[MUSIC PLAYING] [MUSIC PLAYING]
Daniel Kraft:
Good evening, everyone. I'm Dr. Daniel Kraft. And I'm lucky to chair this panel on exponential technologies and health and medicine. We have a pretty amazing, rock star panel across the spectrum of fast-moving technology, so I'll introduce shortly. But as we gather in, I thought I'd give a two-minute, quick framing of where technology is, where it's taking health and medicine, and the applications that many of these folks are pioneering. So we're listening. So you know, we live in this pretty amazing, connected, digital age. Some of that hasn't always caught up to health care. We're pretty much still in the sick care era where we're still waiting for disease to happen. But we do live in an era where things are moving very, very quickly. I have in my back pocket, for example, an antique 10-year-old iPhone 2, which still works when you plug it in.
10 years ago, this was the cutting edge of exponentials. Moore's law, the exemplification of exponential technologies, makes this look antique. And my iPhone X will soon look antique as well. And soon we'll have computers in our contact lenses and beyond.
So the world's moving quickly. And we can apply these technologies, whether it's computation, 3D printing, robotics, artificial intelligence, big data, low-cost genomics, in lots of interesting ways across the health care continuum. And the focus of our panel tonight is to meet some pioneers and thought leaders in that space. And we're going to go from the small side of things, with genomics, all the way to data from our bodies and our homes, to how we're applying AI machine learning to analyzing that and making it useful, to how it's being applied in the clinic in Boston Children's Hospital and beyond, and finally, how new players like Amazon are coming into this space and catalyzing new technologies, as exemplified by Babak Parviz and others.
So I'm going to start off. I'm going to let each of our amazing panelists spend about three minutes or so highlighting some of the amazing work they're doing, how it touches on exponentials, and where it's going to take the future of medicine. We're going to start with Eric Schadt, who is sort of a god of assistance biology in medicine. He's the dean of precision medicine down the street at Mount Sinai and the CEO and founder of a new company called Sema4. He's going to share a bit what's happening in the genomics and omics space and where that's heading. Eric. care continuum.
Eric Schadt:
Great. Thank you, Daniel. So you know, next-generation sequencing technology. You heard from the gene panel a pretty exciting technology enabling us to characterize all variation in the genome and its association to disease. And what many don't appreciate is that next-generation sequencing is one of the only technologies, the only technology known to humankind, that's moved at a super Moore's law pace. So the level of disruption, the level of insight, and how rapidly it's evolving to allow us those insights is something we've never before experienced in human history.
[INDISTINCT SHOUTING AND LAUGHTER]
Awesome. Having fun. That's awesome. We have a great view, so. So the idea in being able to think of next-generation sequencing as being on the super Moore's law path-- think of cameras, like the resolution increasing dramatically every year, enabling deeper and deeper insights, better views of what's going on, even to the individual cell level. And the way we think about that is we can now generate massive amounts of data in populations, and we can start interrelating how that system is put together.
You heard from the gene panel this idea like gene editing and gene therapy. Well, those are technologies that can target things, but you need to know what those things you want to target are. And biology is no more really thought about as a linear pathway of one thing going to the other. It's viewed more as a complex, nonlinear network.
And so what we think a lot about is how to organize all of that information to understand the components that the operating systems in each of our cells are executing. These are executing programs. And if we really want to get at effective medicines or therapeutics or preventions for common diseases like Alzheimer's, we have to be thinking about how we target, simultaneously, multiple dimensions of cells and organs and whole systems and how we do that over time. And so that's sort of been the embodiment of my work over the last several years and now at Sema4.
Daniel Kraft:
And just quickly on that, on the exponential-- cost of sequencing a genome was millions of dollars a decade ago. Where is it today, and where it might be in five years?
Eric Schadt:
Yeah, so the cost of sequencing the first genome, which was published maybe around 2001, was around a billion, $1 billion to $3 billion dollars depending on who you ask. The cost of sequencing a genome today is well under $1,000. And you know, the technologies that are coming that enable terabase sequencing capabilities on a single transistor-- and imagine millions of those transistors on a small chip-- you'll be able to do terabase scale sequencing in seconds. So in five to 10 years, it'll be more like a photograph, getting sequenced, than what we go through today. And it will be effectively free.
Daniel Kraft:
And clearly, we're going beyond the genome. There's other omics-- the microbiome, the exposome. And now we're in the internet of medical things that are becoming connected. And Deborah Estrin, a professor at Cornell and co-founder of Open Health is a pioneer in making sense of that data. So Deborah, tell us a bit about what's happening in your world.
Deborah Estrin:
Certainly. And I'm at Cornell Tech right in the middle of the East River here in New York City. So precision medicine based on advanced genomics doesn't yet, certainly is not intended to, address mortalities whose primary determinants are behavioral, some of the things we just heard Michelle Obama talking about. And even given these amazing approaches to cancer and drug development, everybody agrees that reducing risk factors should be the first line of defense. Atul Gawande, if you haven't read his work, I think really put it best, of course, in an Atlantic piece he wrote where he wrote that we need to expand our focus from rescue medicine to lifelong, incremental care or we risk leaving behind millions of people to suffer and die from conditions that increasingly can be predicted and managed. The more capacity we develop to monitor the body and the brain for signs of future breakdown and to correct course along the way, the greater the difference health care can make in people's lives as well as reducing future costs. So incremental care can, over time, lead to exponential health improvement.
And particularly, what I work on is powering those feedback loops of care with what we refer to as our small data. So by small data, I mean the data that emanate from your sensor-laden mobile phone and all the other online activities we do, the language we use, the things we consume, how mobile we are, and also what keeps us up at night, whether it's too much late-night Netflix or Seamless or Twitter or all of those things combined. So small data, processed into what we refer to as digital biomarkers, can come to inform how doctors actually prescribe and titrate medications, but also how patients manage and adhere.
So consider something as prosaic as lower back pain. According to the NIH, 80% of adults experience it. And it is among the most common disorders associated with prescribed opioid use. There are alternative treatments to lower back pain than opioids-- specific exercise regimens, walking, weight loss, even acupuncture. But patient-perceived progress in these things are very slow, and non-adherence is the norm. And the health care system largely kicks in only after another relapse.
So AI-powered personal agents could use an individual's small data to both anticipate and even preempt relapse by nudging patients towards their aspirational adherence behaviors. That's the connection between the small data we generate and actually helping individuals in the behaviors they need to prevent many of these chronic diseases. And this applies to back pain or hypertension or depression and so forth.
So the data are there. The challenge remains to extract meaning and significance, because they're very noisy and they're very privacy sensitive. And we need to create evidence of what actually works, how, when, and with whom.
So we are in New York City, not Silicon Valley. I can diss Silicon Valley. I lived in California most of my life. So we can acknowledge that tech is not by itself going to solve health care. But it might be our only chance to achieve better quality and reduce cost at scale. there.
Daniel Kraft:
Thanks, Deb. And you know, what's interesting, whether it's genomics or small data or your Apple Watch or your sleep data, the trick is how to put it together. No clinician now or individual can make sense-- our brains have not upgraded in two million years. But companies like Google, as [INAUDIBLE] by Sri and his team at Google Brain, are helping make sense of that. Maybe, Sri, share some of the latest in AI and health care, what Google's doing and where it's taking us.
Sri Madabushi:
Sure. So I think the points Eric and Deborah made were very pertinent. So there's an explosion of data that's happening, whether it's now what we are calling at genomic scale as opposed to astronomical scale, right? There's a whole bunch of sensor data, user data, that's coming up. What do you make of all of this? How do you interpret it? And that's where the big data piece of how technology companies like Google or Amazon or whatever can help. So we kind of accidentally stumbled upon this, to be perfectly candid, where one of our product managers happened to be in India, was visiting eye hospital, and the line was longer than what you see over here. And he started looking into it and figured out that it was because there was not enough doctors screening these patients. And Googlers being nerds, we looked at the problem. It was a bunch of images that were being analyzed and just did not have enough resources. So we figured there might be an approach of analytics and artificial intelligence looking at it.
Fast forward a few years. We took thousands of those images, trained our machine learning, deep learning, models to interpret it, and it turns out that without feature engineering-- that's telling that if you see this, then it's disease-- you automatically just train it at scale. And what we found is we are able to diagnose or screen for this disease, which is diabetic retinopathy, the number one cause of preventable blindness in the world, at scale, right?
And the beauty of it is because it's not a doctor, it's a machine, it's going to do this consistently. Most people who are doctors here know that you're not going to agree with yourself about a third of the time for the same patient you're seeing, right? This is because our brains are wired that way. But a machine can avoid some of these issues leading to medical error.
So the goal here is, can we put this technology on a variety of different devices at scale and put it in the hands of a village social worker? To the point that Michelle Obama was talking about, how can you enable access at scale so that somebody visiting New York Presbyterian here will get the same level of care as somebody in sub-Saharan Africa with a village social worker and a smartphone being able to screen and diagnose at scale? That's one.
Now, that is basically doing things accurately and with more access, what currently developed medical systems do. Now, if we take this further, we published a paper last week in Nature where we looked at the eye and predicted a variety of surprising things. So it might not be surprising for some of the expert physicians here, but we could predict cardiovascular events. We could predict blood pressure. We could predict smoking status.
And we could predict gender. Not that you have to look at the back of your eye to see if you're male or female, but the point is that there is no known correlation of gender with the back of your retina. Now, this obviously needs to be validated significantly in the future. There's a lot of clinical trials that need to be done. But it's very promising.
And it's prognostic in nature, right? Can you now generate hypotheses about a variety of diseases, whether they're cardiovascular, neurodegenerative, to predict things that humans cannot that the machine sees? So I think that's a very exciting field. And that's where we are looking to explore. It's early days still.
Daniel Kraft:
Thanks. And John Brownstein, you're chief innovation officer at Boston Children's Hospital, where I trained in pediatrics. And now you have the opportunity to start making some of this data useful, bringing it into the workflow of the doctor, et cetera. What are some examples of what data is being useful clinically at Boston Children's? Thanks. Thanks.
John Brownstein:
Yeah. Thank you. And you know, it's fortunate, because I'm right sandwiched between Amazon and Google here. And in fact, these are two companies that we think are incredible in terms of working together and leveraging the huge data-intensive environments that we exist in. In my role, we think about the patient journey from end to end. So how do digital tools influence what patients decide to do, how their care is received in the hospital? But even when they leave the hospital, how do we maintain a connection so that you're connecting with a patient not just one-off, once per year, but continuously? And so we're developing technologies and tools not only to extract and utilize that data, but to communicate and interact with patients.
So we're doing this in the home. So for instance, we think that there's huge opportunities in conversational, AI-driven bots, tools that connect with patients even if a physician isn't available to them. So we actually are developing skills on Alexa and with Google, this idea that you can actually start to build a way to engage with a patient even before they ever reach a hospital.
So Alexa, my child has a fever of 102.4. What does that mean? Give guidance that comes directly from Children's and then allows us to triage and impact capacity as people come into the emergency department. So we are now using these technologies not only to provide education, but also around resource management.
In the hospital itself, we live in a very data intensive environment. So very similar to what Sri mentioned, you know, we collect huge amounts of data in our ICU, which is our highest-cost environment. But we really don't utilize the data in any real, significant way to make decisions. But we know with the use of that data, we can develop predictive tools that essentially allow us to understand, for instance, length of stay, make predictions on how long a patient will be actually in the ICU, and make capacity decisions, but more importantly, predict, when that patient needs to be extubated, any outcomes that might occur.
Not only that, we're now integrating this kind of information with tools like Alexa. So a physician who is in a hands-free environment could quickly access information, data that comes from the EMR, but even things as simple as, how do I page so-and-so for a consult? How do I get specific dosing instructions for a particular patient?
That's time that costs huge amounts of money to a hospital. All this type of data is locked away all sorts of parts of our institution. We can organize that content with machine learning tools, but then also make that data available in these voice frameworks that are really transforming medicine. So in my view, you know, we have these exponential tools like voice. I think health care actually is a great place for the first real applications of these technologies to exist. And so--
Daniel Kraft:
And voice can even predict depression, mental changes, even heart disease.
John Brownstein:
Yeah, exactly. So there's all sorts of opportunities that, once you have these listening devices in the home-- and I know there's discussion about how much you want a device. I know there's debate about having something that's in the home listening. But of course, we know that by 2020, over 50% of the US population will have a smart speaker in their household. The reality is this is an incredible medical device in the home that can detect changes in voice and tone, in sentiment, detect disease, mental health status. So there's huge opportunities that we're still just scratching the surface on.
Daniel Kraft:
So these technologies are moving quickly. And I think Babak Parviz exemplifies engineering and converging all these technologies to address some amazing challenges. Maybe share some of your perspective on what you have worked on, what you're working on. Where are things coming together?
Babak Parviz:
Yeah, sure, I'd be happy to. So before I start, though, raise your hand if you can hear me. I wonder, actually, how many people can-- yeah, so we're a little bit acoustically challenged here, so hopefully you can hear me OK. I'm Babak Parviz. I'm with Amazon. And since this is our exponential medicine panel, I want to share with you two things. And they relate to two requests, actually, that I have for this group of pretty prominent, smart people that we have assembled here.
The first one relates to how we envision medicine operating. So a human body is a complex system that involves flow of information, processing of information, but also a flow of physical material, flow of mass, flow of molecules. In terms of exponential trends, we've just talked about this, that with the advent of computing, massive scale computing, access to data, being able to process a lot of data, we are beginning to see, as Sri was mentioning, we're beginning to see the glimmers of what machine learning and these new AI techniques can do in medicine.
And I think this is just the beginning of the exponential. And my expectation, and maybe I'm too optimistic, is that this is going to radically change health care and medicine in the next 10 years or so. So that's on the information side. And you know, there's a number of companies that are at forefront of this. And we also enable a lot of the tools that go into this, from cloud computing to the core machine learning techniques.
But as I mentioned at the beginning, medicine and health care is not just about information flow. It's also about the flow of material, flow of physical things, so we have this amazing ability today to move information around and process it. We should also remember that very recently, and this has also improved exponentially, we have come across the ability to do the same thing with material. So it's not just about moving bits around anymore. It's also about atoms.
So the global logistics network that's in place today in many advanced countries and expanding is an incredible asset. And I think that can also radically change how medicine is practiced, not just in the developed world, but also in the developing world. And my invitation to you is that-- and you know, Amazon is one of the entities that runs a significant chunk of this global logistics that moves physical objects, billions of physical objects, around the planet.
So I'm here today. I'll be, after this panel, around for a few minutes. If you have any radical thoughts of how that can be used, come talk to me. So that's my first request. But there's definitely a significant exponential trend
The second request relates to something that, you know, we have looked at for some period of time and we deeply care about. And that relates to what happens to older people. So we have looked at the older population in the context of health, obviously.
But we know that this group has a lot of issues. They have a lot of unmet needs. Some of them relate to health, but their health and the broader issues that they face are all interrelated. They're not operating in isolation. They're operating inside an ecosystem that involves a number of other people.
So the community in which our older adults operate really is a significant contributor to their health and wellness and well-being. And what we're seeing today is that, unfortunately, even though we've had this amazing exponential growth of connectivity through computational devices, the community around our older people has crumbled. So they're getting lonelier and lonelier, and they're paying a pretty heavy price for not having that community around them.
So my second request to this group is, this is something that we should pay a lot of attention to. And we need radical exponential solutions to recreate community for our older people, and hopefully for our younger people, too. Meaningful communities, that's absolutely needed. And it has a very direct impact on health and well-being of everyone in our society. Thanks, much.
Daniel Kraft:
Thanks, Babak. And one of the interesting things is connecting, we can connect community in new ways, whether it's on Skype or-- you're well-known for inventing now an antique device in my pocket, Google Glass, right? So my question for you-- this is a lot of exponentials coming together. What were some of the lessons from developing this sort of technology, which could be used to communicate and even foster community, that apply across innovation?
Babak Parviz:
Actually, this brings a lot of memories back. This is antique, that's true. This is many, many years old. I think the board in that is from 2010 or so, so many years ago. But I think we learn, actually, through doing this that introducing radically new things into the broader society is challenging. So in my humble opinion, this was a massive step forward in computing and communication. But we learned that it needs to get rolled out in a very different way in terms of making it broadly successful and get broad adoption. So it was not just about technology. It was about how technology is introduced and managed after it's introduced to the broader public.
Daniel Kraft:
And along those lines, talk about disruption and change and exponentials-- it's often about connecting the dots. And all of us have worked on technologies, but how do you get them out there and making a difference? And everyone is familiar with Uber. It's an exponential company. Couldn't have existed before smartphones, GPS, online maps. They didn't invent any of those. They connected the dots. John, you have some experience in Uber-izing health care. Maybe share some of that and how that applies.
John Brownstein:
Yeah, absolutely. I mean, I think, again, it's an example of an exponential technology that there's obvious direct application into health care. Obviously, we had a really bad flu season this year. Despite that, we still have a fraction of the population that's getting vaccinated, despite the fact that you can't walk a block in this city and not see free flu shots available at a pharmacy. So our idea was, if you can't get people to walk over and get a vaccine, why not get vaccines to people? So we somehow convinced Uber to do something called Uber Health, which was essentially put nurses in Uber cars and deliver vaccines on demand. And by doing that, what we found was it was pretty simple. The idea that you could get a vaccine on demand meant that it was simpler. It was more convenient.
And in fact, people were actually getting vaccinated that had never gotten the flu shot before, not because they're morally opposed to the flu shot, just because it wasn't convenient. So from that perspective, of course, flu shot is one example, but we also recognize that transportation is a barrier to access to care. It's one really important social determinant to accessing critical care.
So as much as it was great to think about flu shots, we expanded that concept now to think about, how do we deal with the huge population that's missing their appointments because of lack of transportation, or the real challenges in having to use existing transportation to get to their medical appointments. So now we launched, actually, this with Klick, a company called Circulation, which is essentially solving for that issue. And now we're in over 40 states, in over 80 health systems, where we're actually solving this problem of getting patients to their appointments, helping them get home, and ultimately dealing with these edges of health care that don't usually get a lot of perspective.
Daniel Kraft:
So one challenge-- I'll start with Eric and open up to all of you-- is we want exponentials to go into action. And one of the challenges in the clinical environment-- you know, where I trained at Mass General, we're still seeing fax machines on the wards. And Eric, where you've done pioneer work at Mount Sinai, you've helped connect the dots between clinicians and data. What are some of the opportunities and lessons there, and where's that heading? How do you get these things into the workflow, as well, for all of us so that it's not just overwhelming data becomes actionable?
Eric Schadt:
Yeah, it's a great question, because it can be pretty overwhelming. And these health systems are so conservative, so risk adverse, that trying to embrace a technology, especially overwhelmed doctors where the last thing they want is for you to alter their workflow in a way that makes their job harder or more demanding or that takes more time. So we've had to spend a lot of time-- you know, all these, the amazing exponential growth of these technologies, are-- you know, even today, the simplest of tasks-- we just came out with a 193-disorder pediatric screening test that can identify 193 diseases, all of which are treatable, some of which, if they're treated early enough, will eliminate the course of disease in those children. And this is genetics 101. These are hard-hitting, highly penetrant. And if you can have that test straight out of the gate, you can live a better life, have lower health care burden, and so on. But even getting something like that introduced as standard of care into a big health system is very, very complicated.
So I view one of the big challenges not so much about deep learning, the access to community-- that's all commoditized. I think we're going to see the algorithms that are used to process all the information and derive the insights. Those are all going to become rapidly commoditized by companies like Google and Amazon. And the real trick is going to be how to introduce that into the workflows in the health system so that they actually have an impact.
And so we've had to pioneer a lot of tools, digital engagement tools, with the physicians to basically ease their workflow, make the engagement of that data easier. And incorporating technologies like Alexa to, again, make the engagement of the information easier. So it almost is like we need a Steve Jobs-type figure in health that's going to solve that kind of interface problem to make that engagement more seamless.
Daniel Kraft:
Deborah, any lessons there from taking small data from my watch or my mattress? My physician doesn't necessarily want to see that data. How do we align the incentives?
Deborah Estrin:
Right. And they certainly shouldn't. They don't have enough time, right? So the challenges now are taking that data and turning it into something actionable. It's not every step you take. It's how your steps have greatly declined, or greatly improved since you've had a hip replacement, or whatever it might be. So all the work now is actually in creating, really, evidence engines for how to turn this capability that we have into something that is meaningful for self-care and for clinical care.
Daniel Kraft:
And the clinical trial piece. Just having your smartphone look at the back of your eyeball could be part of a clinical trial anywhere in the world. We could all be democratizing-- a lot of folks here from pharma. How do we take that data and translate that? Any--
Sri Madabushi:
Yeah, I think one of the things that we want to be cautious about over here is, like Eric was saying, this is just a technology. It's a hypothesis-generation machine. You still need to have the clinical validation through prospective trials, which is the first one, right? So engaging in the regulatory bodies is a critical piece. We cannot just throw a piece of technology across the fence and expect it to get adopted, right? We all know of a litany of failed startups that have tried that. Second is the point that John was making. Where this needs to get integrated into the current clinical workflow in a way that it does not add more burden, right? This is not another app or another screen you need to look at in your EMR that you are being forced to engage with. So this needs to be actually pretty seamless and pretty regulated in order to have the impact and scale that you want it to have.
Daniel Kraft:
So we've got about five minutes left. So sort of rapid fire-- maybe we'll start with Babak-- on how you see some these exponential technologies coming You've done some work in robotics, as well, that can help the elderly move around all the way to picking up your socks. And so I'd ask you to-- all of you, maybe think about what some of your favorite convergences and what your predictions might be where that's going to be in, let's say, five or 10 years.
Babak Parviz:
Yeah, actually, robotics is something I'm a big fan of. And a few years ago, we started an effort, actually, in robotic surgery. Hopefully, you can hear me better now. So we've been looking at, once the diagnosis is done, what are the typical things that we can do to help the patients? So we have all the biochemical routes to help a person. But the other big part that we pay attention to in medicine is doing surgery. And there's two issues with that.
One is, even today, there's millions of people in the world dying from known diseases that are curable with surgery, and we just don't have enough surgeons. So how do we scale the surgeons and make surgeons more available? So that was one big push that we had towards robotic surgery for that reason-- automated, actually, robotic surgery for that reason.
And this is my work, actually, in prior years. Has nothing to do with Amazon. So big, important disclaimer-- not Amazon.
And the second one was that the fundamental limitations that we have when we have a human surgeon. So a human surgeon has two hands, has limited spatial abilities, has very limited abilities to decipher molecular composition of the tissue. And we were trying to overcome those. So the question was, could you make a surgeon that can operate in millisecond time spans and do things very quickly and do things with 50 hands as opposed to two hands and see molecules as they do surgery and kind of get closer to some of the fundamental time constants of biology?
And I'm very bullish on that. Unfortunately, we don't have enough effort in robotic surgery as I would like to see. We have a handful of companies that are pursuing this. Hopefully, we'll see more of that.
Daniel Kraft:
And a great example of blending surgery, now, with the future of education is adding on virtual reality to train surgeons to scale. We had a great session at my exponential medicine conference where now you can scale medical education and eventually train the robot. So any quick closing thoughts and predictions?
John Brownstein:
Yeah, I mean, I think that we have incredible access to data, but we're still not quite there at this concept of the digital phenotype, this idea that the data that we generate in our day-to-day lives, the breadcrumbs we leave through our searches on Google, interactions on Alexa-- I think that's still a huge opportunity. It's a continuous monitoring of an individual that ultimately can help us understand outcomes. So you know, that's my prediction. We're going to get there. Still a little ways, but we have all the assets at our disposal.
Deborah Estrin:
And just to follow up on that, that individuals would be treated as n of one. So you're not just treated because a randomized controlled trial said that 75% of the people do better on this medication. You're actually treated on a medication, you see how you respond to it and how much of it you respond, because these digital traces can give you precise and objective feedback.
Sri Madabushi:
And I personally feel that whether you're n of one, whether you are a provider, whether you're a self-insured employer, the current situation is that it's a lot of silos, right? Which means that you're not able to integrate that data and extract the value, the tidbits that you were talking about, John. I think that is going to go away in due time, because there is a lot of effort moving in to establish standards, like FHIR, HL7, et cetera, to extract the information. And you're bringing computing at scale to the problem, so you can brute force it, too. So I do think that I'm pretty optimistic within the five or 10-year frame that a lot of the silos that have limited our ability to have breakthrough innovations will actually go away.
Daniel Kraft:
And after you align the incentives for those silos to mix. And Google spin out, Verily Health, is now doing that baseline trial to understand what some of those siloed informations might mean in the real world. And there's this idea of a digital twin. I have my body MRI, my microbiome, my genome data, and it can be accessed to do that sort of true, individualized prevention, diagnostics, and therapy. Eric, any kind of closing predictions of where you see systems medicine taking us?
Eric Schadt:
I think we haven't really seen anything yet. I think as the next wave of technologies come, again, just remember this super Moore's law speed and the kinds of technologies coming next that would allow you to walk around and seamlessly sequence everything you're being exposed to. The scales of data are going to be unimaginable. That they serve as sensors for every dimension of your health, not just from high digital imaging of the eye, but any of this high dimensional data and environmental context and inside you personally is going to inform on every aspect of your health, every aspect of your ecosystem. And it's just an exciting time to be in the game.
Daniel Kraft:
And not just individual health. You've done work with city health, like sequencing the sewers of New York City, not just for alligators but--
Eric Schadt:
You got it. The PathoMap project that I helped Chris Mason launch at Cornell is doing exactly that.
Daniel Kraft:
So remember, the pace of exponentials is often deceptive. 15 exponential steps is 16,000, but 30 is a billion. And then 31 steps is two billion. And so it's imperative that all of us, I think, look at not just the space of change, but the convergent and how we might apply that new lens to innovating across the health And I think all of us in the room are lucky to play a role in helping forge that future together. So I want to think our incredible panelists. And hopefully, you can mix it up and meet with them afterwards. And hopefully all of us will go create that bright future.
Sri Madabushi:
[APPLAUSE]
