Most companies are using generative AI in some capacity these days. But corporate leaders said it’s not actually having all that much impact, according to a new study from the National Bureau of Economic Research.
Researchers in the study asked executives at nearly 6,000 companies how generative AI has affected their productivity and employment. And 80% of the time, neither of those things seem to have changed much, said Gregory Thwaites, University of Nottingham economist and the study’s co-author.
“But the answer is very much ‘yet,’” Thwaites said. “Because they’re expecting a bigger effect over the next three years than over the past three years.”
That expectation could be because the technology is improving so quickly. Lindsey Raymond, a postdoctoral researcher in economics at Microsoft Research, said she noticed that improvement firsthand in December, just two months ago.
“I write a lot of code, do a lot of analysis,” she said. “Just the coding aid and capability, there was this dramatic shift over the holidays… At least, I feel like I’m getting more productive.”
Even if the technology fails to live up to the hype more broadly, Raymond said the mere possibility could affect company staffing decisions.
“They are probably anticipating there will be impacts of AI on productivity and thus, in anticipation, may slow down their hiring,” she said.
For most companies today, AI may be reshaping expectations faster than it’s reshaping their actual productivity.