AI as Doctor: How ChatGPT and Artificial Intelligence Are Transforming Healthcare
Hundreds of millions are turning to artificial intelligence for wellness advice,and the impact on both patients and practitioners is already profound. from identifying rare conditions to streamlining doctor’s appointments, AI is rapidly reshaping the landscape of modern healthcare – but not without risks.
The potential of AI in medicine gained widespread attention recently with stories of individuals receiving potentially life-saving advice from platforms like ChatGPT. Bethany Crystal,a New York-based consultant,recounted in a viral video how ChatGPT alerted her to a possible bleeding risk after she noticed red spots on her legs. “You need immediate evaluation,” the AI told her, prompting a visit to the emergency room where she was diagnosed with a rare autoimmune disorder. Crystal believes the timely intervention, spurred by AI, was critical.
OpenAI, the creator of ChatGPT, has responded to the growing demand with the launch of chatgpt Health in early january, a platform designed with enhanced security for sharing medical data. Other tools, such as My Doctor Freind, are also emerging, promising to partner with patients in navigating their healthcare journeys.
The Limits of Human Diagnosis
Doctors and patients alike acknowledge the transformative potential of AI. A common saying in medicine emphasizes focusing on the most obvious diagnosis – “If you hear hoofbeats, think of horses, not zebras.” However, this approach can sometimes overlook rarer, more complex conditions.
“I’ve heard from a number of patients who said, ‘Well, guess what? I’m a zebra,'” says Dave deBronkart, a cancer survivor and patient advocate. Unlike time-constrained physicians, ChatGPT possesses the capacity for exhaustive inquiry, potentially identifying subtle symptoms that differentiate unusual conditions from more common ailments.
Moreover, AI’s diagnostic capabilities extend beyond generalized med
ical knowledge. It can analyze vast datasets of patient records, identifying patterns and correlations that might escape human observation. However, this power comes with a critical caveat: AI is only as good as the data it’s trained on. biases in the data can lead to skewed results and potentially harmful recommendations.
The Dangers of Self-Diagnosis
The most significant risk associated with AI-driven healthcare lies in the potential for misdiagnosis and inappropriate treatment when used without the guidance of a qualified physician. without clinical oversight, misdiagnosis, misleading advice, and human misunderstanding pose significant risks. Dr. Robert Wachter, chair of the Department of medicine at University of California, San Francisco, and author of the forthcoming book A Giant leap, recounts a case where AI advised a patient to use the anti-parasitic drug ivermectin to treat testicular cancer.
“It probably wouldn’t hurt you, but what would hurt you is you not getting appropriate treatment for your cancer that is treatable,” Wachter warns. “So,the capacity for badness here is pretty high.” A separate incident involved a 60-year-old man experiencing paranoia and hallucinations after following AI’s advice on reducing salt intake.
Despite these dangers, Wachter remains optimistic about AI’s potential.Studies indicate that large language models are competitive with human doctors in simulated diagnostic tests, with a recent New England Journal of Medicine study showing AI frequently identifying difficult cases. While a follow-up comparison showed a slight advantage for human diagnosticians, Wachter calls the AI’s performance “remarkable.”
AI as a Tool for Doctors, Not a Replacement
AI is already improving the workflow of medical professionals. Wachter now uses AI Scribe, a tool that allows him to maintain eye contact with patients during consultations, and his colleagues are widely adopting OpenEvidence, a “ChatGPT for doctors” providing access to exhaustive medical knowledge. “I use it all the time,” he says. “We all do.”
Patients and doctors agree that the pace of AI integration into healthcare is staggering. Rosen believes that future generations will seamlessly embrace AI as a core component of their care. “Two generations from now,” he predicts, “no one will give it a second thought.”
wachter acknowledges broader concerns about AI’s impact on society, including its potential effects on politics and employment. However, he argues that the uniquely flawed state of the U.S. healthcare system necessitates innovative solutions.”I just see a system that is falling apart and can’t possibly meet the needs of people without this kind of help.”
