The Great Shrinking of corporate America

Post Reply New Topic

In June, Amazon CEO Andy Jassy had a blunt message for his 350,000 corporate employees: There were going to be fewer of them in the near future, thanks to the "efficiency gains" he expected from AI. The proclamation generated big headlines and an uproar from staff. But it struck me as merely honest. He was acknowledging something that pretty much every CEO who sits atop a large white-collar workforce is quietly hoping to achieve sooner or later.

After all, Jassy hasn't been the only executive to hint at a future of lower headcount. The head of JPMorgan's consumer and community business predicted in May that AI will reduce the number of employees in its operations division by 10%. That same month, the CEO of Klarna said that the company's investments in AI has already driven the company's headcount to shrink by 40%. And the CEO of Ford — a company that employs tens of thousands of white-collar professionals — declared that AI will wipe out "literally half" of all white-collar jobs. Meanwhile, Kian Katanforoosh, the CEO and founder of the software startup Workera, tells me that he never wants to have much more than the 80 or so employees he has today, no matter how successful his business ends up becoming. "I truly believe we can go super super far without growing more," he says. "I'm an engineer. I don't want to have to manage so many people if I don't need to."

It's not like CEOs ever enjoyed shelling out for the salaries or navigating the personnel headaches that come with the sprawling bureaucracies they employ. But for more than a century, armies of office workers were a necessary cost of doing business. To grow from tiny upstarts into titans of industry, companies needed an ever-multiplying number of HR reps, accountants, marketers, engineers, analysts, and project managers.

In recent months, that 100-year-trend is starting to come undone. Everywhere you look, AI appears to be helping leaner teams take on work that used to require more people. And executives are talking about their large workforces — once their greatest competitive advantage — as if they're an unfortunate holdover from a bygone, bloated era.

If today's corporate giants shrink their ranks, and if tomorrow's giants never need to bulk up in the first place, we may well be witnessing the end of a defining feature of corporate America: the mega-employer. That could give rise to a whole new generation of nimble companies that innovate faster — but also leave workers navigating a world of diminished career paths and fewer jobs.

Before the Industrial Revolution, most Americans worked for themselves as farmers or craftsmen. And those who didn't worked for very small operations — say, a few journeymen training under a master shoemaker. The resulting economy was a patchwork of all these tiny businesses.

That started to change with the advent of capital-intensive industries like textile manufacturing, which required organizing larger groups of people under a single employer. Then came railroads in the late 19th century. With projects that took many years to realize and stretched over thousands of miles, vast numbers of workers needed to be on the same page. "If you mess it up, there's a big explosion," says Louis Hyman, a political economy historian at Johns Hopkins. "You needed to really coordinate your mechanisms and make sure that people are doing things exactly the same way."

As mass production developed, Hyman says, many of the most consequential innovations during this time weren't so much technical breakthroughs: They were social inventions to coordinate the labor of all the people it took to get the most out of the new machines. The assembly line broke down complicated work into simple, repeatable, standardized tasks; scientific management emphasized the importance of monitoring, measuring, and optimizing everyone's performance; and the M-form corporate structure created a blueprint to manage sprawling bureaucracies through a clear chain of command. In the 1930s, about a tenth of the labor force worked for businesses that employed at least 10,000 people. By the end of World War II, that share had surged to about a third.

By the 1970s, some of that bigger-is-better ideology started to change. A new management philosophy set in, normalizing layoffs that took aim at bloat. And as robots automated many blue-collar jobs, IBM mainframes and word processors eliminated a whole set of white-collar clerical roles as well. Still, there was plenty of work that technology couldn't automate, which meant that companies needed large teams of college-educated professionals to keep them going. Even the most tech-forward companies saw their people — especially their coders — as mission-critical to their success. "Hiring great people — especially engineers — is one of the biggest challenges that any technology company has," Mark Zuckerberg lamented in 2013. "Our country doesn't produce the volume of engineers that the companies would want to hire." Tech giants often hired more than they needed to make sure they had a steady supply of talent, and to attract and retain the best of the best, they treated their employees like gods.

If you were to pinpoint one moment the gods turned mortal, it would probably be November 9, 2022 — the day Meta laid off more than 11,000 employees. From there, virtually every tech company followed suit, with employers across other industries close behind. At first, the cuts were chalked up to overhiring in the pandemic. But two and a half years later, the layoffs haven't stopped and hiring is still down. More and more, AI appears to be driving those austerity measures. In an industry that once hoarded talent like gold, the shift is striking. CEOs no longer seem to view the bulk of their workforce as indispensable, and they say as much: A common refrain among tech leaders from Mark Zuckerberg to Elon Musk to Dara Khosrowshahi now is some version of "If you don't like it here, you should leave." Companies like Microsoft, Meta, Google, and Salesforce had reliably increased their headcounts year after year. Now, according to the workforce analytics provider Live Data, all of them employ fewer people than they did at their 2022 peak.

J. Scott Hamilton, Live Data's CEO, says this is probably just the beginning. To gauge how much deeper the cuts could go, his team recently analyzed the detailed responsibilities of most roles at Microsoft to estimate the share of tasks that could, in theory, be done by AI. Their conclusion: If Microsoft were to offload all of those automatable tasks to AI, it would eliminate 36% of the work currently done by employees. That would mean the company could lay off some 80,000 employees.

On the one hand, that's an aggressive scenario: Companies are rarely able to overhaul their workflows to take full advantage of a new automation technology's capabilities. If they do, that transition takes a very long time. And besides, some work is simply too high-stakes to entrust to error-prone AI — even if it's technically possible. On the other, the estimate may prove conservative: Live Data's predictions assume that AI will remain at 2025-level capabilities. Given how much better the leading large language models have become over the last two years, the best tools will almost certainly be able to handle more than what they can today....