U.S. workers report saving significant time on the job by using generative artificial intelligence, according to a joint study by the St. Louis Fed, Vanderbilt University, and Harvard University. The paper found: “Workers were 33% more productive per hour on average using generative AI.”
The researchers measured the impact of generative AI on workplace productivity by the extent and intensity with which workers use the technology, based on what they believe is the first nationally representative survey of generative AI adoption in the United States. They found that users saved a lot of time.
Among respondents who used generative AI in the last week, 21% said it saved them four hours or more this week, 20% said it saved three hours, 26% said it saved two hours, and 33% said it saved one hour or less.
More frequent users reported greater time savings, suggesting that applying this technology requires a learning period. Among employees who used generative AI every day in the previous week, 34% said it saved them four hours or more of time, compared with 12% of those who used it just one day that week.
Alexander Bick of the St. Louis Fed, Adam Blandin of Vanderbilt University, and David Deming of Harvard University found that if a person used AI to work 40 hours a week, they would typically save 2.2 hours. When non-AI users were added to the mix, the total time saved across all survey participants due to generated AI was 1.4% of total time.
Research shows that saving time is highly relevant in certain occupations. Information service workers spend the most time working using generative AI (14%) and save the most time (2.6%). The report shows that leisure, accommodation and other services use generative AI for the lowest proportion of time worked (2.3%) and the lowest time saved (0.6%).
With the U.S. population aging and concerns over the federal budget deficit growing, productivity growth is seen as key to driving real wage growth, corporate profits and government tax revenue. However, the report said widespread use of AI is a recent phenomenon and its ultimate impact on overall productivity improvements remains uncertain.