As artificial intelligence continues to advance, its application in business sectors is under increasing scrutiny. This is especially true in the healthcare sector, where AI is used in a variety of ways to help the sick.
Many people can now receive AI-generated health information with tools like ChatGPT, which are “helpful in learning more about certain conditions or symptoms,” VeryWell Health reported. The chatbot can “answer questions in seconds and instantly generate easy-to-understand responses.” Clinical applications of AI are also being explored in medical practices and hospitals. In 2022, the FDA approved 91 AI or machine learning medical devices for widespread use.
However, controversy simmers over the purpose of artificial intelligence in medicine and whether it does more harm than good. As with many AI implementations, its use in medicine has not been without its problems, leading some to question whether treating disease is a field that should be left to humans.
AI-generated medical content
Many AI chatbots are used in home settings to help people trying to self-diagnose. Many of these patients have taken matters into their own hands with the help of AI. Benjamin Tolchin, a neurologist and ethicist at Yale University, told Scientific American that “at least two patients have already told him they use it to self-diagnose symptoms or to look for side effects of medications.” . Some researchers are confident that by the end of the year, “a major medical center will announce a collaboration using LLM chatbots to interact with patients and diagnose disease,” the outlet reported.
While that sounds promising, AI-based medical chatbots aren’t without their flaws, especially when it comes to eating disorders. “I recently asked ChatGPT what drugs I could use to induce vomiting,” Geoffrey Fowler reported in an experiment for The Washington Post. “The bot warned me that it should be done under medical supervision, but then decided to name three drugs.” Google’s Bard AI, meanwhile, “has produced a step-by-step guide to ‘chewing and spitting,’ another eating disorder practice,” Fowler added. My AI, a program developed by Snapchat, also wrote Fowler of a “weight loss meal plan totaling less than 700 calories a day – well below what a doctor would ever recommend.”
It appears the AI ”learned some deeply unhealthy ideas about body image and diet by browsing the internet,” Fowler concluded. “And some of the best-funded tech companies in the world don’t stop it from repeating them.”
AI-assisted medical devices
Artificial intelligence has extended its use to hospital computer systems and, in some cases, to hospital equipment itself. This may include technologies that can “diagnose melanoma, breast cancer lymph node metastases, and diabetic eye disease better than specialists when they work well,” according to a study published in the medical journal EBioMedicine. Advances in AI have also enabled medical centers to be “equipped with passive, non-contact sensors [that] can help clinicians and surgeons improve the quality of health care delivery,” the study states.
Although this seems like another benefit for the industry, “the development of artificial intelligence also has the potential to produce negative health effects”, according to a study by BMJ’s Global Health, by The Guardian. These could include “the potential for AI errors to cause harm to patients, privacy and data security issues, and the use of AI in ways that worsen social and health inequalities”, such as a pulse oximeter that overestimates blood oxygen levels, leading to hypoxia in some patients.
While AI may “ultimately become a mature and effective tool for the healthcare sector”, a study published by Springer Nature stated that “barriers arise at all levels of AI adoption. AI”.
The future of AI in healthcare
There is no evidence that AI in medicine will soon disappear. Even if AI tools were “much less accurate at providing diagnoses than doctors,” the Harvard School of Public Health reported, “we can see a future where people frequently turn to these types of tools to get advices”.
Although fears about the future remain, AI will likely never fully replace humans in healthcare, Tom Lawry, national AI director for healthcare and life sciences at Microsoft, told Forbes. While AI is “great at sifting through massive amounts of data,” Lawry added, “humans are gifted with wisdom, common sense, empathy, and creativity, all of which are a vital importance” for health care.