AI could save your life. But your doctor is likely using it the wrong way. | Opinion
AI has extraordinary potential to transform the practice of medicine and to help patients lead longer and healthier lives.
Doctors are starting to integrate artificial intelligence tools into their practices. Already, nearly 4 in 10 physicians use AI, according to a recent survey from the American Medical Association.
But for the most part, doctors are using AI all wrong.
Doctors typically employ these new technologies as "virtual secretaries," using them to prepare clinic notes, respond to emails or write responses to insurers' denial letters. Nearly half of physicians surveyed had the greatest enthusiasm about using AI for simple documentation work and insurance authorization.
AI is great at administrative work. But if that's all it's used for, we're wasting its potential. Much of medicine is pattern recognition − in labs, imaging and symptoms. That is what supercomputers and machine learning excel at.
The most promising tools involve software and algorithms that can sort through staggering amounts of data to detect subtle patterns that doctors either can't detect on their own or can't detect nearly as quickly.
AI could help reduce medical errors
Consider this: There are nearly 800,000 preventable medical errors every year. About 40% of these are from misdiagnoses of common conditions such as pneumonia, strokes and lung cancer.
If AI can help reduce these errors, why wouldn't we embrace it? It would save countless lives and make physicians' jobs easier and more efficient.
Unfortunately, widespread adoption of AI is hindered by regulatory roadblocks and doctors' own pride.
Part of the problem is with the Food and Drug Administration, which admits that its standard regulatory process "was not designed" for integrating AI fully into medicine. Though it has proposed frameworks that factor in how AI continually improves itself, it hasn't solidified any approach.
Too often, regulators set a standard of near-perfection when evaluating AI tools, which can stifle innovation and make research overly complex.
Human doctors aren't expected to be perfect. We shouldn't expect perfection from AI.
To address AI-related issues, the Biden administration expressed interest in appointing a "chief artificial intelligence officer." But what we actually need is more investment in research and a practical, consistent regulatory framework.
Too many doctors also have a negativity bias and remain hyper-focused on AI's capacity for error. I'm constantly reading the newest studies and articles in medical journals, and that concern is omnipresent.
Those concerns are certainly valid, as are many others regarding the ethical use of AI. But I would urge physicians to have a more positive outlook and consider that on balance, these tools can improve the status quo of diagnosis and treatment − which is, of course, now far from perfect.
AI will only get better, but the speed with which it improves depends on how quickly doctors seriously engage with it.
Doctors are worried about job security
Let's also address the elephant in the room: Some doctors are concerned about AI because they secretly worry it will replace or minimize their responsibilities.
Once again, I suggest they welcome the positive changes that AI can bring to clinical care. Imagine how AI can help doctors in under-resourced communities design personalized treatment plans for their patients − without having to carve out several hours or wait days or weeks to consult with multiple specialists.
I’m excited about the prospect of using AI to enable virtually everybody to receive a second opinion on cancer diagnoses.
To those who worry about job stability − and the long-term relevance of physicians − look at the travel agent industry.
After websites like Expedia entered the picture, many people thought that travel agents' days were numbered. Instead, the business model changed. Technology freed agents from dealing with straightforward booking requests, allowing them to specialize in more complex trips that demand a human touch, like honeymoons, luxury vacations and adventure travel. Last year, the travel agent sector grew by 7.7%.
AI has extraordinary potential to transform the practice of medicine and help patients lead longer and healthier lives. But to make that a reality, doctors and regulators will have to view it in a new light.
Dr. John Whyte is chief medical officer of WebMD.