The rise of generative artificial intelligence (AI) is a hot topic in tech circles. But latest research from the Federal Reserve Bank of St. Louis, Vanderbilt UniversityAnd Harvard Kennedy School reveals the true extent of generative AI's infiltration into on a regular basis work – and the outcomes are eye-opening. According to the paper: The technology has adopted more quickly than previous transformative technologies equivalent to the laptop computer (PC) or the Internet.
Here are five surprising findings from the study, which surveyed hundreds of staff across the U.S. to find out their use of generative AI within the workplace and at home.
1. Generative AI is already more widespread than PCs at this point
Generative AI is spreading faster than anyone could have predicted. Just two years after the discharge of ChatGPT, 39.4% of Americans ages 18-64 reported using generative AI, with 28% using it at work. By comparison, it took three years for private computers to achieve a 20% adoption rate.
“Generative AI has been adopted faster than personal computers or the Internet,” the researchers write. “This is as a consequence of the faster adoption of generative AI in the house in comparison with the laptop computer, likely as a consequence of differences in portability and price.” Easy access to tools equivalent to ChatGPT and Google Gemini has played a critical role on this faster adoption.
2. Generative AI is utilized by everyone – not only technicians
While one might expect generative AI for use primarily by software developers or data scientists, research shows that it’s widely used across industries. In fact, one in five “staff” – those in construction, installation, repair and transportation – recurrently use generative AI at work.
“Generative AI is mostly utilized in management, business and computing jobs, with a usage rate of over 40%,” the paper says. “However, one in five staff and one in five non-college-educated staff also recurrently use generative AI at work.”
This shows that AI is not any longer reserved for highly expert or specialized roles. From writing reports to generating creative ideas, generative AI is getting used in a surprising number of tasks across the occupational spectrum.
3. The use of AI reflects the trend towards increasing inequality within the workplace
Just because the laptop computer revolution led to greater inequality within the workplace as computers supplemented highly expert staff and automatic routine tasks, the introduction of generative AI could speed up this trend. The study found that younger, better-educated and higher-earning staff usually tend to use AI at work.
Most notably, staff with a bachelor's degree or higher are twice as prone to use AI as those without (40% versus 20%). The researchers warn that this might exacerbate existing inequalities within the labor market.
“The use of generative AI is more widespread amongst younger, higher educated and higher-earning staff. This is notable since the PC revolution was followed by increasing inequality within the labor market,” the authors write.
4. AI already saves time on quite a lot of tasks
When it involves concrete tasks, staff are using generative AI for greater than just coding or technical work. The commonest uses of AI within the workplace are writing, administrative tasks, and interpreting text or data. In fact, 57% of those that use AI within the workplace said they use it to assist with writing tasks, and 49% said they use it for information searches.
The researchers note that “at-work usage rates exceeded 25% for all ten tasks on our list,” underscoring how useful generative AI has turn out to be across all job roles. Whether it's summarizing reports or coming up with latest ideas, AI is already saving employees a whole lot of time.
5. AI could increase labor productivity within the US – however it continues to be too early
Perhaps essentially the most exciting finding of the study is that generative AI could significantly increase labor productivity. Based on current usage patterns, the researchers estimate that between 0.5% and three.5% of all U.S. work hours are currently supported by generative AI. They further estimate that this may lead to a rise in labor productivity of between 0.125% and 0.875%.
“If we assume that generative AI increases task productivity by 25% – the median estimate from five randomized trials – this may correspond to a rise in labor productivity of between 0.125 and 0.875 percentage points at current levels of use,” the study says.
However, the authors indicate that these estimates are speculative given the early stage of adoption of generative AI. While the technology's potential is immense, its long-term impact on the economy will depend upon how deeply it becomes embedded in on a regular basis work processes.
