How Amazon and Columbia University are collaborating to advance AI in healthcare

Mishra: At Amazon Health Services, our mission is to make it easier for people to find the products and services that they need to get and stay healthy. We believe that the experience of accessing care just needs to be a lot easier. Collaborating with one of the nation's top-tier universities is a great opportunity to expand our exploration of artificial intelligence and large language models as tools for some of the things that we're setting out to do.

Armstrong: We share a very clear commitment to using digital tools to improve access, quality, and equity in healthcare. Building upon existing partnerships with the engineering school here, Amazon and Columbia have an opportunity to develop new ways to study how we can improve healthcare access with digital tools; to understand the ways to build trust and enhance communication; and to really become a model of a partnership between an academic medical center, a university, and a tech giant that is able to disseminate the knowledge we can create to almost everyone.

Armstrong: We at Columbia bring this innovative energy of young, bright physicians and scientists who are actually at the front lines of healthcare issues. And then Amazon has an extraordinary ability to think broadly about how to scale their incredible vision for what it can mean to create a digital world that we all live in. Any great idea that we can co-develop has the opportunity to have the greatest possible impact.

Mishra: Collaborations like this one with Columbia are not unusual at Amazon, and they're valuable, because they open up the opportunity to collaborate with people or institutions that have a whole different set of expertise than we do. At the symposium in October, there were participants from Columbia's medical college, their business school, and their engineering school. To be there with people who were all looking at a similar area with a different lens was super energizing.

Mishra: AI is among the most transformational technologies of our time, and I believe generative AI can support the transformation of the healthcare experience. One of the areas I am most interested in is how we can use artificial intelligence tools and large language models to make the job of the clinician easier. This would allow clinicians to spend more time building relationships with patients and not on the administrative tasks that we know can take up nearly 50 percent of their day.

At Amazon Pharmacy, we have applied large language models to improve the transparency of price so you're never surprised by the cost of what you've been prescribed. That approach would be really valuable in other aspects of care, where you could help people understand what their copay is or how much of their deductible is remaining.

At the symposium, the team at Columbia brought up that we could use predictive analytics to really understand the waxing and waning of demand from patients and use that to make sure we're staffed up appropriately.

Armstrong: Over the last 30 years in medicine and in science, we've made amazing advances in what we can offer patients. The problem is, we haven't been able to get all of that out to the people who really need it. So what we've been thinking about for the last decade is, how do we take these tools and create new models that are going to get everybody access to the information they need to make the right decisions and then build upon that with new models of delivery?

Recently there was a new gene therapy approved for sickle cell anemia. Imagine if we could create tools that identify and inform candidates for treatment. Then we could use AI and other tools to identify how best to do that, to follow patients, and to make better decisions as we go forward.

So that's what I'm excited about. It's not AI on its own. It's really about how this helps us address this fundamental crisis that's driven me forever, which is: Why on Earth can't we help more people?

Armstrong: We have faculty who have been studying, working, and living everything about AI for decades. What is remarkable now is that we've democratized it. So, as opposed to a small group of individuals testing and advancing it, we're at this point where AI in healthcare is becoming something that we're thinking about across every aspect of what we do.

My sense is that we are going to see many parts of healthcare — particularly how we manage and use data — move fairly quickly to better models, whether that's how we collect information from patients, how we provide information to patients, or how we use that information for quality and safety efforts in hospitals and physician practices.

I think it's going to take longer for it to really become part of how we do medical education, which is where I spend a lot of time. And it's going to take some really Innovative people to come up with ways that AI truly transforms our ability to address the fundamental challenges of healthcare access, quality, and cost. That's going to take some big, bold moves.

Mishra: The potential is huge; AI is a very powerful tool, and in many ways, Amazon is already improving customers’ lives with practical, useful generative AI innovations. When it comes to AI in healthcare, I'm optimistic about it, and I have been impressed with the care and thought with which people in the industry are approaching solutions.

Amazon is a place where you can make a significant impact on the lives of customers, and to be able to help make it easier for people to engage with their health is tremendously meaningful work. Most primary care clinicians will see their patients a couple of times a year. We have people who visit the Amazon Store much more frequently. That relationship gives us an opportunity to surface things that are important for their health and to close care gaps. For example, what if we could remind people about checkups or screenings that they may be due for, or just make it easier to schedule appointments or refill their prescriptions? That will make a huge difference. That's what gets me out of bed every day.