We're going to jump right in here, it's it's a pleasure for me to introduce our first speaker, Dr. Tom Lowry, and he will be speaking on artificial intelligence and healthcare. What does the future hold? Dr. Lowry serves as the National Director for AI for Health and Life Sciences at Microsoft, and previously served as Director of Worldwide Health at Microsoft. Tom works with providers, payers, life science organizations, and planning and implementing innovative analytical solutions that improve the quality and efficiency of healthcare services delivered around the globe. Tom focuses on strategy for digital transformation applied to performance optimization, including AI, machine learning, and cognitive services. Tom, thanks for being here with us this morning. The floor is yours. Hello, I'm Tom Lowry. I'm the National Director for Artificial Intelligence at Microsoft. Great to be with you today at this virtual event. So we're all getting used to virtual events. I am sheltering in place in my home office just outside of Monterey, California. Hope you're doing great wherever you are today. So the topic of my session today is AI and healthcare. What does the future hold? What I'd like to do is provide a few things for you to think about, but more importantly, create a framework around which some of the other great speakers that come after me can build upon. With that, let's get started. At Microsoft, we're making a very big bet on the cloud and artificial intelligence. Why? Well, we believe artificial intelligence represents one of technology's most important priorities and also believe that healthcare is perhaps the most urgent application when it comes to taking what we're doing today and helping make it better. Let's start by defining artificial intelligence. It's interesting to me how many times I can have a conversation with someone, get to the end of it, and find out when it comes to artificial intelligence, we're actually talking about different things. So for the sake of this session, artificial intelligence is defined as IT systems that sense, comprehend, act, and learn. Intelligence demonstrated by software with the ability to depict or mimic human brain functions. I'm going to emphasize mimic human brain functions. We'll talk about that more in a minute. When it comes to artificial intelligence, it's actually many things, not just one thing. And what we're seeing is all of those components that comprise artificial intelligence are evolving at warp speed, which is to say very fast. So when it comes to the capabilities we have, increasingly, we're getting those capabilities to reach human parity, which means they are just as good as an average human being. Just in the last few years, we've seen things like vision, speech recognition, reading, translation, speech synthesis, language understanding, and other things reach human parity. When it comes to the use of these growing applications in health and medicine, we're seeing a variety of applications and use cases ranging from very high-end things such as precision medicine, clinical trials, all the way down to things that seem simple and yet are so important when it comes to improving the operational efficiency of what you do. Things as simple as being able to predict which patients will not show up for their appointments and then getting out ahead of that to try and do something to make it different. When it comes to the application of AI in your own practice, there are a number of things that are already being published to demonstrate value in the use of AI. Some of the things on your screen are things that perhaps you're already involved in your practice. These things are benefiting from the likes of things like image analysis and being able to build intelligence into images. It's putting good use to things like natural language processing and knowledge extraction tools to be able to sift through massive amounts of data, things like EMRs, to come up with trends and predictions that you can, in fact, use in your practice. When it comes to looking at how AI is used to improve, say, your clinical practice or even your operational practice, what's important to understand is the value from AI comes from doing one of two things. It's either going to automate the way you work or it's going to augment the way you work. In automating work, the simple principle is work done by humans now will be done by smart machines and devices going forward. These typically involve highly repetitive activities with small variance in how work is done. This is going to be important, but frankly, it's not going to be the area where we see the greatest use and value. The greatest use and value comes by augmenting work, which means it improves the performance of humans but does not replace them. These are typically work activities that have high variance, require critical thinking, reasoning, judgment. It's going to help aid knowledge workers in completing tasks and workflow more effectively. It's not replacing, but instead, look at it as how to make whatever you do more efficient, more effective when it comes to the clinical practice of care. Improving the way you practice is an important part of what AI will be doing going forward. Another critical aspect when it comes to making the whole experience better is what we're able to do with the consumer themselves. In this regard, I'm reminded of a quote from Gandhi who said, there go my people. I must follow them for I am their leader. Well, as you know, the world is becoming more intelligent and mobile thanks to the cloud, the Internet, and a cornucopia of AI enabled apps and devices. So when it comes to better serving patients and health consumers, one, we have so much opportunity to use AI and smart applications to be better. But two, this survey shows we have a lot of room to grow. A survey from Change Healthcare and the Harris Pulling Organization just from this summer shows that finding, accessing, and paying for health care in America requires so much work that half of all consumers just avoid seeking care altogether. More than two-thirds of consumers said every step of the health care process is a chore. Nearly all said they want health care to be as easy as shopping for other common services, including making it a fully connected digital experience. All of this brings me to the belief when it comes to the future of health care of what we're seeing traditional health care systems really falling under the weight of the old ways in which things are done. Increasingly, what I'm seeing and predicting is what I call the rise of an intelligent health system. So the difference between traditional health care and intelligent health care systems is this. An intelligent health system is an entity that leverages data and AI to create strategic advantages through the efficient provision of health and medical services. And key to this and its success is in doing so across all touchpoints, experiences, and channels. When I say that, we're talking about consumers. We're also talking about how to make that experience of practicing medicine even better using the same techniques and the same ways of thinking and working. So I'm going to do a little call out here. I've got a new book called AI and Health, A Leader's Guide to Winning in the New Age of Intelligent Health Systems. This is a book that HIMSS came to me to ask me to write where they basically said, we've got lots of books on technology. We have virtually nothing when it comes to all of those elements that leaders need to understand to take AI, make it real, drive value across a practice or across an enterprise. So what I'd like to do is just throw out a few other things to think about that come from this book. So the first thing to consider is this. The data explosion we are seeing in health care gives health care leaders superpowers if they understand the value of the data and choose to use it wisely. So let's just take a look at that for a second from a clinical perspective and look at the time it takes for medical knowledge to double. So if you were a newly admitted physician starting your practice in 1950, you would go your entire career before medical knowledge doubled. By 1980, that was down to seven years. By 2010, 3.5 years. And today, in 2020, it is estimated that medical knowledge is doubling every 73 days. So you can be the best physician that's just coming out of medical school and the best residency, knowing that a couple of months from now, you're going to have twice the amount of data to deal with. That's where the combination of your training, your wisdom, your experience as a human mashed with AI to help sort through all of that data and empower you to be better at anything you care to do. That's the value proposition between the humans and artificial intelligence in making health care better. Well, in my work with clinical and health leaders, I occasionally hear a leader say that data is, in fact, the new currency in health care. And whenever I hear that, it always makes me pose the question of if data is new currency in your health care organization, are you managing data the same way you're managing your finances? And oftentimes when I ask that question, I get a chuckle and it allows us to have a deeper discussion about the value of data and the importance of developing, curating those systems that allow you to make the best use of that data that's growing today at an exponential rate. Another important topic is recognizing the value of artificial intelligence, but also recognizing the difference between what it can bring to make health care better and the unique role that individuals, humans and knowledge workers will always play. In this regard, I'm encouraging everyone to stop having what I call the humans versus machine debate. It is all about how humans harness the power of AI to do good. If we look back in history, the first industrial revolution was all about automating repetitive physical work. Today, the intelligence revolution is really focusing on knowledge workers, cognitive services, and having a similar effect when it comes to innovation through the provision of health services done on a more intelligent basis. To further differentiate this point between what humans and machines are best at, I want you to think about the fact that computers and AI are really good at certain things like pattern recognition, but there are many other things they're not good at, nor will they be at any point in the future. In this regard, I often like to quip about how everyone's talking about artificial intelligence, but have you ever heard anyone talk about artificial wisdom? Probably not, because it really doesn't exist. So think for a moment about the clinical care process. While it involves a lot of data and pattern recognition, so much of it involves things that are uniquely human, things like reasoning, judgment, imagination, empathy, creativity, and problem-solving skills. All of these things are critical to that care process that still involves data that AI can still help with, but these are the qualities that remain uniquely human. So to put a finer point on this, I'm not a data scientist, but I'm going to play one for this quick example. You see, I've gathered together a lot of data, I've used machine learning to make some predictions, and I've discovered that there's maybe a silent killer among us. You see, I've discovered there's an almost perfect correlation between per capita cheese consumption and the number of people who died by becoming tangled in their bedsheets. So let me pull back from being facetious and simply state that with the right data and the right tools, I can create correlations around virtually anything. But in reality, and particularly in the practice of medicine, there are so many variables that go beyond correlation when it comes to actually determining and understanding causation. It is with this regard that all of those skills, the wisdom experience you have, are critical to taking AI and having it drive value on a consistent basis in your practice. Well, there's one final topic I'd like to cover before I wrap up my session and we get on to some of the other great speakers you have today. And that is this, what are the societal implications of the technology we are creating? And particularly when I think about health and medicine, I believe that AI will be a force for good for all or for some. And key here is we all get to decide in terms of what we're doing with it and how we're managing it. Well, unfortunately, in the real world, AI is not neutral. There's a growing body of literature and books, including these two, which really go deeper in looking at why it is that AI is not neutral, number one. And number two, in the development, deployment, and use of such things as algorithms, we all must be mindful of this fact. When it comes to the practice of medicine, what we're seeing is data that suggests that many of the inequities and disparities in the real world are beginning to cross over into medicine through the use of algorithms in clinical decision support. Probably one of the most eye-opening articles I've seen recently came from the New England Journal of Medicine and one of their June issues, where they looked at 12 separate algorithms used in clinical decision support. And what they found is bias in a number of areas that makes those predictive capabilities good overall, but very uneven when it comes to being able to predict well with one group, but not another. So bias or unfairness in AI comes from many factors. It comes from the values and biases of those who create the algorithms, the data used to train it, or how it's actually used in practice. So at Microsoft, we are champions of what we call responsible AI. We've developed a set of principles to guide responsible data and AI. We apply these things every day to our own developments, and we encourage anyone using AI to do the same. Well, to help facilitate that, we created a digital or e-book called The Future Computed. If you're interested in learning more about principles of responsible data and AI, you can go out to any search engine, type in The Future Computed and Microsoft, and download a copy of this e-book. Well, let me wrap up this session by saying I would guess everyone listening would agree that there's lots of change coming when it comes to the way health and medicine works going forward. I also believe that AI is going to be a big part of this. For those who agree and want to be part of this new AI leadership, on the screen are some of the characteristics that I write about in my book that are critical to success in terms of making that transformation from where we are to where we wish to be. Well, let's wrap up by telling a story about how change happens. So I love data, and I also love history. So the picture on your screen is New York City, 1905. The Flatiron Building had just been built, and this is a picture looking at the Flatiron at the intersection of Broadway, Fifth Avenue, and East 32nd. If you look closely, what do you see? What you see is an economy, a lifestyle, totally driven by horses. At this time in New York, it took 100,000 horses in New York alone just to deliver goods and supplies. There were tens and thousands of workers who cleaned up after horses, who fed horses, who built buggies, who were saddle makers. Decades worth of careers and professions handed down through generations. And it was a great life, and across America, one quarter of all agricultural output went to feeding horses. So here's that same view 20 years later. There's a different type of horsepower that has now taken hold. What we see is new industries, new ways of living and working that have sprung up and become impervious to all of that past history that went back 100 years. In this regard, just as New York changed, I believe healthcare and medicine is on the cusp of radical, dramatic change. And I guess I'd leave you with the thought of the mission of your organization, the motivations for why you want to practice remain the same. How those things work going forward will face lots of change, lots of transformation. And it really is up to you to decide where and how you will play throughout all of that change. And so that's it. I'd like to thank the conference organizers for the invitation to speak with you today. If you're interested in continuing the conversation, we'd love to have you join me on LinkedIn. If you're a Twitter user, follow me on Twitter at TCLorry, or email me at my name, TomLorry at Microsoft.com if you have questions or thoughts. With that, I hope you have a great rest of the day and a great conference.

artificial intelligence

healthcare

AI applications

automation

augmentation