The AI Imperative According To Leading Physicians And AI Executives

When it comes to technology, the American Medical Association is very definitive about its position on AI. The term is a carefully chosen one – one meant to highlight the fact that while it might be artificial, it is augmented intelligence. There's no limit to how machine learning technology can contribute to breakthroughs in healthcare research, but it can't do it by itself.

“In health care, machines are not acting alone but rather in concert and in careful guidance with humans, [like]… physicians,” said Jesse Ehrenfeld MD, AMA chair for the board of trustees. ”There is and will continue to be a human component to medicine, which cannot be replaced. AI is best optimized when it is designed to leverage human intelligence.”

Become a Subscriber

Please purchase a subscription to continue reading this article.

Subscribe Now

Ehrenfeld spoke on the topic at the AMA State Advocacy Summit, where he addressed the points raised by the National Academy of Medicine's (NAM) report called "the hope, hype, promise, and peril of AI."

In a speech given to a crowd of physicians and medical staff at the Bonita Springs, Florida, event, he emphasizes that developments in recent history hold the wisdom for how best to approach challenges raised by AI. He argues the advent of new technologies like genetics, EHRs, and digital medicine all raise very similar questions on topics like policy, regulation, liability, training, and professional development.

The takeaway, Ehrenfeld says, is the importance of settling these critical policy issues before asking physicians to incorporate these technologies into their practice. If the problems are not solved beforehand, it can disrupt practices and create challenges for implementation. Pointing to trends in research and investment dollars, Ehrenfeld believes that essential areas for AI adoption will be diagnostic tools and health care administration.

Another panelist, Sonoo Thadaney Israni, focused on a different aspect of AI technology in healthcare. As the director of Stanford University's School of Medicine's Presence, Israni runs an interdisciplinary center that brings together the art and science of human connection in medicine. She is also a member of the NAM panel on AI in health care and spoke in her presentation about moving away from a “rote medical history” that medical providers go through to prepare for a patient visit.

Interestingly, Israni imagines an AI-driven EHR system that presents information in the form of graphics and animation. The intent behind the idea is to make it possible for physicians to envision exactly where this patient is in their life without having to slog through mountains of data.

She warns that AI comes with blind sports – the technology is not suited to sort information that is not representative of a population. Israni cited the NAM report, where AI-powered tech attempted to tackle anti-hunger efforts but ended up optimizing total production calories by acre instead of optimizing nutrition – ending with people following diets consisting of empty calories. She cautions against AI that leads to "empty health care."

Another challenge addressed at the event was the attribution of responsibility for those companies who create AI solutions for the medical space. As the founder and executive chair of IDX Technologies, Michael Abramoff warns against "glamour AI" – tech that looks exciting but doesn't actually produce better patient outcomes. His firm developed a device for physicians to detect diabetic retinopathy using an AI algorithm.

“We see liability as essential,” said Abramoff. “As a doctor, you’re liable for your diagnosis, your medical decision. As a creator of AI, if you say: ‘It does what a specialist does,” you have to assume liability.”