"Is Google's AI the Future of Healthcare? Find Out if You Can Trust It with Your Health Questions!"

 Do you have a headache or sinus infection? What does a stress fracture feel like? Should you be worried about pain in your chest? If you google those questions now, the answers are probably written by artificial intelligence.

Image : Google.com/images.app.goo.gl/j6DimYS61Zr2grJCA

This month, Google launched a new feature called AI Overview that uses generative AI, a type of machine learning technology that is trained on information from the internet and produces conversational answers to some search questions in a matter of seconds.

In the weeks since the tool launched, users have encountered a variety of inaccurate and strange answers on a variety of topics. But when it comes to how to answer health questions, experts say the risks are huge. This technology can lead people to healthier habits or need medical care, but it also has the potential to provide inaccurate information. AI can sometimes fabricate facts. And if the answers are formed by sites that are not based on science, then they may provide advice that goes against medical guidelines or poses a risk to a person's health.

This system has been proven to produce bad answers that appear to be based on flawed sources. When asked “how many rocks should I eat,” for example, AI Review asked some users to eat at least one rock a day to get vitamins and minerals. (The advice is taken from The Onion, a satirical site.)

“You can't believe everything you read,” says Dr. Karandeep Singh, chief health AI officer at UC San Diego Health. In health, he said, sources of information are very important.

Hema Budaraju, Google's senior director of product management who helped lead work on AI Overview, said that health search has “additional guardrails,” but declined to explain them in detail. Searches that are considered dangerous or explicit, or that indicate that a person is in a vulnerable situation, such as self-harm, do not trigger an AI summary, he said.

Google declined to provide a detailed list of websites that support the information in AI Overview, but said the tool works in conjunction with Google's Knowledge Graph, an existing information system that has pulled in billions of facts from hundreds of sources.

Take the question “Is chocolate healthy?” Google Answers pulls information from research on heart health, mental health, and more.

Answers to health questions like these often come from reputable sources. However, in this case, the response also cited Venchi, an Italian chocolate and gelato company.

Similar searches — “Is chocolate healthy for you?” — generated answers from a variety of sources including the website of a company called ZOE, which sells at-home “gut intelligence tests” and nutrition apps.

The new search response specifies multiple sources; for health questions, there are usually websites like the Mayo Clinic, WebMD, the World Health Organization, and the scientific research center PubMed. But this is not an exhaustive list: These tools can also be pulled from Wikipedia, blog posts, Reddit, and e-commerce websites. And it doesn't tell users which facts come from which sources.

With standard search results, many users can immediately differentiate between a reputable medical website and a candy company. But a single block of text that combines information from multiple sources can cause confusion.

"That's if people look at the source," said Dr. Seema Yasmin, director of the Stanford Health Communication Initiative, added, “I don't know if people are looking for it, or if we've really taught them enough. to look." He said his research into misinformation has made him pessimistic about the average user's interest in looking beyond a quick answer.

Regarding the accuracy of chocolate's answer, Dr. Dariush Mozaffarian, a cardiologist and professor of medicine at Tufts University, says that there are several facts that are mostly true and summarize the research on the health benefits of chocolate. But it doesn't differentiate between strong evidence provided by randomized trials and weak evidence from observational studies, he said, or provide any caveats to that evidence.

It's true that chocolate contains antioxidants, says Mozaffarian. But the claim that consuming chocolate can help prevent memory loss? This has not been clearly proven, and “a lot of warning is needed,” he said. Listing such claims side by side gives the impression that they are more established than they actually are.

The answers may also change as the AI ​​itself develops, even though the science behind the answers given does not change.

A Google spokesperson said in a statement that the company attempted to provide a disclaimer for the required response, including a note that the information should not be considered medical advice.

It's not clear how exactly AI Overview evaluates the strength of evidence, or whether AI Overview considers conflicting research findings, such as whether coffee is good for you. "Science is not a static collection of facts," said Yasmin. He and other experts also questioned whether the tool would take advantage of old scientific findings that have been debunked or not capture current understanding of an issue.

“Being able to make important decisions – to discern the quality of the source – that's what humans do all the time, what doctors do,” said Dr. Danielle Bitterman, artificial intelligence physician-scientist at Dana-Farber Cancer Institute and Brigham and Women's Hospital. “They are outlining the evidence.”

If we want tools like AI Overview to play that role, he said, “we need to better understand how they navigate multiple sources and how they apply critical viewpoints to arrive at a summary,” he said.

These unknowns are worrying, experts say, considering the new system improves AI Overview responses over individual links to leading medical websites such as the Mayo Clinic and the Cleveland Clinic. Such sites have historically topped the results of many health searches.

A Google spokesperson said that Overview AI will match or summarize information that appears in top search results, but is not designed to replace that content. Rather, the spokesperson said, it is designed to help the public understand the information available.

Mayo Clinic declined to comment on the new response. A representative from the Cleveland Clinic said that people seeking health information should “go directly to known and trusted sources” and contact a health care provider if they experience any symptoms.

A representative from Scripps Health, a California-based healthcare system quoted in several AI Overview summaries, said in a statement that “citations in Google's AI-generated responses can help in establishing Scripps Health as a reputable source of health information.”

However, the representative added, “we have concerns that we cannot guarantee content generated through AI in the same way as we can for our own content, which is vetted by our medical professionals.”

For medical questions, it's not just the accuracy of the answer that matters, but how the answer is presented to users, experts say. Take the question “Am I having a heart attack?” The AI ​​response has a useful synopsis of symptoms, says Dr. Richard Gumina, director of cardiovascular medicine at Ohio State University Wexner Medical Center.

But, he added, he had to read a long list of symptoms before the message advised him to call 911. Gumina also searched “Am I having a stroke?” to see if the tool could generate a more urgent response – and it did so by asking users at the front of the line to call 911. It said it would immediately advise patients experiencing symptoms of a heart attack or stroke to seek help.

Experts urge people seeking health information to approach this new approach with caution. Basically, they say, users should pay attention to the details below some of the AI ​​Overview answers: “This is for informational purposes only. For medical advice or diagnosis, consult a professional. Generative AI is experimental.”

Location:

Share :