Job loss estimates over the past year suggest minimal displacement of human jobs in 2023.

The economic conditions facing U.S. companies in the aftermath of interest rate hikes, record inflation, and global conflict were the most significant job displacers. After that, companies cited the closing of a business unit and bankruptcy as the next largest reason they cut head count in 2023. Only a little more than 4,000 job cuts were considered AI-inflicted, according to the Challenger, Gray & Christmas analysis of layoff data.

Investment bankers, analysts, and researchers have made educated guesses regarding the scope of labor displacement over the next decade or more—and they've varied fairly dramatically.

A report from global consulting firm McKinsey estimated that nearly half of all work performed today by humans could be automated. A research memo from investment bank Goldman Sachs projected that two-thirds of jobs today will be exposed to some form of automation and one-quarter could be wholly substituted with generative AI. An even more recent analysis from researchers at the Massachusetts Institute of Technology suggests only a quarter of jobs where vision is required will be capable of full automation.

Most experts agree some fields will be more ripe for displacement early on in the development of AI, including repetitive tasks like data entry. An extension of the automation efforts that have been ongoing in factories and warehouses for decades, the new crop of AI tools has taken aim at lower-level, white-collar work.

Buy-now-pay-later payment provider Klarna boasted in February that it was able to have AI do the work of 700 employees—roughly the size of a layoff the company undertook almost two years before the announcement. The tools allowed the company to provide customer service that was in some ways better received by Klarna users than human customer service. In the same release, the company's CEO noted "the profound impact on society that AI will have."

But when will the robots come for the remaining 25% to 50% of jobs?

Perhaps the AI-induced stock price bubble will normalize, and layoff fears will dissipate as generative AI's limitations become too great to ignore. The leading models still grapple with the tendency to invent misinformation or challenge direction from human users. Or perhaps we are only now seeing the earliest consequences of the AI transformation.

In the nascent days of automation in the automotive manufacturing sector, prolific inventor Norbert Wiener was able to join computer processes with machines and generated mass hysteria when he warned the implications of what he dubbed "cybernetics" could one day create mass unemployment. It wasn't until decades later that automotive manufacturing jobs began to be culled, as hydraulic robot arms were erected in their place.

Story editing by Shannon Luders-Manuel. Copy editing by Tim Bruns.