Preparing for generative AI’s impact on productivity may include rethinking performance management

In the ’90s and early aughts, “computer rooms” began to appear in US households; entire rooms—former bedrooms, studies, or dens—were refitted with oversized furniture with enough nooks and crannies for every piece of hardware needed to run America Online, RollerCoaster Tycoon, and Clippy. When new tech is widely adopted, it marks drastic changes to the way we do things.

As generative AI makes its way into the world of work, it’s bound to drastically change the way companies and their employees think about everything.

“There’s no question in the minds of the HR professionals…that AI will disrupt hiring, training, performance management, but they’re not quite sure how,” said Ben Granger, chief workplace psychologist at Qualtrics.

A recent study of call center workers from the National Bureau of Economic Research found that generative AI can be especially helpful for novice and lower-skilled workers and can improve workplace productivity by as much as 35%.

That same study, however, revealed that the AI took its cues from the more productive and senior employees; their prowess and experience was instrumental to the overall productivity bounce that the generative AI was able to help kindle at the call center.

“It’s digitally capturing some of that expertise that they were previously using for that one customer, and now other customers are going to benefit from it. Even ones that never talked to that high performer,” said Erik Brynjolfsson, senior fellow at the Stanford Institute for Human-Centered AI and one of the authors of the study.

The productivity of higher performers was impacted less dramatically by the adoption of generative AI, the study found.

“Much of the previous wave of computerization disproportionately helped college-educated and more high-paid workers, and increased income inequality,” Brynjolfsson said. “So, this technology is going in the opposite direction.”

The study’s findings reveal another HR-managed system that might be impacted by the widespread adoption of generative AI: performance management.

“A good manager will realize that their top employees are actually even more valuable now than they were before,” Brynjolfsson said, because they’ll “serve as training data for a system, and that expertise, that insight, that skill will—in part—be reflected in all the other calls as well.”

“That’s great for the company,” Granger said. “But is that great for the highest performers who have seeded the AI?”

Granger suggested HR teams may need to restructure incentive and reward systems that measure both the performance of the individual contributor and their impact on the team and the growth of its lower performers.

System changes that accompany the new technologies often take years and sometimes decades. Companies are in the very early stages of thinking through how generative AI will impact their systems, Brynjolfsson said.

Brynjolfsson also pointed to another paper he co-wrote for the Brookings Institution, which suggests that the widespread adoption and productivity gains with generative AI will likely be much faster than other technological advancements because it can run on existing internet infrastructure and is easy for nearly everyone with an internet connection to use.

So, as companies think about how they’re incorporating generative AI into the business, HR leaders should already be thinking about what performance management change it might require.

“The companies that get that right are going to pull away from their competitors,” Brynjolfsson said.