Reliance on artificial intelligence reduces the capabilities of doctors and software engineers.

computer scientists
artificial intelligence

Reliance on artificial intelligence reduces the capabilities of doctors and software engineers.

As professionals increasingly rely on artificial intelligence tools in their work, could their hard-earned skills fade away?
This possibility is a growing concern for medical professionals, computer scientists, and other workers.

For example, 70% of nurses and 77% of doctors expressed concern about losing their skills due to over-reliance on artificial intelligence systems, according to a survey of healthcare workers in the United States published earlier this month.

According to a report published on the Nature website, their fear may be justified.
Evidence suggests that AI-induced "skills decline" is beginning to occur in medicine, computer science, and other fields.
Researchers are now discussing how to preserve important human expertise in the age of artificial intelligence.

"Just being aware of this phenomenon is likely to prompt people to think carefully about which skills they want to retain and which they are willing to delegate to AI tools," says Kevin Crushton, an information scientist at Syracuse University in New York.

Corrupting capabilities

A study conducted on doctors in Poland specializing in endoscopy – the use of flexible probes to examine the inside of the human body – has shown how quickly artificial intelligence tools can impact human capabilities.

The doctors, each of whom had performed at least 2,000 colonoscopies during their careers, were given access to an artificial intelligence system that analyzes endoscopic images in real time and identifies a type of precancerous intestinal lesion called a benign tumor.

The tool was available to specialist doctors on some days but not others.
Once doctors started using it, their performance declined significantly whenever the system was unavailable.

In the three months prior to the introduction of the artificial intelligence tool, specialists detected at least one adenoma in 28.4% of colonoscopies.
In the three months following the introduction of the tool, the rate of adenoma detection in colonoscopies performed without the assistance of artificial intelligence decreased to 22.4%.

Highly skilled people

The findings, published in The Lancet Gastroenterology and Hepatology journal, suggest that even highly skilled professionals may see their performance in the tasks required by their jobs decline as they become more reliant on artificial intelligence tools, says Robert Wachter, MD, of the University of California.

The study's researchers say that continued exposure to these tools may make doctors "less motivated, focused, and responsible when making cognitive decisions without the help of artificial intelligence."

Yoichi Mori, a research physician at the University of Oslo, says more studies are needed to confirm this phenomenon.
He adds that users of AI tools must realize they risk losing some of their skills. There is currently no proven solution to address skills decline.
This topic is expected to be a major focus of scientific research over the next decade.

Computer Science

To investigate whether skills are being lost in the field of computer science, researchers at Anthropic AI in San Francisco designed a randomized controlled experiment in which 52 software engineers were asked to perform a basic programming task.
During the exercise, all participants were given access to online searches and instructions on how to perform the task.

Half of the participants were also asked to use an AI assistant.
After that, all the software engineers were asked to complete a short quiz about what they had learned from the task.

Participants who used the AI ​​assistant performed significantly worse on the test compared to those who did not: the AI ​​group averaged 50% compared to 67% in the other group.

Participants who used the AI ​​assistant performed particularly poorly on questions that required them to diagnose errors in the code, indicating their failure to grasp the concepts behind the code they wrote.

Croston, a researcher investigating how the use of generative AI tools is changing the way software developers learn and retain programming skills, says the findings are worrying, especially for students and young professionals in the field.
He adds: "There is now a strange gap between performance and learning. People can perform at a very high level because they borrow skills from artificial intelligence, but they do not develop those skills themselves."

External sources of perception

Tapani Renta-Kahela, an information systems researcher at the Hanken School of Economics in Helsinki, points out that other technologies have rendered certain skills obsolete in the past.

For example, GPS navigation systems have led to a decline in people's navigation skills.
However, generative artificial intelligence tools are "the first technology to automate various cognitive abilities related to thinking and interpretation, which were long considered uniquely human skills."

Necessary preventive measures

To prevent the erosion of skills by artificial intelligence, Renta-Kahela says people need to be aware of the scale of the tasks they delegate to generative AI tools.


Post a Comment

0 Comments