Latest / CXOTalk / AI in Healthcare: Reality, ROI, and Patient Outcomes | CXOTalk #874
Transcript
- 0:00We're exploring AI and healthcare today on CXO talk
- 0:04#874 with three experts shaping this future.
- 0:11Doctor David Bray is distinguished chair of the
- 0:15accelerator at the Stimson Center, Teresa Carlson is
- 0:19president of the General Catalyst Institute, and Doctor
- 0:23David Rich is chief clinical officer of Mount Sinai Health
- 0:28System and president of the Mount Sinai Hospital in New
- 0:32York. We're going to explore what's
- 0:34real, what's hype, and how AI impacts decisions, data and
- 0:41patient outcomes. So, thoughts on solutions and
- 0:45how AI can support the overall healthcare system?
- 0:50The answer is to really analyze from the perspective of the
- 0:55providers what is a true return on investment.
- 0:58I have many people in the startup world who are often
- 1:01approaching and presenting me with the problem that they can
- 1:04solve, but I don't actually have that problem.
- 1:07And so they're trying to convince me I do.
- 1:10And there are many problems I have that they they don't really
- 1:13work with. And the, the other thing, of
- 1:15course, is that people in the early stages want to work with
- 1:18the health system. And they say, well, we're, we're
- 1:19with you for free. Well, free isn't free because my
- 1:22friends in information technology are, you know, very,
- 1:27you know, upset with me as as a clinician.
- 1:29I come with them and I say, look at this bright shiny object and
- 1:32they try to explain to me the amount of effort that would go
- 1:35into integrating it into the IT workflow, let alone the clinical
- 1:40workflow. So I'll stop there for a moment.
- 1:42I want to amplify what what David was saying.
- 1:44It's bringing back a lot of flashbacks and a past life.
- 1:47I was with the bio terrorist prepares response program.
- 1:49This was back in the early 2000s to 2005 and we were trying to
- 1:53plug into hospital systems, state systems and things like
- 1:57that to basically monitor. And as David so well put, the
- 2:01irony of electronic health records is they've actually
- 2:02created more human labor to complete.
- 2:05And, and I think also that's what David was saying, any good
- 2:09solution in this space has to fit into the workflow that's
- 2:12already being done by clinicians.
- 2:14Now there's a bit of a catch 22 because technology also changes
- 2:17are the possible. And so it's really you have to
- 2:19involve clinicians and then to, and I'll give a nod to what
- 2:22Theresa's talking about. We're now in a place because
- 2:26technology is changing what's possible so quickly doesn't mean
- 2:29it's being adopted, but it's changing what's possible so
- 2:31quickly. Policy makers almost have to
- 2:34iteratively learn also from what's even possible in the
- 2:36technology, both to remove barriers, but also to put in
- 2:40place incentives. If we're going to move forward
- 2:42and think about third and 4th order effects almost in real
- 2:45time, which makes this a really hard space to find solutions,
- 2:48it's doable, but it's a very hard space.
- 2:51David, so right, don't bring a solution to an area there's no
- 2:54problem. And I think that should be the
- 2:57number one use case of the companies that are coming down.
- 3:01What problem are you actually solving in?
- 3:02And there's got to be an ROI because you can't add more cost.
- 3:06Our job should be able to make our clinicians lives easier, our
- 3:09physicians be able to service more of these patients faster,
- 3:13more effectively and provide them the data that they need to
- 3:17get their job done more rapidly. And for us, I would tell you
- 3:21it's, it's really very much about ensuring that we're doing
- 3:25value based care and outcomes based care.
- 3:27And I'll just share one example. We have a couple of companies
- 3:31that we were with yesterday that we've invested in that are
- 3:33really thinking about patients that are really underserved
- 3:38populations. I'm from Kentucky, the hence the
- 3:41accent. I'm from a very rural part of
- 3:44the United States and one of our founders is focused on rule
- 3:48based care and ensuring that we're not only just training and
- 3:53teaching, but we're actually using technology to force
- 3:57multiply the capabilities that are there.
- 4:00Because a lot of times you have this specialist, you don't have
- 4:03the access to education or advocacy for these individuals.
- 4:06And they have individual and unique needs in their healthcare
- 4:09and they are usually the last to get it sometimes.
- 4:12And so she's thinking about a problem.
- 4:14And there's a lot of America that is in rural America, not
- 4:18cities or Urbans. And then the second part is for
- 4:23a company that we have called City Block that's taking care to
- 4:26the most underserved in there. There's a Medicare, Medicaid
- 4:30overlap with what they do that they literally do not get paid
- 4:35until the patient gets better. So it's a value outcomes based
- 4:39model and both of these companies, Homeward and City
- 4:42Block are doing this model that's very unusual.
- 4:45And when we talked to the legislators on Capitola about
- 4:48that yesterday, there's like, what?
- 4:50I didn't even know this existed. So, you know, one of the
- 4:53physicians from City Block Toy and who started this company
- 4:56said she's a physician herself. And she said, look, I could see
- 5:00a lot of patients during the day.
- 5:02I could give them a prescription and hope that they leave my
- 5:05office and go fill that prescription and they take it.
- 5:09But I can't ensure they're doing that.
- 5:11But I get paid for that visit regardless.
- 5:13And their model is much more about ensuring that there are
- 5:16outcomes from what they're actually prescribing and doing.
- 5:20And that's what I call a shared responsibility model between
- 5:23that physician and the patient. So we're really looking for
- 5:26creative ways to ensure we're getting healthier outcomes, just
- 5:29not charging the system more. David Rich, your thoughts on
- 5:33this kind of relationship between innovators and
- 5:41hospitals, hospitals, I mean you're president of of a major
- 5:45hospital. So I'm assuming you're dealing
- 5:47with these practical issues all the time.
- 5:49One of the the key. Things that we're.
- 5:51Working on Michael is a a seamless pipeline for sort of
- 5:56digesting the the multiple. Groups that are coming to us
- 5:59with fantastic ideas. General Catalyst being we know
- 6:02one of the the top groups that we we pay attention to.
- 6:06We regard anything that comes through an organization like
- 6:09General Catalyst is being highly it was something that's highly
- 6:13thought about and you know, it, it comes to us with a degree of
- 6:17how should we say it's, it's curated in a way that we should
- 6:21pay close attention. And as we do this, we try to
- 6:26break them down into the different buckets, you know, for
- 6:29example, is it? Something that would improve
- 6:31patient experience, Something that would improve workflow,
- 6:33Something that's focused on a patient outcome.
- 6:36And so that we can look at what our ROI is related.
- 6:41To this particular. Idea and ROI isn't always
- 6:45strictly financial because in our world we have penalties for
- 6:51failing to meet certain criteria.
- 6:54And so sometimes penalty avoidance is an ROI, sometimes
- 6:58better outcomes and shorter length of stay is an ROI because
- 7:01we can backfill with other patients.
- 7:04So it all comes down to whether the idea integrates into a sort
- 7:10of a category of ideas that make sense for us and is something
- 7:14where we can rationalize an ROI and sometimes partner and since
- 7:19we're an organization. That likes to think of.
- 7:22Ourselves as entrepreneurial potentially work with startups
- 7:27and other groups and put in in our case, what is often the
- 7:31sweat equity in helping develop and making a product
- 7:34commercially viable. Check out cxotalk.com, Subscribe
- 7:38to our newsletter, Join our community.
- 7:41David Bray I'm hearing a core issue here is alignment between
- 7:47groups that have somewhat overlapping agendas and goals,
- 7:52but also distinct. Is that a correct way of looking
- 7:55at it? And how do we bridge that gap?
- 7:57Yes, I think this would be the definition of what we would call
- 8:00a wicked problem where you have not only different folks aiming
- 8:05for different goals that sometimes overlap, sometimes
- 8:07don't, but the very goals and how you're going to pursue it
- 8:10are changing at the same time too.
- 8:11That's why, you know, everyone's like, there's a magic wand that
- 8:13will magically fix healthcare and solve it by the next hour.
- 8:16I think we're all we're all trying to say what you really
- 8:20can do is if you get the stakeholders together.
- 8:22And I think because things are moving so quickly and, and
- 8:25things are shifting, do three things, which is think about
- 8:28first and foremost, how does this improve the practice?
- 8:31And, and, and David Rich was very good at pointing out, think
- 8:33about ROI. And ROI is not just in terms of
- 8:35financial return, but giving back time to the clinicians,
- 8:39healthier outcomes for the patient in terms of getting them
- 8:42released earlier, and then also penalty avoidance.
- 8:45And so practice first and foremost.
- 8:47Then on top of that, then thinking about the new program
- 8:49and having some curation. So it's not just every shiny
- 8:52idea gets tried out because you just can't absorb that.
- 8:55And the organization would be stressed and it actually might
- 8:57disrupt up the organization more than help.
- 8:59And then finally the policy layer.
- 9:01And that's where Teresa was saying, because we have
- 9:03policies, you know, our laws don't have expiration,
- 9:06expiration dates usually. Sometimes they do, but usually
- 9:08there's no expiration dates. And so there's things still in
- 9:10the books from the 1940s, nineteen 50s and 1960s that may
- 9:13have made sense then that don't necessarily fit now.
- 9:17And I think what I would submit is 1 area in particular that I'm
- 9:20personally passionate about coming from a background with
- 9:22the People Center and Internet Coalition and, and what we're
- 9:24doing at the Simpson Center now, data, in my opinion, is a form
- 9:27of human voice. And I think we've got to think
- 9:30about instead of data being something that's opaque to both
- 9:34the patient but also the clinician, about where it came
- 9:38from, how it's informing and everything like that, we've got
- 9:41to think about how it gives both more visibility and stakeholder
- 9:44ISM to the clinicians as well as the patient because this is
- 9:48going to be helping to influence outcomes.
- 9:50And if you're in a rural state where maybe things aren't even
- 9:52digitized yet, they're not even accessible yet to help inform a
- 9:55clinician with your care, that's an unfair thing that we have to
- 9:58figure out how to dress. But at the same time, we don't
- 10:00want patients and clinicians to lose the ability to choose when
- 10:03and where that data is used. If we want to take advantage of
- 10:07these most transformative technology companies that we
- 10:10hope will help David in his practice and our patients.
- 10:13The one thing I will share is that we also can't have a
- 10:17patchwork, have a regulatory environment that doesn't enable
- 10:22this. And like is a good example I
- 10:24think is healthcare. But if you think about there
- 10:27could be a federal AI model and then fifty state models.
- 10:32And then if these companies work outside the US, they have
- 10:35another model that they have to adhere to.
- 10:37That is not if you're really it's an extension, it's an
- 10:41existential threat to these really transformative
- 10:44technology. And we want the doctors and
- 10:46physicians, the physicians and the nurses and the clinicians to
- 10:49have the best, best technology for what they're trying to
- 10:52achieve. But what we actually adhere to
- 10:54is we believe Healthcare is a really good example where you
- 10:58can allow the industry to take control of how they want that AI
- 11:02to be enabled. And I'll just give one quick
- 11:04example. The FDA has been doing their job
- 11:07a long time. Let's let them actually work
- 11:10toward what the right policy is for their environment for
- 11:13healthcare. And they're actually kind of
- 11:15already doing it. We have companies working in
- 11:18with FDAAI Doc was with us. They work directly on the
- 11:22physician to look for capabilities to allow the
- 11:27physician to see if there's medical errors to help them with
- 11:30that. And we know that medical errors
- 11:31is one of the largest cost that they're moving fast.
- 11:34They need technology to support their efforts that if we can let
- 11:38those industries kind of operate, then we can take this
- 11:41AI capability industry by industry with the policies and I
- 11:45think it will allow us to move a lot faster and let the experts
- 11:48who know that industry work on the right policies.
- 11:52There's an organization called the Consortium for Healthcare.
- 11:55AI Mount Sinai's involved all of the major health systems.
- 12:01And one of the. Aspects that we're all concerned
- 12:04about is that we can be regulated out of existence
- 12:07before we exist. And so, and it isn't just at the
- 12:11federal level and the state level.
- 12:13We also have Joint Commission, other certifications, we have
- 12:16local health boards as well. And So what we've done often is
- 12:22trying to come together as, as a group of, of healthcare
- 12:26organizations and bring all of the three letter federal groups,
- 12:30at least as a start, together with the concept that we might
- 12:34develop assurance labs, such that as the AI tools are
- 12:39developed, that we have a way of assuring their quality and
- 12:43assuring that they, you know, fit into the, I would say the,
- 12:47the culture as well as the, all the different matrix elements
- 12:52of, of working in complex health systems.
- 12:56And I think the, you know, one of the ways we get around it
- 12:59right now is often a lot of the tools that we've implemented,
- 13:02especially once we've developed internally for decision support,
- 13:05We try to think of bringing the right team to the right patient
- 13:08at the right time. We don't actually make the
- 13:10diagnosis, but we try to identify the patients who most
- 13:13likely have that diagnosis and then bring the clinicians there.
- 13:17We'd like to move beyond that at some point.
- 13:20But that's going to take a little help.
- 13:21We need to have a real partnership between government,
- 13:25industry and healthcare providers in order to be
- 13:27successful at that. We've had John Halamka as a
- 13:31guest several times on CXO Talk, who's one of the founders of
- 13:35Chai. Looks like an interesting
- 13:36organization. We have some questions that are
- 13:39coming in from LinkedIn and Twitter, so why don't we jump
- 13:42there? And this is a question from
- 13:46Arsalan Khan, and I'll toss it out to the three of you, whoever
- 13:49wants to grab this. Has anyone mapped the entire set
- 13:54of healthcare workflows? And what about having AI
- 13:58recommend which workflows are archaic?
- 14:03This is actually a needed space and there are different groups
- 14:07talking about how do you translate what is business
- 14:13process modeling in the healthcare domain into some way
- 14:16that can then be analyzed rigorously across different
- 14:20systems. There are different vendors that
- 14:22have tools for business process modeling in healthcare space,
- 14:24but they're each in some respects proprietary.
- 14:26And so if we can get to that and actually have almost a universal
- 14:31way of expressing business process modeling in healthcare
- 14:33space, we can actually do another thing too, which is we
- 14:35can go to the journals. And this is where I'll I'll give
- 14:38a nod to my other fellow colleague David to talk about.
- 14:42Sometimes the journals actually have peer reviewed articles that
- 14:45recommend different ways of achieving outcomes that might
- 14:48actually be different. But because it's right now in
- 14:51text as opposed to business process models that can do
- 14:53comparison, you actually can't compare and say which of these
- 14:57leads to better efficacy for the patient.
- 14:59So this is actually an area where with AI, if we can
- 15:02actually get to some way of not being locked into proprietary
- 15:06vendor systems, but instead have a way of having interoperability
- 15:09to express business process modeling across both hospital
- 15:13systems and IT systems. But then also translated out of
- 15:16the text journals in terms of what is the peer reviewed gold
- 15:18care, then you can actually start to do some very
- 15:21interesting things and actually achieve that.
- 15:24However, first we've got to figure out what's that place
- 15:26that encourages people to think more broadly than their own
- 15:28specific system, but think about holistically across systems of
- 15:32systems. I'd like to address the the word
- 15:35archaic. One of the.
- 15:37The jokes in medicine is that the way you learn is you.
- 15:40If you've done one case, you say in my experience, if you've done
- 15:452 cases, you say in my series, and if you say done 3 cases, you
- 15:48say in case. After case after case.
- 15:50So one of the problems that we have in medicine is this culture
- 15:56where experience that leads to best practices that are not
- 16:02always evidence based. And as we develop the evidence
- 16:05base in getting the physicians to change, getting the nurses to
- 16:09change, getting the transporters to change, everyone who's
- 16:12involved in care in a healthcare setting requires some sort of
- 16:18persuasion. The common term now is the nudge
- 16:22unit. And the nudge unit is intended
- 16:25to identify and to gently or maybe sponsor gently correct
- 16:30behaviors that are are headed in the wrong direction.
- 16:34Often those are things that are are wasteful either clinically
- 16:37or, you know, sometimes financially.
- 16:39Like a perfect example was University of Pennsylvania
- 16:44managed to switch prescribing behaviours to less expensive
- 16:49drugs and we're very successful and we see similar things at
- 16:53Mount Sinai where we've had some success.
- 16:55So I think going back to archaic, it regards the, the
- 17:00most important thing I take into account is we have to think
- 17:02about behavioral change and sometimes that'll be based upon
- 17:07behavioral economics. And so the behavioral economics
- 17:11perhaps are the the answer that I would give to starting to look
- 17:16at ways of overcoming archaic processes.
- 17:19We have another question that has come in on a related topic
- 17:23on the topic of interoperability.
- 17:26And this is on Twitter from Chris Peterson, who says we're
- 17:3030 plus years into healthcare interoperability standards, but
- 17:35we're still not there in many respects.
- 17:40And so how does this lack of standards affect the advancement
- 17:46of AI in healthcare? We do hear a lot from our
- 17:51companies that is that it is because they are so creative.
- 17:57I mean these entrepreneurs, I can't stress, they are trying to
- 18:00solve a problem, not just look at a problem and jump over it
- 18:04and step in it. I mean, they're really trying to
- 18:06solve a problem. But I would tell you that they
- 18:09do say interoperability is, is a top challenge for sure.
- 18:14But I think what we're also seeing, Michael, are creative
- 18:18ways that they are trying to work around it and look for new
- 18:20ideas, even starting from scratch on some things, which I
- 18:24know nobody wants to do, but they're like, OK, we got to
- 18:27solve it. Not just keep talking about the
- 18:29problem, but look for creative solutions.
- 18:31And I will tell you AI, because the Internet in these data sets,
- 18:37these large language models are getting so good that you're
- 18:41almost now beginning to create new systems of record without
- 18:46even knowing it's being created, right?
- 18:50So you're seeing these ideas. And also now they're beginning
- 18:52to access systems that are there in a way that's creating new
- 18:58ideas to look at that data and look for solutions.
- 19:03So while it would be we, we talked about freeing the data a
- 19:06lot and I think we need to free that data.
- 19:09Data needs to be free, you know, our physician David needs to be
- 19:15able to take advantage of that in every aspect of his practice.
- 19:18He should never not have access to the information that he needs
- 19:22at his fingertips to treat a patient.
- 19:24On the more research side or Doctor David Bray, we want our
- 19:29researchers and our new, the entrepreneurs creating new
- 19:32technologies to have those capabilities that they can bring
- 19:35back to that clinical work. So I would say whatever we do in
- 19:39this country and around the world, we've got to make sure
- 19:41that we have open and accessible data.
- 19:44And just by the way, not just for the clinician, but for me,
- 19:48the patient, I want 100% access. I want to make decisions about
- 19:52my own data. I think you're seeing consumers
- 19:54get a lot smarter, so we need systems that are open and
- 19:58interoperable in order for the patient themselves to make the
- 20:02right decision about their care. David Bray, let me ask you a
- 20:07question then, as Teresa was just describing.
- 20:11Well, as David Rich and Teresa were just saying,
- 20:15interoperability is very important and Teresa was just
- 20:19describing the need for flowing, the free flowing of data in
- 20:24order to support innovation. But on the other side, the
- 20:29economic incentives for healthcare information systems
- 20:34in many cases is to keep that data because he or she who holds
- 20:40the data holds the money. So what about that?
- 20:43I think we need to unlearn the meme from about a decade ago
- 20:46that somehow data was the new oil.
- 20:49Oil, use it up, it's gone data, use it, it's still there.
- 20:52And in fact, if you involve people associated with the data,
- 20:54whether it be a patient or a clinician, they'll find things
- 20:58wrong in the data, they'll fix it, they'll make it better,
- 20:59they'll create more data, they'll create better quality
- 21:01data. So in some respects we need a
- 21:03different metaphor than data's new oil.
- 21:06I also think we need to simplify the consent mechanisms, but also
- 21:11the transparency. So a patient can see where is my
- 21:14data going because right now you fill some forms, but they're
- 21:17paper based forms usually and you have no idea where it's
- 21:19going to go. And then similarly, a clinician,
- 21:21as Teresa says, needs to be able to get access to that data to to
- 21:25do what they need to do. And so there are technically
- 21:27policies that say you're not supposed to do information
- 21:29blocking and, and, and it, this is one of the sort of tensions
- 21:33of the US system given that we have a free market, which is
- 21:36great, but we also then are trying to think about how we
- 21:38overcome individual self-interest to.
- 21:41Hoard the last thing I would say on the interoperability
- 21:44challenge, Michael, I often use the phrase that standards.
- 21:47Are like toothbrushes. Everybody says they want to use
- 21:50one, they just don't want to use anybody else's.
- 21:52And so you have to be careful because not all standards are
- 21:56created equal. Some standards are really
- 21:58vendors trying to build a Moat. Sadly, it's it's standards that
- 22:02are propped up by them. But other standards groups
- 22:04really are about trying to level the playing field, whether it's
- 22:07for established companies, startups and the like.
- 22:11And so usually I look for, do they charge or not to use that
- 22:14standard? I think the last thing I would
- 22:16say is also technology is, is advancing just what's even
- 22:19possible in medical medicine. And we've gone from about 20
- 22:23years ago, there were 12,000 CPT codes for claims.
- 22:27Now we're up to 72,000 CPT codes for claims.
- 22:31It's getting where it's really difficult for a human to know
- 22:34which claim code to use, whether it's an experience coder, let
- 22:38alone a patient trying to figure out when they actually are
- 22:40looking at their bill. So this may actually be
- 22:42something where we need to think about a new strategy, almost
- 22:44like what Theresa was saying. What would it look like if you
- 22:47had your own personal chat bot to help you understand your
- 22:50bill? If you got your bill and or pre
- 22:52authorization to navigate what is a very complicated standards
- 22:56based process and how do you make it more accessible to
- 22:59everybody, not just someone who's a technological expert.
- 23:03We have a question from LinkedIn from Ravi Karkara who is the Co
- 23:10founder of the AI for Food global initiative and David
- 23:16Risch. He says this what AI skills will
- 23:20be needed for medical and nursing schools to teach in
- 23:25order to create this AI medical revolution?
- 23:29We don't teach so many things in medical and nursing.
- 23:32School, other than perhaps the traditional things that emerge
- 23:38from anatomy, Physiology, mythology.
- 23:43And we need to teach people about systems based practice and
- 23:48that it isn't just AI. The whole point is technology is
- 23:53here. I would argue that Healthcare is
- 23:55a technological industry that has failed to recognize it for
- 23:58decades, and so I couldn't agree more.
- 24:02With the individual that asks the question that as we reform
- 24:06the curricula, it is about teaching people how to best
- 24:10integrate with technology, with. The proviso that you have to
- 24:15teach also that AI and other technologies that it's not a
- 24:20threat, it's actually a way to augment the way you deliver care
- 24:24and. I have a friend who's chair of
- 24:26radiology here at Mount Sinai and he says that AI won't
- 24:29replace radiologists, but radiologists that use AI will
- 24:32replace. This radiologist that do not.
- 24:34So I'll leave it at that. I just want to remind everybody
- 24:38that you can ask your questions on Twitter, Use the hashtag CXO
- 24:44talk if you're watching on LinkedIn, pop your question into
- 24:47the chat. And truly, when else will you
- 24:51have the chance to ask such an esteemed group of people?
- 24:55Pretty much whatever you want. So I hope you take advantage of
- 24:57it. And this is from again from
- 25:01Arsalan Khan who comes back and he says regarding healthcare
- 25:05data, do we need more healthcare data or less?
- 25:08And what can the government do at the federal policy level to
- 25:14help this data and interoperability issue?
- 25:18I've been in the world, I've been working in the public
- 25:20sector world for over 28 years and I'm used to a lot of data.
- 25:24So I've seen it just continue. So I so I don't know that we
- 25:28have a limit on the data anymore.
- 25:31And artificial intelligence is really a key solution for us to
- 25:34be able to enable and understand that data, research on it, get
- 25:39immediacy from it. So I don't, I don't think data,
- 25:42I think I would be personally OK.
- 25:44I think as much as we need, it's OK because we're going to keep
- 25:47building on it. So I don't know that there's a
- 25:49limit when you're working with folks like the intelligence
- 25:52community and defense and you see over the years the amount of
- 25:55data in healthcare, I think we're OK.
- 25:58But to your point, what are the policies?
- 26:01I would tell you we need to educate a lot more our policy
- 26:05makers. I do think they want to
- 26:08understand, but I don't know that we've done a good job
- 26:10actually explaining to them that if the data is not free flowing
- 26:16and there's not accessibility to it, what's the harm that it
- 26:19does? Why does it inhibit our ability
- 26:23to make America healthy again? If that is the goal, if that is
- 26:27this administration's goal, what does it do?
- 26:29So I see our job as an example at the institute working with
- 26:33partners like this to go in with examples.
- 26:36And, and to your point, David's point earlier, there are
- 26:40policies in place that are supposed to make sure that the
- 26:44data is open and free flowing. But for some reason we hear from
- 26:48a lot of our companies, a lot of big and small companies, it's
- 26:51just not working. And not just them, we hear from
- 26:53the physicians and the clinicians and the hospitals.
- 26:56It's just not working the way it's supposed to work.
- 26:59So when we need to figure out why that is, and then we need to
- 27:02ensure that the policies that are in place already are being
- 27:06activated, activated, and somebody is watching out for the
- 27:11patient to make sure that that data is open and available to
- 27:15them. The clinicians and the hospital
- 27:18systems and the vendors and partners that actually have to
- 27:21have it to make America great again.
- 27:25And healthy, sorry, healthy. We're already great.
- 27:28I think America's already great. Let's make it health.
- 27:30And Michael, if I can build on what Teresa said, 'cause I want
- 27:33to give a nod, General Catalyst, the General Catalyst Institute
- 27:36put out a report. And what was interesting is they
- 27:39were talking about what Teresa mentioned earlier with some of
- 27:41the examples shifting from right now, what is a service where you
- 27:45get paid a fee, but doesn't necessarily mean that the
- 27:47patient followed through with the healthcare advice, followed
- 27:49through, didn't get actually get healthier.
- 27:51Shifting from that to value based care and outcome based
- 27:54care. And I would submit this is where
- 27:56the data is so crucial because if you're going to do value
- 28:00based care and outcome based care, you almost first have to
- 28:03understand the the sort of the long, the long tail of that
- 28:07patient and understand are they pre diabetic?
- 28:09And therefore, if you get involved now you can avoid some
- 28:12of the more costly things that they actually become diabetic.
- 28:15But that's one of those things again, where you know, it's both
- 28:19as what what David Rich was saying about how do you help fit
- 28:24into the existing workflow for clinicians?
- 28:26Because if you give them a 40 or 50 year olds patient record,
- 28:30they don't have time to wade through it all to make sure they
- 28:33catch everything. And so how can you give them
- 28:35confidence if they're working with an AI augmented tool, it's
- 28:39not replacing them, but augmenting them confidence that
- 28:42the AI has brought the relevant things to help them inform that.
- 28:45And then when you. Teach.
- 28:47Future generations of doctors and nurses, how do you make it
- 28:51so it's not just about the individual clinician, but part
- 28:54of a larger system to make sure that hand off for all the people
- 28:57that are providing that value based, outcome based care are
- 29:00thinking about that patient's neutrality.
- 29:02And so that's a major revolution to try and fit into the next 5
- 29:06to 10 years that's already got a system, that's already got
- 29:09strains. But you know, that's what also
- 29:11makes this space very interesting.
- 29:13And that's why we need to have these conversations now about
- 29:16let's make sure we don't lose sight about empowering the
- 29:18patient and then also making sure the clinicians have the
- 29:21tools at their hand to actually help with that.
- 29:23When we talk about maximizing physical responsibility in U.S.
- 29:28healthcare, the only other point that I want to make here, if we
- 29:32want to cut the red tape out and do that, we don't want to make
- 29:35it more expensive to get to the data and be able to access it to
- 29:41get those outcomes. Because again, let's remember,
- 29:44AI and technology over time should drive the cost way down,
- 29:48right? We should think about, I know
- 29:49people talk about it driving it should drive the cost down
- 29:52because you can do it faster, more efficient, get to the
- 29:56outcomes. And then let's just not just
- 29:59less expensive, but then you get to the results faster.
- 30:03So that allows our clinicians to be able to make decision, the
- 30:06decision making much faster. So I just wanted to say that
- 30:08because we we talk a lot this this administration is very
- 30:12focused on cost cutting. So we can't make it more
- 30:16expensive. And that's why technology should
- 30:18be the enabler both to move faster, get the resourcing there
- 30:22and again, don't make getting to the data more expensive.
- 30:25Let. Me give a couple of practical
- 30:27examples if I could that I think illustrate the the points we're
- 30:31just making About four years ago we started maybe the longer we
- 30:36started working on a predictive algorithm for delirium in
- 30:39hospitals, which is a big problem and the structured data
- 30:43were not very helpful didn't did not end up with a predictive
- 30:47model that was you know all that impressive.
- 30:50But as soon as we added natural language processing to the
- 30:53unstructured data in Physician and Nursing notes, the model
- 30:57became. Predictive.
- 30:59Almost scary predictive in terms of how accurate it was.
- 31:02And so I think that as someone who was working with Snow Med
- 31:06and Itsy Doo years ago, trying to establish standards and then
- 31:09watching, watching the moats as David Bray described being built
- 31:14by the different electronic health record providers, at
- 31:17first I thought, oh, this is just the end of standardization.
- 31:20We'll never have interoperability.
- 31:22But now I'm starting to. See that the tools, especially
- 31:25large language models that are able to create.
- 31:28Some summarizations of massive data sets be very important.
- 31:31I walked into the OR this morning and the patient had had
- 31:36five previous cardiac operations.
- 31:38I'm a cardiac anesthesiologist. Just imagine trying to summarize
- 31:42that data. It's challenging, but honestly,
- 31:45if a large language model had been available to me and had
- 31:48given me a 5 paragraph summary of the most important
- 31:52unstructured data in all of those records, it would have
- 31:55made my job much simpler. I would not have spent 45
- 31:59minutes last night pouring through electronic records.
- 32:01I think those are practical examples of how we've lost the
- 32:05battle in terms of having the data perfectly structured.
- 32:08But with interoperability and using newer tools, I think we
- 32:12can overcome those sins of the past.
- 32:15We'll say what I find fascinating is that we even have
- 32:19to have a discussion where we remind ourselves that we have to
- 32:23put the patient first. It's just, there's so many
- 32:27different stakeholders in this system.
- 32:29And I do think it is a, you know, again, because we separate
- 32:32our, our, our public sector from our private sector, which is a
- 32:34strength. But it does mean you have to
- 32:36think about, you know, how do you empower the clinicians,
- 32:40which are plural, which are working across different IT
- 32:42systems that are provided by different vendors.
- 32:45And then also the fact that the patient has the ability to
- 32:48choose in some cases where they go, and they may go across state
- 32:51lines. And for some mysterious reason
- 32:52in the Constitution, because nobody mentioned who got to
- 32:56oversee healthcare, it defaults to being a state right unless
- 32:59the federal government uses some preemptive clauses.
- 33:01And so I actually think in some respects I liked what David Rich
- 33:05was saying because I also have those battle scars from the
- 33:07standards battles about 20 years ago, which sadly, there were
- 33:11people that were trying to, to, to create a standard identifier
- 33:15for everything in the known universe.
- 33:16And I would submit that is never going to be doable.
- 33:19But now with using unstructured data and using new approaches
- 33:23that can actually say I don't have to have everything
- 33:26specifically specified, but I can help the clinicians with
- 33:29understanding whether it's through a summary of the text.
- 33:32But could also be using computer vision to go through 1000
- 33:35computer images and actually say these are the ones you should
- 33:37pay attention to. I think if anything, it may be
- 33:40we're in a world that's drowning in data but lacking in insights.
- 33:44And if you can give the insights back to the clinicians and the
- 33:47patient, that's how we're going to get through this.
- 33:49We have an interesting question from LinkedIn from Greg Walters
- 33:56around this set of issues. Before I take that question, I
- 34:00just want to mention to everybody who's watching.
- 34:02You guys are an amazing audience.
- 34:04You ask incredible questions and we want you to join the CXO Talk
- 34:08community. So just subscribe to our
- 34:10newsletter, go to cxotalk.com so we can tell you about our
- 34:13upcoming shows because we have incredible shows coming up,
- 34:16really awesome ones. So Greg Walter says this, it's,
- 34:20he says that he believes AI can be the glue in healthcare, not
- 34:27only accessing old data, but creating and tracking new data
- 34:32and therefore could being that being that connective tissue.
- 34:36And he wants to know what's the level of investment dollars and
- 34:41the commitment to the, regarding this mission of AI that the
- 34:45healthcare industry is currently putting towards these
- 34:50technologies and what is the, what is the engagement level as
- 34:54well? We look at ourselves using
- 34:58standard surveys, I believe Gartner did one for us.
- 35:01We, for example, we look at how much we spend on cybersecurity
- 35:04versus the banking industry and it's much less surprisingly and,
- 35:09and especially with the challenges in healthcare, we're
- 35:12an industry with very low margin.
- 35:15And in fact, with the downward pressures that we're going to
- 35:18see related to, you know, a desire on the part of the
- 35:22administration to reduce healthcare spending, that we
- 35:25really wonder if we're going to be able to afford to stay up to
- 35:30the level of spending that other industries have related to
- 35:34information technology writ large.
- 35:37And so the answer I think has to be that we always, when we look
- 35:42at the ROI on a tool, have to see a way that it actually
- 35:45reduces some aspect of the expense.
- 35:49And so for example, finding tools that are better able to
- 35:54engage patients as partners in their own care.
- 35:58Truly personalizing medicine isn't just looking at a genomic
- 36:03markers and biomarkers and environmental factors in
- 36:06deciding what's best for that patient.
- 36:09You also have to analyze something that we've never
- 36:12really looked at, which is the receptiveness of the individual
- 36:16or the family to actually being a.
- 36:18Partner in that. That care so as we move from,
- 36:22you know, good old fashioned, you know, case after case after
- 36:25case that I referred to previously towards population
- 36:29health, towards personalized medicine, we have to really
- 36:33start to bring it down not just to the level of the provider and
- 36:38the patient, but the patient and their desire and interest to
- 36:43interact with the provider. Making America healthy again is
- 36:46actually a partnership and. Michael, if you look at venture
- 36:50capital, venture capital's investing, I think alone about
- 36:5311 billion this year in AI, taking a look for healthcare and
- 36:58about, you know, could be small in a $4.5 trillion health
- 37:02market, but it's getting there and we believe in it.
- 37:05And I will tell you our CEO Hemantineja has been investing
- 37:09in healthcare, I mean for so long, he's a believer, He's
- 37:11created companies. We we believe in this model.
- 37:16So much of AI transformation in healthcare, we have a whole
- 37:20group just dedicated to this. And the other thing I'll share
- 37:23is we are seeing on the healthcare side, the six CE OS
- 37:29and founders we just had up here yesterday, their business is
- 37:32growing at a really rapid rate, all focused on AI.
- 37:37And we're seeing the healthcare industry that the payers and
- 37:40providers really lean in because they see the opportunity.
- 37:45So I expect this, the AI efforts in healthcare to grow
- 37:51exponentially. And I'll just give you one
- 37:53example that we've been talking to this administration about,
- 37:56and we put in our white paper that we just put out a week ago,
- 38:00which we advocate for sandboxes where we can allow healthcare
- 38:05institutions in the US, federal government and states to see the
- 38:09value. So they try out, they realize
- 38:13how, you know, they need to understand it.
- 38:15And we're a big believer in experimentation because we don't
- 38:18want back to Doctor David in D's point.
- 38:22We don't want people just selling him stuff.
- 38:25He should see the value in a technology before he adopts it.
- 38:30So he should have the opportunity to try it out in the
- 38:32sandbox. Because he may say to his
- 38:35credit, he's a smart guy. He works every day very hard
- 38:38with. You might say, I'm not sure, but
- 38:39I'd like to try it. So we need ways to allow for our
- 38:44clinicians, healthcare providers to try out tools and show the
- 38:48value quickly and then they can acquire it and scale it.
- 38:51And I think in today's world of cloud and AI, it's not a license
- 38:55based world. You should be able to try things
- 38:57out and then scale it. David Rich, how do you measure
- 39:02the clinical value, clinical impact of tools such as Teresa
- 39:09was just describing and, and the outcomes, the impact on
- 39:13patients? Well, it really comes back to
- 39:15the, you know, the, the many ways that ROI can be calculated.
- 39:20A simple one for a very complex tertiary or quaternary care
- 39:24hospital is can I assure a better outcome, a lower
- 39:28mortality, a lower infection rate?
- 39:31Can I empty a bed more quickly and fill it with another patient
- 39:34that generates another DRG? Just at the simplest level, at
- 39:38the population health level, something that we have not
- 39:42really been successful at at least much less so on these East
- 39:46Coast a little bit. Better on the West Coast is
- 39:48value based care. So if I had a panel of patients
- 39:52and they really had a narrow network and they really had to
- 39:55use our health system for their care and I'd better manage let's
- 39:59say their diabetes, their hypertension, etcetera.
- 40:02You know, can I reduce their spend?
- 40:04Can I truly bend the healthcare cost curve?
- 40:08When you have. A wide open sort of rodeo here
- 40:13in the United States where if I'm not happy with Mount Sinai
- 40:17Health System, I can go to the health system a mile down, you
- 40:21know, 2nd Ave. Or 3rd Ave. and go to a
- 40:23different 1. You know, we don't really have
- 40:25the ability. To to really model and measure
- 40:28at that level. So it's, you know, really to
- 40:32answer your question a bit, Teresa and Michael, it's, it's
- 40:36such a mess because we don't really control the, you know,
- 40:41the, the spend and we don't have a good way of measuring the
- 40:44outcomes in an environment where people can get their care
- 40:48wherever they. Choose.
- 40:50And that's part of what we value as Americans is freedom of
- 40:53choice. But if you go to Europe, then,
- 40:56you know, they've taken a different approach with National
- 40:58Health insurance and they've created, you know, sort of a
- 41:01captive market. And they can for, you know,
- 41:05it's. Not a word that we'd like to
- 41:07hear, but they they ration care by deciding what's in the
- 41:10national basket, what drugs will be provided.
- 41:14What service will will be provided and can apply standards
- 41:18that are not really acceptable in in US culture.
- 41:22So I think we have to also think about these cultural barriers to
- 41:27bending the healthcare cost curve.
- 41:29Doctor Faith Mehmet Ghoul, who is CEO of The View Hospital in
- 41:34Doha. He says he appreciates the
- 41:38valuable medical examples shared for AI applications during this
- 41:42discussion. However, we should not overlook
- 41:44the significant potential of AI in non medical areas within
- 41:49healthcare as we continue to face challenges around resource
- 41:53allocation and affordability. AI can play a transformative
- 41:57role in driving efficiency and improving access to care.
- 42:02We know, for example, with this incoming administration, they
- 42:05want a focus on eating healthier.
- 42:08Also trying to change your lifestyle, you know, go outside,
- 42:11enjoy the sun. Those are all things that you
- 42:13don't need a medical doctor or a nurse to both prescribe for you.
- 42:17You don't have to prescribe. And two, it's hard for them to
- 42:19follow up on it, but you could imagine if you did have an AI
- 42:22app that was actually about helping inform the patient so
- 42:25that they could actually do the behavioral changes that David
- 42:28Rich was talking about so much of health, especially if you
- 42:31want to sort of in that cost curve he was talking about
- 42:34there. There there's things that
- 42:35traditionally in the United States we called public health.
- 42:37They were behavioral changes. But again, because the strength
- 42:41of the United States is where highly decentralized.
- 42:43That's also the challenge of the United States is we are highly
- 42:45decentralized. You have freedom of choice.
- 42:47The question is, how do you bring how do you bring the Nexus
- 42:51of both choice and agency to the individual to let them pick what
- 42:54they want to do, but also. Find a way.
- 42:56To bring together all those different parts that are not in
- 42:59the immediate medical setting that can improve health
- 43:02outcomes. And I think that's where there's
- 43:04a long tail, both before you see a doctor and a clinician, but
- 43:07then also afterwards too. That is that is huge
- 43:10opportunities for innovation in the space if it you know, and
- 43:14recognizing again, different countries have different
- 43:15choices. EU s s strength is where
- 43:17decentralized, but any solution has to take that into place.
- 43:21This is from Philip C on LinkedIn, who describes himself
- 43:25as a builder, investor, and advisor, and he says Amen to
- 43:30sandboxes. Is there a realistic pathway to
- 43:34bring together user data isolated in healthcare systems
- 43:38and user data scattered across their consumer apps?
- 43:42It seems like that's a natural next step for greater
- 43:45contextualization of people's healthcare needs via Gen.
- 43:49AI and other AI techniques. So going back to what we were
- 43:53talking about both 4 David Rich, you want to jump into this one?
- 43:57The how to how do we bring data together?
- 44:00Well, I, I have a friend who taught me the expression that no
- 44:03one lies to their search search engine.
- 44:09And, you know, and it's, you know, when we stopped laughing
- 44:12about it there, you know, there's a certain truth to it
- 44:14because, you know, you know, we, we, we really want to know
- 44:18certain things. And I think that, you know,
- 44:21looking at, you know, consumer information, marrying it with
- 44:24medical information, it, it certainly is going to provide
- 44:26additional insights. And so it's something that I
- 44:29would look forward to. The other thing that we really
- 44:32haven't? Addressed so much in this talk.
- 44:33Is is the concept of privacy? What is privacy?
- 44:36Did we give it all up once we started tracking ourselves with
- 44:39our phones? Of course we gave up some
- 44:41aspects of it. And there's a balance in society
- 44:44between the respect for privacy and the need for things like
- 44:47national security. So where are these balance is
- 44:50going to occur and as genomic information proliferates, Mount
- 44:54Sinai has 300,000 genomes into a 1,000,000 genome collection
- 44:59project. There's no such thing as a de
- 45:02identified genome. So we have to bring together
- 45:05what can be de identified and perhaps work on that.
- 45:08And I think we also have to create safe harbors as.
- 45:13Something that's in the national interest where we can bring
- 45:16together data that have to be protected so that individuals
- 45:20genetic information is not released, but to truly
- 45:23personalized medicine to truly bring together it's the
- 45:27information outside of the medical records, it's the it's
- 45:30the information inside the medical records and it has to be
- 45:33shared in a way that we are comfortable.
- 45:36With as a society. I'll just say one thing.
- 45:39Hippocratic AI. I'm going to answer our
- 45:42gentleman from Doha and this new gentleman by saying I agree that
- 45:47there's so many health applications out there that
- 45:50actually not not just the day-to-day clinical care that
- 45:54we're talking about, but help in many other ways.
- 45:57And one of the solutions that I just love is the one of our
- 45:59investments called Hippocratic AI.
- 46:01And their mission is really to close the growing shortage of
- 46:04healthcare workers by helping them scale.
- 46:07And they don't do, they don't do prescribing.
- 46:09They don't do, but they do follow up and like they call
- 46:13down, there's an agentic AI that calls the patient and says, ask
- 46:18them questions about their health.
- 46:20Have you checked your blood pressure?
- 46:22Have you taken your medicine? Did you get your prescription
- 46:25filled? Have you eaten healthy today?
- 46:27Back on nutrition where you could scale a nutrition as to a
- 46:31million people at a time, but then merging those two data
- 46:35sets. They're a great example of being
- 46:39able to open up the World Wide Web of the Internet of
- 46:43information to also some clinical details without going
- 46:48into prescribing or doing the work of a doctor.
- 46:51But helping that doctor and nurse be better, have more time
- 46:56them on their hands and them getting that information about
- 46:59back about what is the patient doing, how when I actually see
- 47:02them face to face or do a telemedicine call.
- 47:05What are the core things that I have to focus on now?
- 47:07So we're seeing so many great applications now that are truly
- 47:12helping our physicians and nurses scale their work effort.
- 47:16But also in the back end, it's more administrative work.
- 47:19It's helping educate and advocate that things honestly,
- 47:23I'm sure Doctor David would love to do, but doesn't have time to
- 47:25do too. He's he just doesn't always have
- 47:27time to do that. So this is where AI is really
- 47:31going to help us, I think in our world of of care for patients.
- 47:35David Bray, very quickly question from Twitter, how do
- 47:39you deal with ensuring ensuring that AI is used ethically in
- 47:44healthcare? There are competing agendas,
- 47:46including extracting value from patients to business.
- 47:52So how do we manage these competing agendas and where do
- 47:56the, where do the ethics come into play?
- 47:58And very quickly, please. Two things that I think we need
- 48:00to see more of. And, and I've I've seen this
- 48:03already work in the UK. We can do it in the US, the UK
- 48:06they call them data trust. Here in the United States, we
- 48:08call them data cooperatives. I volunteer with an effort
- 48:10called Birth to Three that tries to make sure every infant in the
- 48:13United States gets a necessary physical, mental and emotional
- 48:15care they deserve. Obviously data about infants is
- 48:18very sensitive data. And So what we did is we used
- 48:21the existing contract law to create a data cooperative that
- 48:25includes members of that community, including
- 48:28representatives, and they have choice as to how the data is
- 48:31used where. And we explicitly said at no
- 48:33point in time will this data ever be monetized.
- 48:35But then you can actually then choose which systems are done.
- 48:38And we're increasingly we're, we're pushing for the algorithms
- 48:41if they're going to train, come train in situ with the data as
- 48:45opposed to porting the data somewhere else.
- 48:46This also solves the interoperability problem because
- 48:49it's not about interoperability of the data elsewhere.
- 48:51It's about can you bring the algorithms.
- 48:53Now there's challenges with that, but you can do that.
- 48:55The last thing I would also say is we may need to have.
- 48:58More boards that include members of the patient community being
- 49:02served by a clinical setting be involved maybe every three
- 49:06months, every six months worth. Just let's explain what we're
- 49:09using the AI for and you can ask questions.
- 49:12So stakeholders in both on the data and on the AI?
- 49:14David Rich, very, very, very quickly, you mentioned the use
- 49:18of Gen. AI to summarize complex patient
- 49:23histories. Question from Twitter, who
- 49:27should be held reliable if the AI recommendations are
- 49:33hallucinations and how would you even know that they are
- 49:36hallucinations? Very quickly, please.
- 49:38We have. A real obligation as we develop
- 49:41the tools to assure that they are accurate.
- 49:44I often refer people to a study done by radiologists showing
- 49:47that when in then mistakes were intentionally entered into a
- 49:52predictive algorithm, that it LED people astray and the
- 49:55residents were wrong 80% of the time.
- 49:57Even the experienced radiologists were wrong 50% of
- 50:00the time when the AI LED them astray.
- 50:03So I think it is a prerequisite that the tools we develop have
- 50:08to be overseen, because when we build tools, we build them on
- 50:12data that are frankly biased in nature because of inequities
- 50:16within our healthcare system. So that's part of the quality
- 50:20assurance process. We make a lot of errors in
- 50:23medicine, and we're always struggling to try to make less
- 50:25and less and to improve things. The same thing applies to any
- 50:29technological tools that we have at our disposal.
- 50:32As we finish up, I'm going to ask each one of you a separate
- 50:37policy related advice related question and I'll ask you to
- 50:40answer quickly. Teresa, let me start with you.
- 50:44Given all of this that we've been discussing, what advice do
- 50:50you have for policy makers when it comes to technology such as
- 50:54AI and healthcare? My top advice to them is they
- 50:58have to learn. They, and I'm seeing them open
- 51:01to this, but they really have to learn and their policies have to
- 51:05move faster. The technology is moving so much
- 51:08faster than they're instituting policies.
- 51:11And so they've got to learn and understand.
- 51:13They've got to be open to new ideas.
- 51:14They've got to be really on top of this.
- 51:17But last, what we went to earlier is they need to be
- 51:20thinking about AI at the industry level.
- 51:22Let the experts do their job. Don't do broad policies that
- 51:26hurt the industry. Do something that allows the
- 51:29industry to understand. Don't over regulate, but let
- 51:32them do. They're already doing it.
- 51:33Let them just own it and do it well.
- 51:35David Bray, What advice do you have for technology developers
- 51:39when it comes to AI and healthcare?
- 51:42Very quickly. Please learn from what happened
- 51:44with the introduction of ultrasound for sonograms in the
- 51:48healthcare space. Initially, when ultrasounds and
- 51:51sonograms for pregnant women were introduced, there was
- 51:54resistance. There was questions about the
- 51:55efficacy. It had to be proven, But also,
- 51:57how did you work it into workflow?
- 51:59But now nobody would question. In fact, if anything, it might
- 52:02actually be wrong to not do an ultrasound for a pregnant woman.
- 52:06So it's worth understanding that there is a journey.
- 52:10Be patient with that journey and figure out again, as is David
- 52:12Rich was saying, how do you plug into that workflow but also show
- 52:15quality? But David Rich, what advice do
- 52:18you have for healthcare leaders when it comes to AI and
- 52:25healthcare? Surround yourself with a group
- 52:28of. Forward thinking people and
- 52:30people who are perhaps even part of the old guard.
- 52:34You need to have a brain trust and that brain trust has to be
- 52:39there. It has to come together and
- 52:41provide the best advice because it's becoming like the Wild West
- 52:46out there and we have to have a way of digesting and, and really
- 52:52facilitating better healthcare within our organizations,
- 52:55reducing waste, cutting out cost and assuring quality.
- 53:00And you know, technology can be a double edged sword.
- 53:03So we have to be sure that, you know, we do the best that we
- 53:07possibly can. And the way to do that is to
- 53:10surround yourself with the best minds.
- 53:12Makes sense? Surround yourself with the best
- 53:14people. And with that, a huge thank you
- 53:17to Doctor David Bray, distinguished chair of the
- 53:19Accelerator at the Stimson Centre, Teresa Carlson,
- 53:23president of the General Catalyst Institute, and Doctor
- 53:27David Rich, Chief clinical officer of Mount Sinai Health
- 53:31System and president of the Mount Sinai Hospital in New
- 53:34York. And a huge thank you to all of
- 53:37you folks who watched and who asked such great questions.
- 53:41Check out cxotalk.com, subscribe to our newsletter, join our
- 53:45community, and come back, come back next week.
- 53:48We have two members of the House of Lords next week talking about
- 53:52AI and the ethical aspects and regulation.
- 53:55So join us. Thanks so much, everybody.
- 53:58I hope you have a great day and we'll talk to you soon.