00:00So what do you, when you hear about AI, I feel like everybody we're talking to
00:05mentions that it has a lot of potential great uses for health care in the medical community.
00:10How do you see it and how are you seeing it in your world?
00:13Well, I think AI has the power to be transformative in our world.
00:19I don't think that's a hard thing to understand or see, but we have to be cautious as we use
00:26it.
00:26I think, as we were discussing earlier, it's not quite there yet in terms of being trained
00:32to a point where you're certain of its accuracy.
00:35And there are lots of moves within platforms to improve that accuracy.
00:41But you can see the road ahead, how it's going to help us understand, get to diagnosis quicker,
00:50help clinicians.
00:51I think it's going to be an incredible adjunct.
00:53It already is.
00:54You know, I see how much we are using those tools to bolster our capabilities, to check,
01:03to push back, and really try to understand to get to the most accurate diagnoses for patients.
01:10There's lots of uses, obviously, within automation and helping ease of work.
01:15We're using ambient technologies, you know, to be able to record conversations and create
01:20the first templates of our notes so we can focus more on one-on-one time, personal time, with patients.
01:27So I think it's a vista ahead of us, but needs to be done smartly so that we really include
01:34the data
01:36for all populations in the models that we build and the answers that we give.
01:40You mean like women?
01:41Like women?
01:42I think that's a focus of your segment today.
01:46You know, we haven't historically done a great job of including women in clinical trials,
01:53collecting data on women, understanding.
01:55But to your point around technology with real-time data, right, real-time data sets on women,
02:04coupled with studies targeting women's illness, I think we're going to have a much more nuanced
02:11and personalized understanding of disease.
02:15When?
02:16Today?
02:19We have to do this, I feel like.
02:22I do not want my daughter's generation to be sitting here.
02:25With the same conversation, right?
02:27With the same conversation.
02:28And I think it took the unlocking with the AI tools that are coming to us to make that a
02:34possibility.
02:35I guess what I'm curious about is how quickly you see this happening.
02:40Not right now, and to Carol's question, when, but like if we look at this as a moment in time
02:46over your entire career and the way that you're seeing tech actually move the needle
02:51when it comes to what you do and what your doctors and nurses do every single day at Boston Children's,
02:56are we in a moment where we are going to see us be able to make leaps and bounds
03:01when it comes to development of new therapies for certain ailments, new diagnosing tools?
03:08Like, is this a pivotal moment?
03:10This is 100% a pivotal moment.
03:13Like the biggest in your career at this point?
03:15Absolutely.
03:15It's like an industrial revolution is taking place.
03:18You know, it's that type of moment in time, from my perspective.
03:22I think that...
03:23But you see stuff happening that makes you think that, right?
03:25It's also a very important perspective for healthcare, too.
03:27Yeah, totally.
03:27I think generative AI was the first next milestone, right?
03:32And that was, what, three years ago, right?
03:35And then the agentic piece that's come over the last year or two is unlocking all new capabilities.
03:42So I think that it's the evolution and the geospatial AI is going to have its moment.
03:49So when you start to be able to understand where are the medical deserts in care, you're able to use
03:57agentic workflows across data that can bring back information that you really couldn't see.
04:05You wouldn't have the scale and the capability.
04:08Who's building the data sets?
04:10Like, I think about that.
04:11And one of the things that we've, I think, through people we've talked to, how it's shifting to, it might
04:16not be that everybody's using these large language models,
04:18but you'll have very specific vertical data sets, especially for, like, healthcare or medicine, so that when you are working
04:27within it, it's a really smart, appropriate database.
04:31So who's building those databases?
04:33Are you guys doing your own?
04:34Are you working with other medical organizations?
04:36Like, how does that work?
04:38I think it's something that's evolving.
04:41Okay.
04:41There's lots of data sets that have been built over time.
04:46Right?
04:46In terms of electronic health records, data sets within registries.
04:53There are many data sets, and you raised the right point, making sure those are solid, good, evidence-based data
05:02sets, that this is the data you want to be using.
05:05There's the UK Biobank.
05:07There's large data sets that are unlocking this sort of potential.
05:13And I think we're now having this real-time data collection capabilities that is going to be very interesting to
05:22see.
05:22Instead of, say, a glucose test once every three months, you have a continuous glucometer running, you have a continuous
05:30glucose data set.
05:32So I think of it like my perspective, my cardiac ICU.
05:35I'm dealing with continuous data streams.
05:37Now there's an ability to analyze continuous data streams in ways that were just so challenged before.
05:44So I think it's that unique moment in time.
05:48We were told drug development, drug development, drug development is a real opportunity, too.
05:53Is that true?
05:54Yes.
05:55I think it is very much a real opportunity.
05:59So how, I mean, look, drug development's expensive.
06:02It's high.
06:03Slow.
06:04Slow.
06:05It's high stakes.
06:05You know, you see, sometimes, like, it's almost like every day the biotech companies are some of the biggest movers
06:10to the upside or to the downside.
06:12They take big risks.
06:13Very volatile.
06:14Does the development of AI, does it change sort of the drug development cycle meaningfully?
06:23I think that is the hope that everybody has right now.
06:25And there are many companies that are, you know, spinning out and doing that sort of work.
06:32Big Pharma has arms.
06:34They're doing that sort of work.
06:35You know, think about it in terms of the rare disease base and the sort of, like, children that we
06:40care for, say, at Children's.
06:43What an opportunity to be able to potentially bring things to market that, you know, weren't possible.
06:48And then when I think about, you know, the long-term care of patients and understanding the full patient profile,
06:59the phenotype of the disease,
07:00is we have new ways of actually understanding the patient cohorts that may unlock new possibilities in terms of developments
07:08of therapies for them.
07:10You know, we've always talked about very individualized medicine.
07:15Is that still a thing?
07:17And increasingly that is a thing?
07:19And will AI help it so that we can look at each of us, you know, because ailments obviously impact
07:24people differently?
07:25Yes, they do.
07:26They do.
07:28And I think you had Dr. Doug was on here earlier talking about bone health, right?
07:35Yes, yes.
07:36And how do you personalize things to an individual?
07:40I think this is part of this moment in time that you can personalize.
07:45Your physiology is going to be different than my physiology.
07:48And maybe how I live my life needs to be a little bit different to the way you live your
07:54life.
07:54Yeah.
07:54Based on knowing your lab values, your genotype, your phenotype, that then we could get ahead of understanding what disease
08:04you have the potential of developing
08:06and intervening earlier such that you may either delay or not end up with that disease trajectory.
08:15So are we there?
08:16Again.
08:17I think it's an infrastructure build.
08:20Yeah.
08:20And it's not just technology.
08:22I think that's an important point.
08:24Technology is not the only thing that you need to think about in solving for this.
08:29It's market inefficiencies, it's operations, it's regulations, and that spans the globe.
08:36We really need to think about that.
08:39And I think that's where I say we're in a revolutionary time in the sense of everything is changing.
08:44And how do we build the right and most appropriate structures?
08:48Global health is so important to you.
08:49Just 30 seconds.
08:51And this change in the game for underserved communities, underserved populations.
08:55Yes.
08:55It will.
08:57It will.
08:58If they can get the tech?
08:59I think it's the tech that's designed for the local community.
09:06Trust, building trust.
09:08I think it unlocks the ability to, I think, more expanded data around all of humanity.
09:19You know, not just like a U.S. population.
09:22And being able to understand, I think, how you can deliver better, more targeted care.
09:28I mean, we're already using some of this tech in terms of collaborating across geographies.
09:34I think that's one of the things.
09:36It closes the distance.
09:36I mean, we personally, with the foundation, Virtue Foundation, are already using some of that technology.
09:41And you have a book coming out later this year.
09:42I do.
09:43I'm just going to say it.
09:44We have to wrap it.
09:44It's Crossroads in Global Health.
09:46Crossroads in Global Health.
09:48Published with me and my colleague, Dr. Abiyalahi, coming out from Taylor & Francis in September.
09:53You guys have to come back and then talk to us once you have the book out.
Comments