A new study analyzing the Danish labor market in 2023 and 2024 means that generative AI fashions like ChatGPT have had nearly no important impression on total wages or employment but, regardless of speedy adoption in some workplaces. The findings, detailed in a working paper by economists from the College of Chicago and the College of Copenhagen, present an early, large-scale empirical take a look at AI’s transformative potential.
In “Massive Language Fashions, Small Labor Market Results,” economists Anders Humlum and Emilie Vestergaard centered particularly on the impression of AI chatbots throughout 11 occupations usually thought-about weak to automation, together with accountants, software program builders, and buyer help specialists. Their evaluation coated knowledge from 25,000 staff and seven,000 workplaces in Denmark.
Regardless of discovering widespread and sometimes employer-encouraged adoption of those instruments, the research concluded that “AI chatbots have had no important impression on earnings or recorded hours in any occupation” throughout the interval studied. The boldness intervals of their statistical evaluation dominated out common results bigger than 1 p.c.
“The adoption of those chatbots has been remarkably quick,” Humlum advised The Register concerning the research. “Most staff within the uncovered occupations have now adopted these chatbots… However then after we take a look at the financial outcomes, it actually has not moved the needle.”
AI creating extra work?
Throughout the research, the researchers investigated how firm funding in AI affected employee adoption and the way chatbots modified office processes. Whereas company funding boosted AI device adoption—saving time for 64 to 90 p.c of customers throughout studied occupations—the precise advantages have been much less substantial than anticipated.
The research revealed that AI chatbots truly created new job duties for 8.4 p.c of staff, together with some who didn’t use the instruments themselves, offsetting potential time financial savings. For instance, many academics now spend time detecting whether or not college students use ChatGPT for homework, whereas different staff evaluate AI output high quality or try to craft efficient prompts.
