the speed of the answer.”

Now AI does the hunting and details how the information was selected and the source. The agent then reviews and, if necessary, modifies and refines the answer with the caller.

Other companies have registered productivity gains for their workers. A study conducted by Brynjolfsson with two other researchers, Lindsey Raymond and Danielle Li, on more than 5,000 customer support agents at a Fortune 500 firm found that AI assistance allowed workers, on average, to resolve 15% more problems, with less experienced and lower-skilled workers improving most in terms of speed and quality.

At the same time, they found that callers were more polite and less likely to utter the phrase every customer service agent has come to dread: “I want to speak to a manager.”

AI on the factory floor

At an upgraded factory in Le Vaudreuil, about 60 miles north of Paris in Normandy, Schneider is using AI to manage complex industrial processes at a decades-old site that had already been upgraded with robotic and digital tools -- some with a French accent.

The automated guided vehicles, or AGVs, that glide around the factory floor delivering parts, for instance, are named Émile and Victor after great French writers like Zola and Hugo.

AI is not needed everywhere, said Virginie Rigaudeau, a project leader at Schneider. “We use AI only when we know it delivers added value.”

Like in the production of the 74 million silver tips that the factory produces each year to manufacture electrical contactors -- the switches that are used for turning electrical circuits on and off in elevators, motors, electric vehicles, heating systems, lighting banks and more.

The recipe for cooking up the tips includes silver nitrate and sodium. The mixture is whirled in a centrifuge, and the resulting silver paste is then repeatedly washed in large steel-gray tanks to rinse out excess sodium.

But knowing how many washing cycles were enough was always something of a guessing game, Rigaudeau said.

With AI, operators can see a visual representation and learn the precise amount of sodium that remains after each rinsing cycle.

“Now the system tells us when to stop washing, and we immediately know whether the powder meets quality standards,” an operator said.

The savings have been enormous, Rigaudeau said. In one year, the company reduced waste from the process by 73%, and water use has been cut drastically.

Samples from every batch no longer need to be sent to an off-site lab for testing, a process that could take 24 to 48 hours. That has saved thousands of euros in lab costs, while decreasing gasoline consumption -- from trucks that used to ferry samples to and from the lab -- by 22%.

Cameras enhanced with AI are also used to assess the quality of finished contactors within seconds.

Making employees better, not redundant

In some European countries, using AI to make workers more productive -- and not replace them -- is encouraged by strict labor laws that can make it difficult and expensive to lay off employees.

In the United States, Brynolfsson said, government policies often encourage companies to invest in capital and cut workers. He pointed to the tax code.

“If you’re starting a new venture, and you have a lot of labor, you’re going to have to pay more taxes,” Brynolfsson said. “If you only invest in capital, then you pay less taxes.”

Of course, forecasts about the impact of AI on the job market encompass a multiverse of possibilities. And while many economists agree that policymakers and businesses have choices about how AI is deployed, some wonder if those options are narrowing.

It’s “very unpredictable,” said Anton Korinek, who helped lead the Economics of Transformative AI project at the University of Virginia. AI “will create and destroy jobs, and it’s not clear which will be more predominant,” he said.

(STORY CAN END HERE. OPTIONAL MATERIAL FOLLOWS.)

Korinek said he started studying how to develop AI as a tool for helping the labor force more than 15 years ago, but the spectacular advances have made him more doubtful about society’s ability to steer how it is developed and used. “You can’t pick the direction in which it’s going as easily anymore,” he said.

At some point, he said, AI will be “so much more productive and cheaper than humans.” (Korinek, who recently joined the economics research team at Anthropic Institute, the research arm of the AI company, made his comments before taking his new position.)

The built-in dilemma is evident even at Schneider, a company that has found ways to use AI to complement employees’ work.

Sandra Ferraguti, the plant general manager in Le Vaudreuil, showcased a new plug-and-play contactor that Schneider’s AI-assisted workforce developed and that no longer requires an electrician to do the wiring.

“Now a robot can install it,” Ferraguti said. “You don’t need a human.”

This article originally appeared in The New York Times.

Copyright 2026 The New York Times Company