Artificial intelligence might be a relatively new field – industry leader OpenAI was founded just a decade ago – but new research shows that it is already impacting the labor market.
This working paper out of Stanford University found that AI is particularly impacting younger workers in their early 20s who are entering AI-exposed fields such as software and customer service. Among early-career workers ages 22 to 25 in the most exposed occupations, researchers observed a 13% relative decline in employment.
Since AI started becoming more popular with the advent of programs like ChatGPT and AI image generators, there have also been concerns about its rapid growth and potential impact. Those range from its use in school projects to its ability to potentially escape human control or even become sentient.
Reuters/Ipsos poll results released earlier this month showed that 71% of respondents were concerned that AI would permanently put people out of work. That poll came after Andy Jassy, the CEO of online retail giant Amazon, announced that the company would likely shrink its workforce due to AI.
New controversy around AI that illustrates concerns about AI’s impact on creative jobs was sparked this month when retailer J. Crew apparently used AI images to promote its new sneaker collaboration with Vans. Images posted on its Instagram page were met with backlash and the retailer followed up by crediting the work of an AI creator, according to The Cut.
Authors of the new paper out of Stanford said they made six key findings in their analysis of “a large-scale, high-frequency administrative dataset from ADP,” the largest payroll software provider in the U.S. Here’s what they found:
1. Substantial declines in employment for early-career workers in occupations most exposed to AI
2. Growth in economy-wide employment alongside stagnant employment for young workers
3. Entry-level employment declines where AI automates work and “muted” effects where AI augments work instead
4. Declines still observed even “after conditioning on firm-time effects”
5. Labor market adjustments associated with AI seem to impact employment more than compensation
6. Patterns hold in “occupations unaffected by remote work and across various alternative sample constructions”
Notably, the researchers found that more experienced workers seemed to be somewhat protected from the AI-related shifts, as well as people in less-exposed occupations such as nursing aides. They said that the AI-related employment patterns began to appear starting in 2022.
“Why might AI adversely affect exposed entry-level workers more than other age groups? One possibility is that, by nature of the model training process, AI replaces codified knowledge, the ‘book-learning’ that forms the core of formal education,” the authors explained. “AI may be less capable of replacing tacit knowledge, the idiosyncratic tips and tricks that accumulate with experience.”
Of course, AI isn’t the first type of technology to shake up the labor market. Based on previous trends, the researchers said that new technologies tend to usher in adjustment periods for the workforce, with displaced forms of work replaced by new forms. A similar adjustment might already be underway when it comes to AI, “with emerging evidence of shifts in college majors away from AI-exposed categories such as computer science,” they said.
“Past transitions such as the IT revolution ultimately led to robust growth in employment and real wages following physical and human capital adjustments, with some workers benefiting more than others,” the paper authors added.