People are most worried about AI replacing these 2 jobs, according to the Max Planck Institute

You’ve probably heard AI is coming for many of our jobs. But how would you feel about getting a medical diagnosis from an AI doctor? Would you trust a verdict delivered by an AI judge?

A new study of 10,000 people in 20 countries, including the United States, India, Saudi Arabia, Japan, and China, found when it comes to artificial intelligence replacing human jobs, people are most concerned about AI replacing doctors and judges, and least concerned about AI replacing journalists.

The findings, published in American Psychologist by the Max Planck Institute for Human Development, focused on the study participants’ attitudes to AI taking over six occupations: doctors, judges, managers, caregivers, religious leaders, and journalists.

Researchers looked at eight psychological traits—warmth, sincerity, tolerance, fairness, competence, determination, intelligence, and imagination—and assessed AI’s potential to replicate these traits. The study’s findings suggest that when AI is introduced into a new job, people instinctively compare the human traits necessary for that job with AI’s ability to imitate them. The level of fear that study participants felt about AI taking certain jobs appeared to be directly linked to a “perceived mismatch between these human traits and AI’s capabilities.”

For example, the prospect of AI-driven doctors and care workers elicited strong fears in some countries due to concerns about AI’s lack of empathy and emotional understanding.

But when researchers looked at widespread concerns about AI replacing human workers, they found people’s attitudes also varied widely among nations.

For example, people in the U.S., India, and Saudi Arabia reported being most afraid of AI’s role in jobs, particularly of judges and doctors, reflecting concerns about fairness, transparency, and moral judgment. (AI-driven journalists were the least feared, likely because people feel that they retain autonomy over how they engage with the information provided by journalists.)

However, people in China, Japan, and Turkey were least afraid of artificial intelligence overall. And other studies have found that people in China place less importance on controlling AI and more on connecting with AI compared to European Americans. They’ve also found that 47% of North Americans are worried about harmful AI, while only 25% of Southeast Asians and 11% of East Asians have similar feelings.

That is due, at least in part, to different countries having different traditions of depicting AI as benevolent or malicious, as well as different historical interactions with intelligent machines. It’s also affected by people in countries having been exposed to different governmental policies about AI.