To best measure the impact of generative AI on productivity, organizations need to assess how intensely workers rely on the technology, researchers Alexander Bick, senior economic advisor at the Federal Reserve Bank of St. Louis, and professors of economics Adam Blandin of Vanderbuilt University and David Deming at Harvard Kennedy School, explained in a Feb. 27 post.
According to surveys taken in August and November 2024, 9% of U.S. workers aged 18-64 reported using generative AI every work day in the previous week, while 14% said they used generative AI at least one but not every work day in the previous week, the researchers said.
Additionally, workers who reported using generative AI in the previous week said it assisted them during 6% to nearly one-fourth of their work hours.
The researchers’ conclusion: On average, generative AI “is not just an occasional tool for its users, but also an integral part of their work routines.”
Key for employers, the data implied that workers are on average 33% more productive in each hour they use generative AI, the report noted.
Their study showed that generative AI also saved workers a meaningful amount of time: 33.5% of those who used generative AI every day in the previous week said it saved them four hours or more, compared with 11.5% of those who used it only one day in the previous week.
More broadly, generative AI has caught on much faster than the personal computer and the internet, they noted.
For business leaders waiting to see productivity gains they expected from generative AI, the problem may be that they’re treating it as just another technology and missing its true potential, Steven Kirz, an operational excellence senior partner at West Monroe, wrote in a January op-ed for HR Dive.
Instead of approaching AI as a technology to replace human talent, employers should recognize it as a talent type — one that improves talent efficiency at the task level, making other talent types (i.e., employees, contract labor and outsourced labor) more productive, Kirz said.
Another way to look at it is to think of AI as an intern for every employee, he suggested. The AI “intern” performs tasks the employee could have performed but that would have consumed valuable time, such as research or summarizing meeting notes and outcomes.
Similarly in an October 2024 report, professional services firm Accenture recommended that organizations that see productivity remain flat after adopting AI redefine productivity beyond traditional cost measures. That is, organizations can take the lead from productivity leaders and invest in strategies that empower, not replace, the workforce, Accenture said.
This means focusing on innovation and value creation, developing strategies that identify and tackle each challenge, and viewing challenges as intertwined components of a broader system, the firm explained.
The Federal Reserve Bank of St. Louis study found that across industries, information services has both the largest share of work hours spent using generative AI (14%) and the highest time savings (2.6%).
Time savings also vary with usage across occupations, the research found. For instance, workers in computers and mathematics used generative AI nearly 12% of their work hours and reported this saved them 2.5% of work time, according to the study.
By contrast, workers in the personal services industry used generative AI in only 1.3% of their work hours, saving them only 0.4% of work time, the findings showed.
“As more companies formally integrate AI into workflows, we may see these gains materialize more clearly in aggregate productivity measures,” the researchers wrote.
However, “the extent to which AI reshapes the labor market — whether through displacement, wage effects or skill development — remains an open and crucial question,” they said.
