Copyright Fast Company
At its core, healthcare is about people helping people: doctors diagnosing patients, nurses caring for the infirm, surgeons performing life-saving procedures. But technology—from MRI machines to groundbreaking gene therapies—has also played a key role in pushing healthcare forward. Now, artificial intelligence has the potential to move healthcare by leaps and bounds. Indeed, Dr. Marc Triola, director of the Institute for Innovations in Medical Education at NYU Langone Health, believes AI is already making a difference for healthcare providers, their patients, and the healthcare system as a whole. During the recent Fast Company Innovation Festival, NYU Langone Health hosted a dinner where Dr. Triola discussed the intersection of AI and healthcare. The following are edited remarks from that event and a follow-up interview. You’re a big believer in the power of generative AI to improve the American healthcare system. What are the challenges AI might help address? Healthcare is a tremendous expense and investment for the United States, and it’s a worthwhile one—it’s obviously one of the most important and essential things that we can invest in. When it comes to things like access to care, we can’t just throw more money at it and say, “Double the number of doctors; double the number of nurses.” The beauty of AI is that it’s extending our existing healthcare workforce to better meet the needs of the patients they’re caring for now and to better meet the needs that are going unmet across the country. Where are you seeing AI really move the needle in terms of patient care? Mental health is often an area where there can be a disconnect between a patient’s need and their access and availability of healthcare. There was a recent study that looked at using chatbots for supplementing—not replacing—care delivered by licensed mental healthcare professionals. It was a way to really interact with these patients, to gather their symptoms, to provide them with information and counseling without extending or overtaxing the human workforce. And the study found that they would’ve had to have hired four more mental health professionals to do what the chatbot did in a given month. It’s really exciting to consider that AI, working in conjunction with doctors and nurses, can really extend our ability to deliver care, to reach patients, and to stay in frequent contact with them in the outpatient setting. The goal is to have better outcomes and a more efficient healthcare system that’s better tailored to the true needs of people across society. How quickly did NYU Langone Health embrace the new generation of generative AI tools such as ChatGPT? When these models were first released, NYU Langone Health saw tremendous opportunity in their use. From the very beginning, our leadership supported investigating how these models could safely, ethically, and equitably solve many of the everyday healthcare challenges that we face. So, it’s been a full-court press since the very beginning. We’re proud that we were the first integrated academic health system in the country to have a HIPAA-compliant private instance of ChatGPT that would enable us to do this experimentation safely with no risk to patient privacy and confidentiality. We’ve asked everyone in our NYU Langone Health community to help us by identifying problems or opportunities big and small in their everyday work where they think AI could help. We’re moving forward across thousands of different applications, and dozens and dozens of different AI models are running live within our system to figure out how to deliver a healthcare system that is safe, reliable, and vastly more efficient. Looking into the future, what do you think AI will help make possible in the healthcare space? I think that we’re going to see an explosion in humanoid robotics. Hospitals across the country are already using robots to do things like deliver medications to the nursing station and deliver food trays up to the floor. Imagine a robot that could move patients around or clean operating rooms and patient beds between procedures or between hospital stays. Doing things in situations where a patient might be in isolation for an infectious disease or because of a weakened immune system. Doing physical therapy or helping patients at home who have mobility issues. I think the sky’s the limit. I also think that in the next five years, we’ll see examples of these AI systems autonomously interact with patients, in addition to their human care teams.
