Explore the transformative impact of AI-powered automation in healthcare. This session delves into how intelligent automation is revolutionizing the industry by streamlining operations, enhancing patient care, and empowering healthcare professionals. Learn about innovative applications beyond direct care delivery, including revenue cycle management, collections, and data unification. Discover how these advancements are simultaneously improving efficiency, reducing costs, and reimagining the human element in healthcare
Speakers
Yan Chow, MD, MBA, Global Healthcare Leader, Automation Anywhere
Megan Antonelli, Chief Executive Officer, HealthIMPACT
[00:00:00 ]VO : Welcome to Digital Health talks. Each week, we meet with the healthcare leaders making an immeasurable difference in equity, access and quality. Hear about what tech is worth investing in and what isn't. As we focus on the innovations that deliver. Join Megan Antonelli, Janae sharp and Shahid Shah for a weekly no BS deep dive on what's really making an impact in healthcare.
[00:00:30] Megan Antonelli Hi everyone. Welcome to Health Impact, live digital health Talks. Today, I'm excited to have Dr Yan Chow with us. Dr Chow is the global healthcare industry lead at automation anywhere. His impressive career includes roles at Amgen and Kaiser, where he's been at the forefront of digital innovation and technology. Today, we're going to explore the transformative impact of AI powered automation and healthcare. Please join me in welcoming. Dr Yan Chao, hi. Dr Chow, how are you?
[00:00:57] Yan Chow: Thanks, Megan, it's great to be here, to be here again, actually, and we have some interesting things to talk about.
[00:01:03] Megan Antonelli: Absolutely, I always love speaking with you, because I inevitably learned so much about, you know, how health or care organizations are implementing and using automation. And you know, we talked for the first time many, you know, a few years ago. So you guys have been at this for quite a while now. And I know, you know, have seen some amazing results, you know, but it's, you know, it's been so interesting to watch, kind of the trajectory of all of this, right? I mean, you were with us at Health Impact, I think, in November, you know, or January of 23 but, you know, chatgpt was just sort of on the scene, and we were all talking about the impact and the potential, and there was a lot of fear in the room, then a lot of concern. And I think some of that has changed, you know, and the sort of the hype cycle is, you know, maybe we're at the kind of the cresting of that now, but I'd love your perspective on it, having been at this for so long to, you know, and sort of what you're seeing with with the health systems around the country,
[00:02:04] Yan Chow:you know, with the chat GPT, the following hymns, conference, you couldn't walk 20 feet without hearing the word chat GPT. And of course, now we have, you know, we have Claude, we have Gemini, we have a lot of other things and that that's just sort of exploded, you know, since that time. And I think the reason it exposed is because, for the first time in healthcare, doctors and nurses and other folks are seeing that this is a technology that could actually make their work better, and as opposed to all the other technologies and regulatory and billing and things, this could actually help their lives. And so there was a lot of excitement early on, but because, like any new technology, it was early, there was no guidelines. So people started experimenting. In fact, I heard the early apocryphal story of somebody, some physician somewhere, who decided to use chatgpt to appeal a claim denial, and it created a very beautiful document. The only problem is all the citations were fictional. You know, they were not real. And so he got caught for that. But it's part of the early process of trying to figure out what's good and bad about chat GPT, especially the early version. So we saw a lot of interest from especially from academic medical centers. And so, you know, because they're they are able to experiment a bit. And we saw things like pilot projects for a personalized patient after visit summary, you know, which is personalized according to language and culture, educational level, even a neighborhood, you know, you can imagine an after visit summary from a doctor, which I've never gotten that would say something like, Hey, guess what? Your nearest Walgreens is 2.3 miles you take this bus and your discount is waiting for you there, you know? And you would never see that, because it would just take too much work to do that. Now, with computers, you can do that. The after visit summary could be rephrased a couple times for different family members who don't have, you know, who have certain reading levels and things like that. So, so that was very exciting. Now I see a switch over to some of the concerns about generative AI and AI in general. The first of this month, I think, was the enforcement date for the new European Union AI act, and that's a very comprehensive effort at comprehensive governance of AI, both responsible AI and other things. I think it's going to be the first of many. And I think obviously there are a lot of concerns with AI and its impact, and we're just at the cusp of figuring out what to do
[00:04:46] Megan Antonelli: Yeah, no doubt. And I think, I mean, you know what you said about sort of that, even the patient summary, I mean, things that we just couldn't do before as clinicians, you know, being able to. To sort of do those things and amplify them and that they are, you know, incredibly useful, right? You know, and valuable to the patient, but at the same time, you know, ensuring their accuracy, and, in some case, making sure that they're, you know, the things that we start with are a little bit lower risk, right? Tell me a little bit more about what you see in terms of the regulations that are coming and how organizations can, like, prepare now in terms of what they think may be, you know, impacting certainly here in the US, but even abroad.
[00:05:35] Yan Chow: Well, you know, the whole area of governance of AI is really exploding. A lot of organizations are now trying to figure out, either singly or in in groups. We try to figure out what is the right way to deal with AI. And of course, when you do start doing that, you know there are bad actors, and are good actors for that. Bad actors, like the folks that would never follow the nuclear proliferation treaty, those are the guys. But the good guys are trying to, trying to come up with a reasonable way to ensure fairness, transparency, all the good stuff that you want to hit with responsible AI, including sustainability, which I find really interesting. The cost of AI for natural resources is huge, and all the other issues you know, AI itself has been known to be biased based on bias data. It may be unfair. It may make judgments that are not completely objective. You know, it's based on its the studies that go into it, the algorithms are opaque now, sometimes the explainability level is not very high, so you have to ask people to trust in something that you know. But on the other hand, you know, you look at that and you go, Okay, you can't explain it. But if you ask a physician who's been in practice of 40 years, and you ask them why they chose a certain medicine, they probably wouldn't be able to tell you, they kind of synthesize all that stuff from their experience, and they just know, you know from a variety of inputs that that's what they would choose. So that's what our comparison is. But I think the whole area of AI is just evolving so fast now, on the on top of that, there are other regulatory drivers, especially for the federal government, the O and C has decided in the last 20 years to really step up its push for data exchange and data liquidity interoperability between different players in the healthcare system, and that has come across in the form of no surprises act, information blocking rule and a lot of this pricing transparency rules that are going to hit that already hit payers and providers. So what are they to do? The problem is that these rules require you to share data in a fairly comprehensive way. And you can imagine, let's say, for Kaiser, which is where I used to work, they have over 12 million patients. Let's say 100 of them decided to come in and get their data. That's the revenue cycle department's job. So they they have to get somebody who they don't have, you know, to go and collect this data, including financial data that impacted the care trying to figure out what it goes in and what doesn't go in, and then have to deliver that within two, three days. That's impossible. That is something that. And the same thing goes from the no surprises act, creating the argument for your arbitration requires a lot of manual labor. I don't see people doing that without automation. I mean, it's hopeless, you know. So you want to do it so that when they come in and say, I want my record, you click a button and then they get the record the way it's supposed to be. And I think that's sort of the way that regulation is going, and it's good for the patient. I mean, it's good for visibility into your medical record. You can check it make sure it's a correct you can shop, you know, for services. That's what the government is pushing and so and so forth. So in general, it'd be good, but the implementation is a different question, as it always is, so
[00:09:02] Megan Antonelli: right, yeah. And, I mean, without, as you said, Without automation, it almost, you know, the demands from all different sides are almost impossible. And, you know, I think, you know, and we see this sort of push pull between kind of the risk, the safety, the privacy, the security. I mean, certainly this year we've also seen, you know, a number of breaches, I mean, and a number and then there was, you know, even the downtime, disruptions of service and things like that. What are you seeing in terms of applications or even just safeguards that you've been putting in place to kind of combat some of these you know, sort of, as you know, sort of bad actors. And in some cases, you know, not even bad actors. It's just stuff, you know, stuff that happens in the natural course of business.
[00:09:50] Yan Chow: You know, I have a work laptop, and typically what that means, and I work in a software company, so they should know. So this work laptop is locked down to the utmost. I mean, it's, you can't even send stuff out, you know, things like, which is good, but it makes it inconvenient, right? So, so we haven't quite figured out how to totally protect. I mean, total protector is taking yourself offline, and that's a total protection. On top of that, people now are using AI, doing deep fakes and fishing with, you know, people that sound and look just like somebody else. And that's that's very worrisome. People are using AI to fish out the data, the information, a lot of which is on the dark web, to to get access to very sensitive systems. And in particular, I read a really interesting article the other day about Gen AI. And in particular, it's really hard to safeguard against, really hard because the prompt is open. You can put whatever you want in there, including sets of data, and you're asking it to create data, create new data. And I'm not an expert, but apparently that makes it very hard, very hard to to and ultimately, I think that we're probably going to need countervailing AIs. We're going to have to develop ais that can check other AIS, because humans can't do it. We don't understand what AI is doing, and unless you can, and there are companies trying to work on the explainability of AI, but I see the AI is so valuable, it's going to be hard to say everybody has to be explainable before we use it. That that is very difficult to do.
[00:11:28] Megan Antonelli: Yeah, I just read an article the other day. I think maybe this morning, it feels like the other day, but about even the how the models now you know how you have to be careful of training the models on on their own data, because then they can call cause model collapse and things like that. So I mean, as I mean, this gets into the sort of technicality of of these models. But you know how we then, you know, kind of protect against the evolution of them and ensure the quality of them throughout is a huge piece of that.
[00:12:05] Yan Chow: I think. When I get calls from young physicians and they ask, what do you recommend I do in the future? Because I don't know what medicine is going to look like, and I tell them, I don't know what it's going to look like, but one thing for sure, you need to look at AI. Two things, well, three things, AI automation and AI automation and data data science. And if you look at those, and you and you learn something about those, you'll be in good shape. I think, I think all our roles are going to change. You know, when we have aI avatars, we have aI voiceovers, AI content. I mean, it's just mind blowing how, how is going to change? And so what, what that means is your role is going to change. And for a physician, what does that mean? You know when, when AI can can already outperform you in terms of memory, decision making, even maybe judgment in some cases, especially in the imaging fields. What is the role of a physician? Is it to explain it to patients? Is it to, you know, to take accountability in case the decision is wrong, you know, things like that?
[00:13:16] Megan Antonelli: Yeah, well, and as you said before, I mean, that's some of the acceptance and the excitement around AI has been that they, the clinicians, really do feel like it will help them, right? And so in in a world where we have, you know, so much, both burnout and job dissatisfaction and sort of just general, you know, unhappiness within that system, you know, AI does have the potential. But what are you seeing in terms of the actual implementations, you know? I mean, yeah, you know, are there? Are there, you know, real elements of this?
[00:13:53] Yan Chow: Yeah, there's some interesting stuff going on. I mean, physicians, nurses, spend anywhere, depending on scenario, from 20 to 40% of their time on administrative tasks, you know, things that they hate doing, basically, data entry, you know, comparing data, verifying data, checking the check boxes. And so in my travels, you know, talking to different organizations, anything to relieve burnout, usually gets in here and gets a hearing, a real hearing. And one of the, you know, besides the typical thing we think of as automation, for instance, summarizing your patient and so that you can, you can do your summaries and your referrals really fast, or looking at emails to discern content and context, so that you can respond to them very quickly and appropriately and not be tired and do the wrong thing. One of the newest things that's happening, and actually the VA put up an RFP for this, is ambient recording using AI. Ambient recording is really interesting because it's the idea of a physician walking to a room, talking to the patient in a patient. Full attention to the patient, and you have passive monitoring on the side, and the patient shakes hands. Thank you, Mr. Smith, I'll see you next week. Walks out immediately. There's a structured summary, and in fact, with later iterations, you could see even that AI could make recommendations during the visit, maybe on a screen, on a wall, things like that. But this is very exciting, because it's almost like getting us back to the old model, you know, where the patient and a doctor relationship, the human element is the most important. And we're finding that the human element actually is important to help the patient get better. When they trust a physician, turns out, they take the medicine. So it makes sense, kind of they'll they'll come in for the follow up. If they don't trust the physician. Physician doesn't speak the language, they don't understand what's going on, they won't come back. And then, you know, the estimate is 10% of our hospitalizations in the US are due to non compliance, which is ridiculous. It's a huge number. So you can address that in any way. This can be better for everybody, including all the quality folks
[00:16:06] Megan Antonelli: well. And again, it goes to that, that L, you know, in some ways, low hanging fruit, but also, like, what was not possible before, right? And so when I think about that, I think about like, this, the social determinants of health, you know, another thing that there was a sort of a huge hype cycle for like, Oh, we're going to use all this data. This is so important. Then the reality of it came back, and the doctor said, How am I going to use all this data? How am I going to collect all this data? Am I going to, you know, but, but years ago, yes, a physician would have known, you know, that Mr. Smith lived here was, you know, married to so and so, for how many years, you know, had two children and lived on a farm or lived in the city, you know, would know all of those things and have that at their disposal, but now AI could make that possible? Yeah, yeah. So it's interesting to think about, you know, what's possible with that and the actions that you know could come from it.
[00:16:59] Yan Chow: You know, the city across the bay from where I live, San Francisco has over 150 languages, and that, you think about it, today's doctor. They don't know who will walk in. You know they will, let's say they're a Chinese doctor like myself. And so it could be Hispanic. Could be a South America. South America. It could be Latin American. Could be a European and African. How are you supposed to deal with all the different cultures and cultural beliefs? That's something where AI would be superb at doing. And so I mentioned the after visit summary. I asked Gen AI to generate a patient specific after visit summary for a for Hispanic patient with asthma, and it came up with things that, you know, be careful when you eat salsa and chili, because they can trigger asthma. Be careful when you when you play soccer, you know, not baseball, but soccer and so on and so forth. Those kind of things make, what I would think make a lot more sense into people in certain with certain backgrounds that they could relate to. And they'll go, oh, this doctor gets me, you know, I think I'll, I'll trust this guy, you know. And that's something that it's very hard to do that in five minutes, you know, when you only have 510, minutes to see a patient, at least the leave behind says that you know them. You know they you know the culture you would ask about, you know, AI could prompt you to ask about home remedies, which are very common in Hispanic culture. Could ask about, you know, the kind of where they live, you know, does it trigger asthma, things like that, those kind of things you mentioned, the structural determinants of healthcare. Those are the things that we sometimes have access to that data, but I'm sure it would be very helpful, and if we can get more access to that data,
[00:18:42] Megan Antonelli: right? Yeah, no. And it is, I mean, you know, not to say that, you know, it's entirely algorithmic, but there are certainly things that would be, you know, easily trainable for sure, in terms of things that will work, and recommendations and and that type of thing. So, you know. And it is, it's the human element, but it's also that understanding, it's the empathy, it's those types of things that really make it, you know, there's just, there's so much possible, there's so much possibility around the you know. So we've talked a lot about revenue cycle management and the ability for AI and automation to really, you know, save a lot of money for health systems and generate, you know, generate revenue as we get to some of these softer, or more human element elements of care. And you're speaking with, you know, health system leaders about adoption and implementation. Where are they looking to find the value, you know, from a financial side, and where are you seeing it
[00:19:46] Yan Chow: from the revenue cycle perspective? You know, it's been a fairly easy argument, because even even a study we commissioned as it was vendor agnostic, people were seeing, you know, 500 executives from many different. Industries, we're seeing that the return was in the multiples, which is really unusual. I mean, you know, 500% 600% that's to go from nothing to something. So that's the first step. That's why it's such a large, large return. Now people realize that, yes, we should use automation across the enterprise, right? In fact, automation could be used to redesign healthcare, but they don't know where to start. I mean, it's hard, because every healthcare organization is different, right? Everybody has different specialties. Diabetes clinic like Joslin is very different from a cancer center, you know, things like that. And so where do you start? How do you do it? And we're finding that one of the main challenges today is scalability. How do you scale a program that's may have started in rev cycle or some other area, some business unit, and you take those learnings, and how do you persuade other business units you know who are not seeing that example in their in their area, to to consider automation, and in a sense, is a, is a question in innovation change management, which I was an innovation for many years. It's, it's well known to a lot of innovators, how do you persuade, you know, how do you find the early champions who can, who can work with you and are willing to work with you, so you can start to develop the proof points and that kind of thing. How do you get the funding? How do you how do you do it? Because you scale, you're talking about the strategic level. That's sea level. And it's not a, you know, business unit problem anymore. Is actually a vision for the company and and sometimes we have that, sometimes we don't, healthcare is still really early, you know, they see the immediate return for revenue cycle. But there's a lot of potential for regulatory compliance, for burnout reduction, improving the jobs for 1000s of clinicians, for improving the patient experience. You know, from as I mentioned, this is so much, so much potential, and lot of it's not measured in dollars and cents, but indirectly, indirectly. So if you reduce patient readmissions to hospital because they're more compliant, you won't get penalized. You won't get penalized by CMS, you know. And you you may get a better quality score, which will draw more patients into your hospital and things like that, you know. So working that out, developing the business value framework for that is something we're focusing on, and that seems to resonate with the senior leadership, but I think we need to do more. We need to really think of healthcare as a, you know, as something that is now with automation and AI can be changed. In fact, we are still practicing, in some cases, 19th century medicine. And, you know, it's interesting, right? So the question, what do you train? What do you develop physicians for? You know? I mean, when you have automation, you have something doing the work for you. You're not doing the work anymore. So your skills should be managerial, in sense, and it's not autonomous. I'm the doctor. I lead the team. I do this, you know? So it's a very, it's a really threatening mindset, actually,
[00:23:04] Megan Antonelli: well, yeah, but at the same time, I mean, it is, you know, it goes to that ability to rethink and redesign, you know, what was wrong. And I think, you know, it makes me think about, you know, EHR implementation, and we sort of used this tool to sort of fix this, but, you know, then then engineered sort of the the misgivings of that tool into the operations, right? And we talk a lot about getting rid of the stupid stuff, and it goes back to, you know, literally, how do we model these models to do the things that should be done, as opposed to the things we've always been doing, right?
[00:23:43] Yan Chow: You know, it's interesting. Technology is evolving so fast. I've never seen technology evolve this fast, where science fiction has become reality at a pace that I can't keep up with. And so can you imagine the physician creating an avatar that looks and sounds just like him or her, that can advise patients, they can talk to patients, and conversational AI that can empathize, depending on how the patient looks on the screen, and so and so forth. That all exists today. And can you imagine five years How can a physician extend themselves to see 1000s of patients? You can't you know if you're just doing a human face to face now, you get a virtual reality headset, and you're all set
[00:24:26] Megan Antonelli: right, and, but yet, you know, I mean, and if we don't figure it out, you know, it will be done, I mean, and that's sort of, you know, if you look at sort of how slowly even the regulations around telehealth have been able to be rolled out, and in some cases not rolled out, you know, where, where there are barriers to that. But, you know, I know I can see on my Instagram account where people are starting to roll out their own avatars doing certain videos, you know. And they're not high quality, but it's happening, yeah, and it's certainly. You know, it's all possible. And if we don't, if healthcare, from an administrative and a regulatory perspective, can't keep up with that, it will happen anyway.
[00:25:11] Yan Chow: Yeah, And they'll always be the the early adopters of medicine that are going to do it, and they'll, they'll make the mistakes, and they'll make and they'll tease out the triumphs and things like that, so they'll define it. But I think there's quite a few organizations that want to be on the forefront, to be the leaders and to do the studies, and a lot of them are academic centers. They see this as can I be the Mayo Clinic of AI? Can I be the leader for the next generation of healthcare? And that's a very noble goal, but I think we haven't even seen the beginning of the potential or the threats, and so I think it'll be interesting next 510, years.
[00:25:49] Megan Antonelli: And it is, I mean, it also comes down to kind of the information, right? I mean, the way people absorb information. And if you think of, you know, the Mayo Clinic's information clinical tool, you know, sort of resources that there are online where you search for that, and you find those, and they're the most credible. And there was certainly a period of time where they always came to the top, but frankly, now there's so many sources that they don't come to the top. And, you know, and now with what younger generations are searching for that they want the video they want, you know, I always shut off the video. I'm like, I just want to read the directions. I don't want to watch the YouTube. But other, you know, my kids certainly don't want to read it. They want the video. So if we don't, you know, get to that point where it becomes accessible. You know, things that we didn't, you know, weren't possible before, because it was difficult, but it's, it's easy now and then, to go back to your point about translations, to have it in every language and and make that available to the patients is
[00:26:47] Yan Chow: one of the one of the really interesting possibilities is that now people are checking to building their own large language models based on their own data. So in other words, you could be a common spirit, you know, Boston Children's, and you have your unique data from 10 or 20 years, and you have your unique, unique patient, patient population, and your unique capabilities in terms of clinical skills, procedures, you know, technologies, things like that. And then you can actually do take that and use AI to optimize it. And that's something now is possible and within reach, and it was not possible before. So, for instance, the organization could find out, you know, what have been doing that, that, what have we been doing that's been holding us back? You know, what have been do? What? What are we good at? You know? How can we promote that? You know, and things like that, where, where you can have aI summarize data, all kinds of data, across your organization, and actually give you a picture of, you know, how, how you should proceed. And I think that's something that, you know, we used to have consultants do that for a lot of money. Now it's, it's actually going to be built on massive data troves that you actually keep and and I think that's going to be really interesting, because every organization that's a thought leader will start to use that to to really, for instance, here's an example. In June 2023 in nature, NYU Langone publish a study. They use Google's open source Bert LLM to build their own, you know, in house, you know, so there's no leakage of whatever in house. AI Gen AI to see they could improve their prediction for data, for database and prediction for morbidity and mortality in the hospital, who's going to die, who's going to leave, who's going to come back, you know, in a month. And they looked at their 10 years of data, and the interesting thing was that, and as and as you would expect it, it improved the predictability by anywhere from five to 15% which is huge, you know, to say this patient, you better pay attention to because they're the higher likelihood of dying, you know, in this hospital. But the interesting thing was that they implemented it by looking at the context of the physician's practice. Physician didn't have to enter any prompts, so no prompt engineering. What they did was they actually inferred from the context. This is a patient that could die, you know, and so on and so forth. So to make that prediction, to notify its position to all the stakeholders, to pay more attention, and so on, so forth. And so that's the kind of thing where, if you use your own data, you can actually analyze yourself and know yourself and know better, how to be better. And I think that's really intriguing. And, you know, we'll see customized AIS all over the place, and they'll be proprietary, you know, they won't be easily exportable, you know, because somebody else has different data, you know. So, of course, they're still subject to the same warnings about data. You know, they should be high quality data, very. Provide data, etc, etc, and not bias data. But that's really intriguing, because now they actually have a much better idea of how to allocate resources, where the strengths of weakness, where they need to address issues and so on. Is this? It's an interesting time.
[00:30:16] Megan Antonelli: It's fascinating. And I mean it, you know, it is. It's sort of, you know, starting with, you know, sort of predictive modeling, and the, you know, how people, how we use data from, you know, taking data from data to actual knowledge, right to the insights, to actionable insights that is possible you mentioned before. And I'd like to go back to it a little bit around this, just because now, when you're thinking about all of this, and all the sort of possibility of all this, and then the scalability of it, you mentioned sustainability briefly, and I'd love to hear a little bit about, you know, because I feel like that's a topic where there's certainly been some initiatives, some discussion around, you know, smart hospitals and green the greening of smart hospitals, and how do we ensure the sustainability of that? But when you add this whole element of AI, you know from your from where you're sitting and you're seeing that power and that you know the need for all of that, what is your advice to the health systems as they look at that. And where do you see the future of that going? I mean, how is it going to get managed?
[00:31:25] Yan Chow: Yeah, you know, AI computing, first of all, requires massive electricity, and it requires a lot of water, which is really interesting. Those two things are, you know, the focus of a lot of organizations in this world, apart from healthcare and apart from healthcare. So healthcare is kind of downstream from that, you know. So hospitals are using this resource we say they have to use. I mean, eventually the competitors will use it. They'll use it. So what's going to happen is that we'll have, we'll probably have regulatory, you know, control over the cost of computing on a natural on our natural resources. But on the other hand, technology is moving very fast, so I'm seeing almost every day developments of algorithms and systems that don't use much power, don't use as much power. So it's a moving target. It's a concern, and it's probably upstream from healthcare. But healthcare should keep it in mind, because healthcare is a big user. Potentially, we could have millions and millions, hundreds of millions of transactions in AI just trying to figure out what patients are doing was the best thing for them to do, and that is a huge cost to the to the ecology. So I think, I think there's still, like everything else in AI, still very early, but the early models are getting better, and so there are other issues besides natural resources. There's also things like, it's, as far as responsible AI, who's controlling those natural resources, you know, where they come from? Are we going to outsource them to other countries, you know? Are we going to be putting stuff on the moon? You know, there's so much weird stuff going on, who knows? Right? It's, it's like, and then the other worry I have about AI in general is digital divide, you know, there'll be those that have and those that don't have, you know. So like before, and then, so like the internet in the early days. So how are we going to address that? You know? And so the good thing is that AI for healthcare is basically going to be behind the scenes, I think, unless you're dealing with patients directly, so patients will notice it, but I think the main effect will be behind the scenes, and so it'll be in sense of invisible.
[00:33:51] Megan Antonelli: Yeah, yeah. No, absolutely. I mean, I think, you know, as all technology should be right to sort of create the seamless experience, and it should, it shouldn't be something that you feel too much in that well, you know, yeah, it's amazing, and it is amazing to think about that. But, as you said, with healthcare, it is, you know, sort of it'll get figured out up here, and then it'll be something to think about, but, but maybe not something that health systems have to worry about too much in that regard. But so as we have, you know, a few minutes left, you know, tell, tell me a little bit about what you see. It's been a crazy four years. You know, what does the next four years look like for healthcare? And if that's too far ahead, because things are moving too fast, we can just go with, you know, four to four to eight months,
[00:34:43] Yan Chow: four to eight days. Well, I think, I think AI powered your Intelligent Automation is here to stay, of course, but it's clearly just the beginning to to have the potential to cause a fundamental rewrite of the health. Healthcare landscape. And that's on top of, you know, the technology piece is great, but that's on top of regulatory, consumer trends, world chaos, you know, a lot of things are going to be affecting healthcare. So I see it as a new world where AI tools are helping workers to be more productive and to do more meaningful work, like we always say, also helping patients to more personalized and attentive care. But like any new technology, I think it'll bring a lot of challenges, as I mentioned, both inherent in the technology and external to the technology. And I think ultimately, like the internet, is going to be invisible. It's going to be something that we be part of our lives, you know. We'll be go shopping for clothes, and then we have phone hotels. I think you look better in the green, you know, which, okay. I mean, it'll be invisible, and we'll accept it, you know. And so we won't ask too much. How did you get that? Except in healthcare, how did you make that decision, you know. But on the other hand, I don't see any way around it, you know, because it's such it's got such potential. So it's going to be invisible, but it's going to definitely change our lives. And so it's going to be exciting next four months,
[00:36:15] Megan Antonelli: And probably even four days, it is. It's amazing. And I think the you know, sort of that, that what's possible, coupled with, you know, a little bit of, you know, Healthcare's reticence for for moving forward too fast and managing the risk while we do it creates, you know, incredible opportunity for for improvements in healthcare, which is what we're what we're all about Yeah, well, thank you so much for being here again. Dr, Chow, it's always a pleasure.
[00:36:47] Yan Chow: It was a pleasure for me as well. Yeah,
[00:36:48] Megan Antonelli: I look forward to chatting again and seeing you soon at an upcoming health impact. Thanks everyone for joining us on health impact, digital health talks and tune in next time and subscribe below.
[00:37:03] VO: Thank you for joining us for this week's HealthIMPACT's Digital Health Talk. Don't miss another podcast. Subscribe@digitalhealthtalks.com. And to join us at our next face to face event, visit healthimpactlive.com