Blue-Collar Workers More Open to AI in Workplace, MIT Study Says

Workers without a college degree—especially Hispanic and Black workers—are more optimistic about artificial intelligence tools and automation in the workplace than those with a diploma, according to a new Amazon.com Inc.-backed Massachusetts Institute of Technology study.

Researchers expected workers in more routine jobs, and those with less formal education, to be especially wary of automation when they set out to study employee perception of AI, according to Ben Armstrong, executive director of MIT’s Industrial Performance Center. But in both cases those hypotheses weren’t borne out in the study.

In the MIT study released Monday, 27.4% of workers without a college degree estimated that automation will be beneficial for their job security, compared to 23.7% of workers with a college degree.

But the study among US workers found that while the split based on education level was pronounced among Hispanic and Black workers, it was very small among White workers.

Researchers said one explanation could be that employees without a college degree who have seen technologies introduced to augment their work “realize that replacing workers with automation is really hard, and therefore they are not worried.” They may also associate the introduction of new technologies with a “growth outlook for a company” that makes them feel more secure, the study said.

Overall, Black and Hispanic workers in the US were “far more optimistic” about the impact of automation on their jobs than were White workers and those of other racial backgrounds, according to the study.

“This was an unexpected finding without clear explanation. The apparent optimism of African American and Hispanic workers about automation invites new hypotheses that can be tested in future research,” the report said.

Some 38% of Black workers without a college degree perceived automation as having a positive impact on job security, as compared to 34% of Black workers with a degree, according to survey data.

Among Hispanic workers, 39% without a college degree perceived automation as having a positive impact, compared to 31% with a degree.

Among White workers the gap was smaller, as 22% without a college degree perceived automation as having a positive impact, compared to 21% with a degree.

While the study is based on a survey of more than 9,000 workers across nine countries—Australia, France, Germany, Italy, Japan, Poland, Spain, the United Kingdom, and the US, conducted in January—the findings on level of education and AI were only analyzed for US workers.

The study reviewed workers’ perceived impact of automation in five categories: safety and comfort, pay, autonomy, upward mobility, and job security.

US workers were the only ones to report a net negative impact of automation in any category, citing both job security and pay.

Overall, workers in roles that they identified as requiring complex problem solving or new ideas tended to be more positive about automation.

Amazon funded the survey, which was conducted by Ipsos. MIT designed the survey and wrote the paper independently.

‘Tip of the Iceberg’

Amazon has already introduced AI extensively at its worksites. More than 750,000 mobile robots are deployed in more than 300 facilities globally, according to a spokesperson. In March, Amazon tested a humanoid robot called Digit at a warehouse in Seattle.

Tye Brady, Amazon’s chief technologist and roboticist, said the MIT survey aimed to help Amazon evaluate how to use automation to be a better employer.

“I have the goal to automate out any menial, mundane, repetitive tasks, and allow people to focus on higher order,” Brady said.

Amazon is funding a fellowship for post-doctoral students to run independent studies based off the MIT test findings about worker motivation within the retail giant’s facilities.

Brady called the survey the “tip of the iceberg.” The fellowship will allow Amazon to “dig into the details” and be more focused on its own workplace culture.

The program will begin next year and is expected to be supported for three years.

Amazon’s adoption of workplace automation has led to some scrutiny, particularly from contract drivers in its Flex delivery program who claimed they were wrongly dismissed based on an algorithm, according to a 2021 Bloomberg News report.

Amazon spokesperson August Green said in a statement the report was inaccurate. “The fact is Amazon Flex delivery partners are only offboarded when repeat or egregious issues occur,” Green said, adding that the offboarding decisions are made “by a member of our team, not an algorithm.”