Most Americans Are Uncomfortable With Artificial Intelligence In Health Care, Survey Finds

CN-

According to a new survey, most Americans are "quite concerned" about doctors using artificial intelligence to manage their health, but they are generally aware of the potential for artificial intelligence to reduce medical errors and eliminate some of the problems that doctors may face. associate with racial wrongs. . , have superstitions. .

Artificial intelligence is the theory and development of computer programs capable of solving problems and performing tasks that normally require human intelligence. machines that can learn basically like humans from the information they are given.

You probably use AI-based technology every day without even realizing it.

For example, when you shop on Amazon, AI guides the site to recommend cat toys if you've previously purchased cat food. AI can help you unlock an iPhone, drive a Tesla, answer customer service questions at your bank, and recommend the next Netflix show.

Americans may love these personalized services, but when it comes to artificial intelligence and their health, it's a digital step too far for many.

Sixty percent of Americans who participated in a new Pew Research Center survey said they would be uncomfortable if a health care provider relied on artificial intelligence to diagnose an illness or recommend treatment. About 57% said the use of artificial intelligence would worsen their relationship with the supplier.

Only 38% believe that using AI to diagnose or recommend treatment will lead to better health outcomes; 33% said it would lead to worse results. And 27% said it wouldn't make much of a difference.

6 in 10 Americans say they don't want artificial intelligence robots to do some of their jobs. They also don't like the idea of ​​a chatbot working with them on their mental health; 79% said they do not want artificial intelligence to be included in their mental health care. Security is also an issue when it comes to AI and medical records.

“Awareness of AI continues to grow. So one dynamic here is that the public is not very familiar with all these technologies. So when you look at using it in a very personal context where the stakes are as high as your health, I think the idea that people are still learning about this technology is definitely dynamic," said Alec Tyson, associate director of research. : at Pew.

The findings, released Wednesday, are based on a Dec. 12-18 survey of 11,004 American adults using the American Trends Panel, an online survey panel collected from a random sample of residential addresses across the country. Pew weighted the survey to reflect American demographics, including race, gender, ethnicity, education and political party affiliation.

Respondents expressed concern about the pace of adoption of artificial intelligence in healthcare and medicine. Americans generally prefer health care providers to tread carefully and carefully weigh the implications of adopting artificial intelligence, Tyson said.

But they are completely against AI in healthcare. For example, you are familiar with its use to detect skin cancer; 65% believe it can improve diagnostic accuracy. Some dermatologists are already exploring the use of artificial intelligence technology in skin cancer diagnosis, but with limited success.

Four in 10 Americans believe artificial intelligence could also help providers make fewer mistakes, a serious health care problem. A 2022 study found that medical errors cost about $20 billion annually and cause about 100,000 deaths each year.

Some Americans also believe that artificial intelligence can improve health equity.

Research has shown that most providers have some form of implicit bias—more positive attitudes toward white patients and more negative attitudes toward people of color—and this can affect their decision-making.

Among survey respondents who perceived such bias to exist, the dominant view was that AI could help make diagnoses or recommend treatments and base those decisions more on evidence-based data.

Tyson said that when people were asked to describe in their own words how they thought AI would help fight prejudice, one participant mentioned class bias; assumptions about what it will do. . On the way, they were getting dressed for a date.

"So there's a sense that AI is more unbiased or at least less biased than humans," Tyson said. However, artificial intelligence develops with human input, so experts warn that it is not always completely neutral.

Previous Pew research on artificial intelligence has shown a general openness to AI, especially when it complements rather than replaces human decision-making, he said.

"Artificial intelligence is only one part of the process of helping people make decisions, and it supports a lot," Tyson said. "Less Final Decision Making for Artificial Intelligence."

For years, radiologists have used artificial intelligence to analyze X-rays and CT scans to detect cancer and improve diagnostic skills. The survey found that about 30% of radiologists use AI in their practice, and that number is growing, but more than 90% of respondents said they would not trust these tools for autonomous use.

Dr. Victor Tseng, a pulmonologist and chief medical officer at Ansible Health in California, said his practice is one of many that has studied the ChatGPT AI program. His group formed a committee to review its use and discuss the ethics of its use so that practices could take safeguards before implementing it into clinical practice.

This month, Tseng's group published a study showing that ChatGPT correctly answered enough practical questions to pass the US medical licensing exam.

Tseng says he doesn't think AI will ever replace doctors, but he believes technology like ChatGPT could make the medical profession more accessible. For example, a doctor can ask ChatGPT to simplify complex medical jargon so that a seventh grader can understand it.

"Artificial intelligence is here. The door is open," said Zeng.

The results of the Pew survey suggest that attitudes may be changing as more Americans become more comfortable with artificial intelligence. Respondents who were more familiar with the technology were more likely to support it, but expressed caution that doctors could adopt it too quickly.

"Whether you've heard a lot about artificial intelligence, little or nothing, all these segments of the audience are really in the same place," Tyson said. "They express a sense of caution about the implementation of artificial intelligence in healthcare."

What do Europeans really think of Americans?