Latest / StartUp Health NOW Podcast / Vinod Khosla: Investing in the Future of Healthcare & Moonshot Thinking

Transcript

Vinod Khosla: Investing in the Future of Healthcare & Moonshot Thinking · StartUp Health NOW Podcast

• 0:00
  [on-hold music] When you need real out-of-the-box thinking, you can't rely on experts in an area.
• 0:19
  But the, the hard questions they ask are extremely valuable to a good entrepreneur.
• 0:25
  [upbeat music] Hello, and welcome to the ninth season of StartUp Health NOW, our weekly show celebrating the entrepreneurs, the innovators, and the investors transforming healthcare.
• 0:45
  My name is Steven Krein. I'm the CEO and Co-founder of StartUp Health.
• 0:49
  And along with my partner, uh, president, and co-founder Unity Stoakes, um, we have been on a mission since 2011 to improve the health and wellbeing of humanity by investing in a global army of entrepreneurs who are committed to solving the world's biggest health challenges.
• 1:08
  We call them health moonshots. We are incredibly honored today to be welcoming back the Vinod Khosla, a legendary and prolific entrepreneur, investor, and technology fan.
• 1:21
  He's the founder of Khosla Ventures, which is focused on impactful technology investments like those that are in StartUp Health.
• 1:29
  And what's really exciting is that this is the fourth conversation we're having, almost on a pretty regular basis every few years, with Vinod.
• 1:37
  Uh, he first, uh, came to visit us in 2014 when the digital health, uh, market was just evolving and maturing and beginning to start to see some momentum.
• 1:49
  And what's really fascinating is that as we were preparing for today and looking back at some of those videos, Vinod's wisdom, his insights, and his vision is timeless.
• 1:59
  And we are incredibly excited to have and spend the next hour with Vinod.
• 2:03
  Um, as a kickoff, uh, Vinod, to, um, the conversation, I have to obviously start with the fact that, um, one year ago today, we were gathered in San Francisco at the StartUp Health Festival, our annual celebration of health innovation, um, a mere weeks away from the global pandemic really transforming everything about the way we live and think about the future.
• 2:30
  Um, I'd love to kind of start off with, you know, a question around this idea that we had a before COVID world, BC, and now we're in an after COVID world. And what do you think the biggest difference is today
• 2:45
  than it was one year ago when we were together at the StartUp Health Festival? So let's, uh, start with the basic idea that you mentioned, which is we didn't-- we were weeks away from pandemic.
• 3:00
  In fact, we were well into a pandemic. We just didn't know it. And that's a typical situation in healthcare. If you're a cardiac patient, you only know it 20 years after the condition starts.
• 3:16
  If you're an Alzheimer's patient, you only know it 20 years after the situation starts. I could go on. We knew we didn't know, but that didn't mean the pandemic didn't exist. It was already kicking off.
• 3:33
  The fire was burning. Now, the good news I think the last year has brought, to get directly to your question,
• 3:43
  is we have realized we as a society are limited not by what we can do, but what we think we can do. If I'd said to you last year at this time at JPM
• 3:57
  that we can develop vaccines in one year, I would have been laughed out of the room. And why that was the case was only because we thought that wasn't possible, not because it wasn't possible as we've proven.
• 4:14
  By and large, the system has become much more moldable and open to experimentation because of COVID. We tried everything. Um, take question of hydroxychloroquine.
• 4:32
  We didn't decide we weren't gonna use it till we had five years of clinical trials. We tried it and discarded it very quickly because it wasn't the right thing.
• 4:45
  But we didn't prevent us in uncertain information from experimenting with it. Every manner of experimentation has been done this last year to solve this catastrophic problem.
• 5:03
  And I think I, uh, that, that ex-culture of experimentation will live with us. And the moldability of the healthcare system has increased dramatically.
• 5:17
  You know, most physicians would never look to doing a telecare visit. Today, most, I think eighty percent or more, have done a telecare visit. I wonder about the remaining twenty percent.
• 5:36
  Uh, but the, the system has become much more dynamic and flexible, and it bodes well for the future.
• 5:44
  When, when you think about the, the possibilities today, you think about the way you describe it's what we didn't know and we thought differently.
• 5:53
  Do you think something is possible today that is gonna be achieved over the next decade that wasn't possible over the past decade as a result of the pandemic?You know, we today, my biggest beef is we today have artificial
• 6:16
  categories or, of specialties. We have primary care, we have endocrinologists, we have gastroenterologists, we have cardiologists, um, and we have to have them. It's not a bad system
• 6:32
  if a human being has to be an expert in each of these areas. And the cardiologist has to keep track of every study came out in the last year, maybe thousands of them, and know how to apply it to their patient.
• 6:48
  The problem with that is the cardiologist doesn't know what the endocrinologist is doing, and integrative care is not available to the patient.
• 7:00
  I have a hard time imagining that will continue to be the case. Um, so I think there's sort of vision documents.
• 7:14
  It was in 2011 I wrote my first blog that got everybody up in arms, do we need doctors, if, uh, you remember, and then there was a Rock Health talk. Um, I think we are well on our way to leveraging
• 7:34
  AI to do 80% of what doctors do, and that's really what I meant to say and what I've said all along, and use physicians for the 20% that are either the human element of care
• 7:47
  or the best expertise, the top 10% of physician diagnosticians, um, who can, um, be part of our care. So I do think we don't know this is possible.
• 8:01
  Most people don't believe it's possible, but it is happening in narrow specialty areas, um, initially.
• 8:13
  And, and how, how does that apply, do you think, to, one, the activities of entrepreneurs and innovators and, and investors and their thinking about how to take the rapid transformation of the last, you know, 10, 12 months and turn it into action now in 2021?
• 8:32
  Well, the obvious one, which is sort of to me, uh, marginal use of technology is probably the right word. It's very valuable. Telecare. Um, we started with primary care as telemedicine,
• 8:50
  but it was really putting somebody on a video with some guardrails around security and safety and privacy and all that. Um, to me, that was really not innovation, and it isn't.
• 9:06
  Um, we've done something that's, I think, pretty telling. And let me go back a little bit to the general belief that healthcare is very hard to change. And let me go address that because it's a favorite beef of mine.
• 9:23
  Yes, it is, because people have interests, whether it's not wanting to learn something new or a new way of doing things unless you're forced, like COVID face-
• 9:38
  forced most physicians to do telecare, or it's revenue models. I make so much money off cardiologists, uh, or cardiac patients. I don't wanna change my model. What we have seen, and I think the way innovation will happen
• 9:57
  in these verticals. You know, primary care, there's innovation that's going to happen because with telecare we can go to AI, and we can come back and talk about it.
• 10:07
  There's some-- There's one or two really serious efforts and about 90 not so interesting efforts, but still useful efforts in telecare. But go back to verticalization, which I'm a huge fan of.
• 10:23
  You have a traditional healthcare provider with their primary care, with their cardiac care, with their, uh, mental health care. Uh, I think we started to see the real explosion of verticalized care.
• 10:42
  Livongo said, "I'll take your diabetics and handle them a particular way, and almost all remotely." There's companies that have, like Ginger.io, that have done that in mental health.
• 10:55
  Sword Health and Hinge have done this in musculoskeletal. Take every large part of the budget and see what can be done remotely, and could 80% of it be done 10 X cheaper?
• 11:10
  That's not to say in cardiology we won't need interventional medicine or imaging, things like that. Uh, we will. Or we won't need ACL repairs.
• 11:22
  But 80% of the healthcare costs in these verticals can be done remotely in most cases, and I'm generalizing, so don't hold me strictly to it, and that 80% can be done 10 X cheaper.
• 11:38
  And more importantly, when we do it with the right kind of remote, we are going to see rapid increase in the capability of these systems.
• 11:51
  These will be what the Institute of Medicine calls learning health systems, which is not possible the way healthcare is done today. Uh, um, except in a very minor way.
• 12:04
  But in, with AI, which can learn from every interaction, we will have a learning health system, the real vision of the Institute of Medicine.But these verticalizations will allow innovation to happen, suck out pieces of revenue from the healthcare system, which to me is suck out pieces of cost from the system, serve the patient better, serve the patient much cheaper.
• 12:31
  And I think that's pretty exciting, and I think that'll be a d- we'll see a decade of increasing percentage of these verticals being handled. By vertical I mean diabetic care, hypertension care.
• 12:43
  AliveCor's doing an exciting job with cardiac care, musculoskeletal, mental health, you pick. Women's health, uh, pick your favorite area. Right.
• 12:55
  And I think it'll be s- carved out, like biting at small, uh, pieces of the big elephant and eating one little slice at a time by entrepreneurs and innovators. Yeah.
• 13:09
  So, so you're talking about the capabilities or the exponential capabilities being available. A lot of this technology was in development or even developed prior to the pandemic.
• 13:19
  Um, interesting fact, we came out with our StartUp Health Insights Report last week. In 2011, when we launched StartUp Health, there was $2 billion of funding flowing into digital health.
• 13:29
  It's 10X that now, $21.5 billion into, uh, digital health innovation in 2020, and we think it's just beginning.
• 13:39
  Will we see another 10X multiple on in- investment in this sector given the $10 trillion global industry that it represents?
• 13:49
  And what do you think we need to do to not only speed it up, but make sure that there's adoption of these technologies on the other side? Well, so two different questions. Um, first, will we see a 10X?
• 14:04
  I think anybody who pretends to know or claims they know is surely wrong, myself included. So I'll claim I don't know. But are we going to see rapid payback,
• 14:17
  which would result in rapid increases in, uh, investment over the next decade? Yes, we will see that in healthcare. So do I expect a rapid expansion? Um,
• 14:35
  the adoption question is very interesting. There's two types of adoption. One's within the existing system doing something incrementally better, and that adoption is much easier
• 14:53
  but not very revolutionary. It's not gonna result in a change in the way we practice medicine.
• 15:04
  The other kind of adoption is like Livongo and AliveCor and others carving off what are serious large dollar verticals of chronic care, for example, but the same would apply to oncology and other, uh, care,
• 15:24
  and radically innovating, renovating, changing them with new business models as well as new technologies. And I think that's the more exciting piece to me personally, but both are useful.
• 15:40
  So better surgical robot within a hospital system, yes, it's of course useful, and we have companies doing that. But changing the cost of diabetic care by 10X, which is entirely possible,
• 15:55
  um, uh, is, is much more exciting and, from a business standpoint, will be much m- much higher rewards. Um, here's, here's something I would just add. You know, the end of the Kennedy video,
• 16:10
  uh, said something which is, "We're going to the moon not because, uh, it's easy, but because it's hard." I like to say in most areas of technology and business,
• 16:25
  not all hard things are valuable, but most valuable things are hard. And people who take the long perspective
• 16:34
  and the investment and the learning perspective, whether the investment is in a core technology or learning over five years, um, uh, they will have better payoffs over the long term.
• 16:49
  The short term, there's easy pickings that people will mop up, and those are perfectly good business investments. I don't know if they're as interesting to me.
• 17:01
  Increasing billing for somebody with a better statistical system, not that interesting to me personally, uh, but it could be a pretty good business.
• 17:12
  Uh, so I would sort of say fundamental revolution is both possible and likely to accelerate, and the level of care, the vision I have for 2030 is really, really much different and bigger than where we are today.
• 17:30
  And- So then- It's been great. [laughs] Yeah. The second theme we really wanted to dig into is this concept of collaborative innovation.
• 17:40
  And, and over the years we've spoken with you a lot about, um, or heard you say that innovation or transformative innovation rarely happens from inside, uh, rarely happens from inside- Yes... the legacy systems.
• 17:55
  Um, but yet in the last year, as, as you said, um, there's been extraordinary progress with, for example, the vaccine development, where we saw Pfizer and BioNTech and Moderna and NIH collaborate to speed up innovation.
• 18:13
  I wonder if you could speak to this concept of the power of collaborative innovation, if you think there's a, a new potential here for large and small e-...
• 18:24
  uh, and startups to work together to try to speed up health innovation specifically? Um, I can give you the politically correct answer, but the real answer is collaboration is generally BS.
• 18:36
  [laughs] Um, with some exceptions. Uh, let me be more nuanced. Collaboration works at certain much later scaling stages of companies, not early when they're innovating.
• 18:54
  When you have... You know, if Moderna was responding to NIH early, I don't think they would've moved as fast. The experts would've slowed this down, leaving aside the political factors. Um, uh,
• 19:12
  you know, Pfizer ev- obviously helped scale a vaccine,
• 19:18
  but if the fundamental innovation had been done with Pfizer's experts, it would've slowed down, maybe even not been radical, but they were very, very valuable in helping come in late and help scale it.
• 19:33
  So when you need real out-of-the-box thinking, you can't rely on experts in an area.
• 19:43
  Uh, but the questions and the, uh, the, the hard questions they ask are extremely valuable to a good entrepreneur.
• 19:54
  So there is value in, in sort of their point of view, extreme value, because a good entrepreneur will start taking it into account early without changing the radically different point of view.
• 20:10
  Um, in a completely different domain, um, you know, I've been working with a fusion effort. There's real experts fusion effort that's global, collaboration across seven or 10 countries,
• 20:27
  um, and, and trillions of dollars literally. Um, a little startup like Commonwealth Fusion, leveraging the best ideas from there will probably be dead by 30 years, or fail and flame out.
• 20:44
  Either is possible, but that's the consequence of taking risk and radical ideas. You do flame out often, but when you don't, it more than makes up for all the times you did flame out.
• 20:58
  Uh, so I think collaboration has a role in scaling up, but imagine if Tesla had gotten advice from General Motors on how to build a car.
• 21:08
  They would've ended up with the EV1, and I think if we take advice from healthcare providers in radical innovations, we are gonna end up with the EV1's equivalent.
• 21:22
  Marginal cars, maybe if we're really lucky, a Prius of healthcare. Uh, not a Tesla, which everybody kept predicting forever would fail, would go bankrupt, and now- And-...
• 21:37
  is the most valuable car company in the world, but it could have gone bankrupt. Sir, there's some amazing- Uh, I th- I think that will happen in healthcare and healthcare innovation. It's the nature of innovation.
• 21:50
  There's some amazing, uh, questions coming in from the audience, and, and, um, one theme that's coming up, you, you kinda teed up earlier, which was your 2030 vision.
• 22:01
  I'm wondering if you could elaborate into more of what that is. Okay. So let's, let me s- make some general statements, and then, uh, try and elaborate on them. Fundamentally,
• 22:18
  a patient gets diagnosed based on their symptoms, and I think that's a fundamentally bad idea because you're addressing a disease.
• 22:30
  And there are acute care circumstances, you broke your bone, you need to address the symptom, but for most of healthcare span, most diseases you happen, happened, uh, started a decade before you noticed them.
• 22:46
  So eliminating systems, symptoms-based medicine should be a key goal by 2030. You shouldn't have to start losing your memory before you're diagnosed for Alzheimer's. That started much, much earlier,
• 23:05
  and there's measurable biochemical changes that should have been monitored. So, um, I like to say we have symptom-based medicine. We have tradition-based medicine.
• 23:20
  Every, um, every intern learns from physicians. You pass that along, but you don't have global scale. Um, if you, uh,
• 23:33
  if you had global scale, you wouldn't have hundreds of care pathways in each hospital system, each different. You would have the best damn care pathway for any given patient and his
• 23:46
  circumstances, uh, social determinants of his health. Uh, there'd be one answer, not different pathways based on different experts. So experts generally don't learn from each other, but we haven't had a better system.
• 24:03
  And I like to say healthcare today is better than it was 10 years ago. 10 years ago it was better than it was 20 years ago. So it has improved constantly and incrementally, in some areas significantly.
• 24:18
  But we haven't changed the practice of medicine. We learn from history and clinical studies and all that, which are n-But we haven't said healthcare and medicine should be a science.
• 24:36
  So I think the second goal would be change the practice of medicine to the science of medicine. And having humans relearn all of medicine is hard.
• 24:53
  Um, one study I saw said there's 30 cardiac biomarkers that could predict disease well in advance. Now, a decade or two ago, uh, there was one biomarker, LDL par- particle size.
• 25:08
  I think it came out of Lawrence Berkeley labs. It's now the Cleveland HeartLab's test. But h- what percentage of cardiac patients get that test? Not everybody, and that's 20 years later. And why did it take so long?
• 25:24
  And what if there's 30 or 50 biomarkers, maybe gene expression patterns that indicate you're going to have a problem in 10 years? That would be the science of medicine. Uh,
• 25:44
  so I, I'm pretty excited about this notion. So one of our companies just determined that just by a blood test they can determine NASH.
• 25:56
  Now, only way to determine you have NASH is a biopsy of your liver, which you'd only do if your symptoms were really, really severe because the intervention is pretty severe. Um, a- a- and you...
• 26:12
  So but should everybody be screened and determine NASH has started, and you shouldn't do something about it? Uh, absolutely. I think that's what I mean by the science of medicine.
• 26:26
  So eliminate system-based, symptoms-based medicine. Um, change the practice of medicine to the science of medicine.
• 26:38
  Have a learning health system where every interaction with every patient gets into AI systems that benefit every p- other patient, not just the patients of that physician. Uh, those are some of the fundamental things.
• 26:53
  Now, new tools are coming online, so we'll be able to measure more, more. One thing I would predict, I'd be shocked if in 2030
• 27:03
  we did not have an all-omics test for every patient as the annual physical along with an annual full body image,
• 27:16
  and that would be way more effective than today's annual checkup which is, which lots of studies show is pretty worthless, just expensive, not useful. Um, so that'd be my vision for 2030.
• 27:31
  If you had 10,000 biomarkers, you'd also get past trying to pretend humans can look at it, and we'd have AI pattern-based analysis, um, determining... Still very scientist. So complex doesn't mean it can't be a science.
• 27:50
  There is a field called complex systems theory and complex systems science. I'm a huge fan of the Santa Fe Institute approach, which is when I first discovered this approach to medicine.
• 28:02
  There was a researcher who's now at Harvard Medical School, Walter Fontana, asked me the question when I was sort of a summer intern there in 2001 or so.
• 28:14
  He said, "I'm trying to study 3,000 metabolic pathways in the body simultaneously as a complex system. And what's the math of that? What's the science?" So those kinds of things will be very important, I hope by 2030.
• 28:31
  If not, timing is very hard to predict.
• 28:34
  So another, um, picking up on this thread, uh, uh, another person from the audience, Catherine Cheney asks, um, about the most significant developments in digital health over the past year, but specifically with implications, uh, for low and middle income countries.
• 28:51
  Um, you often talk about global opportunities to transform health. Um, I'm wondering if you could, uh, kinda take us in that direction of the impact for low and middle income countries here. Well, so when it's digital
• 29:08
  and when it's remote, it will be a lot cheaper. When you can do 80% of what a physician does, you'll scale the number of physicians. You know, we have crude solutions to this problem.
• 29:24
  In Tanzania, for example, last statistic I saw there was one physician for every 50,000 population. Well, you've got a problem then.
• 29:35
  Uh, now if you scale it 80%, you can get to one physician for every 10,000 people equivalent. Um, and, and there are crude solutions to this.
• 29:47
  I heard a while ago that Mayor Bloomberg was training high school students to do C-sections because if a woman needed a C-section, it was almost always a death sentence.
• 29:59
  And so once they saw 10 of these, they could maybe do the 11th one themselves if there wasn't an experienced physician at hand. That is one brute force approach. Um, it's,
• 30:15
  uh, not requiring 10 years of medical school,
• 30:19
  but saying, "What is the minimum level of training needed to get somebody to do a reasonable job, not the best possible job of, and not all the corner cases of emergencies, for a C-section in Tanzania where there just isn't gonna be more resources?"But my son has a startup called Curade that's working on a primary care AI physician assisted by human physicians, because in California you can't prescribe medication without a physician.
• 30:53
  But can you do most of the work of the patient visit? Do most of the findings?
• 31:00
  And they track percentage of findings that the AI determined before the physician tests stepped in to say, "Here's additional questions I have, additional findings I need. Here's what I need for differential diagnosis."
• 31:16
  But that kind of approach to primary care, over 10 years, I'd be shocked if it wasn't possible to do free primary care, almost like Google Search is today, to every person on the planet.
• 31:30
  And the only way to do it is to do it through something like that. And I'm mo- most concerned about the accessible part of accessible healthcare.
• 31:41
  Most people talk about it, but if you need 10 years of medical training before you can provide that service, instead of relying on an AI with a human relationship person familiar with the area,
• 31:57
  uh, you're not gonna scale the number of physicians in Tanzania or India or, or anywhere else, whether they are primary care physicians or oncologists.
• 32:07
  You know, what's interesting is it just feels to me, and I haven't seen a startup in this area, oncology should be the easiest possible discipline to turn into an AI oncologist.
• 32:19
  So if any listener out there has ideas on doing this, I'm all ears. But my point is scalable medicine will be cheaper medicine.
• 32:30
  And what's interesting is, and I s- wrote my paper on 20% Doctor Included six years ago now. It was a 100-page thesis that's on our website. Uh, it was called 20% Doctor Included.
• 32:45
  I literally said, "I hope 25 years from then," so 19 years from now, 20 years from now, "I will better get better cardiac care in a village in India than I get at Stanford."
• 32:59
  Because Stanford will still rely on the physicians they just hired, cardiologists they just hired, and they'll be experts, while the AI will have 25 years of learning and improving every single day.
• 33:16
  As I tell- You know-... the Cur- Curade folks, your goal should be reduce physician load by 1% a month for the next five years or 60 months. Exponentially, that's hugely valuable in the model.
• 33:32
  So you're not cold cutting the physician out anytime on day one, and you have the guardrails and safety of human supervision, and then continual improvement. Sorry, long answers to short questions. No. I mean, I...
• 33:49
  You could go on and on. I would love to continue hearing more.
• 33:51
  But I wanna shift a little bit to something you'd said, by the way, in, in, in several of our past conversations on this theme around scaling up the capabilities of nurses and PAs and others in the- I call it upskilling.
• 34:07
  Upskilling. Upskilling professionals. Yeah.
• 34:08
  So, so, so has there been enough innovation and, and, and true companies being built to commercialize and help scale up these incredibly already skilled individuals so that we can see some of those results sooner?
• 34:25
  And are you happy with that? Well, I, I, I don't think we've made enough progress, but one of the unusual phenomena, if, if I move to physics for just a second. There's a point
• 34:43
  with sufficient complexity and sufficient component pieces, a systems system becomes autocatalytic. Um, and, and, you know, there's a thesis that life originated when sufficient complexity arrived in an area,
• 35:01
  possibly the volcanic lands. Um, I think we are seeing lots and bits and pieces being developed by lots and lots of companies and entrepreneurs and scientists.
• 35:15
  I shouldn't leave out the researchers who are playing a critical role in this. At some point, there'll be enough pieces where the system goes autocatalytic and you see exponential adoption.
• 35:31
  I think in the case of, uh, AliveCor or the Livongo. By the way, AliveCor has a million users who regularly use their ECG device. Think about it. They get 100,000 ECGs or more a day. A day.
• 35:47
  Nobody can match that scale of data, and most of those ECGs are because somebody's feeling something. I'm feeling odd, or I just exercised and my heart's pounding, or I just ate, or just woke up feeling X or Y.
• 36:04
  That's when those ECGs happen, not in the way the traditional ECG happens, which is I feel something, I call the doctor, I get a appointment a week later. Three weeks later, I get an ECG. The symptoms may have gone.
• 36:19
  So this kind of real-time contextually rich data, all this is happening. I expect at some point it becomes autocatalytic. One can look to the internet. It started in the '70s.
• 36:37
  Uh, frankly, I spend, started spending a lot of time in, in it in '93.But the browser was born in '95 when we... When I was at Kleiner, we first invested in Marc Andreessen and the internet browser.
• 36:53
  And it was two or three years past that, that the web became autocatalytic, and probably 10 years after, the iPhone came, when mobile became autocatalytic.
• 37:05
  I expect in the next decade enough components will be there, the system will be rich enough, uh, we- we'll see this autocatalytic system phenomena. It's why Silicon Valley works and it's been hard to replicate.
• 37:21
  There's such a diversity of talent in Silicon Valley, despite the recent trash talk, uh,
• 37:29
  uh, that, that diversity of talent, points of view are needed for a system to be autocatalytic, and I think that'll happen in healthcare, if I were to add another forecast for 2030.
• 37:43
  No, I, I think that's perfect, uh, transition into Silicon Valley, and specifically the mindset and the resiliency that both entrepreneurs and investors have.
• 37:53
  Um, you talk a lot, um, again, historically as well, about the importance of staying alive, quote unquote, you know, as a company long enough for luck to kinda help you connect the dots and- Yes... see success.
• 38:07
  Can, can you dig in a little bit around mindset and resiliency as a component of not only your decision-making initially with getting involved with a company, but also your continued support of the investments you've already made and the importance of a, of a role that plays in everything?
• 38:24
  Yeah. I'm sure we have a lot of Europeans in the audience. Let me start with that analogy. My general experience is European venture capital has worked for linear innovation.
• 38:42
  And the reason it is, they turn every radical idea and find ways to reduce risk to the point that the probability of success goes up, but the consequences of success are marginal to inconsequential.
• 39:00
  You know, moonshots is a theme here. I like to say I'd rather have a 90% chance of failure and a 10% chance of changing the world. Uh, most very large innovations
• 39:16
  necessarily mean you have to take large risk, whether it's a technology risk, a science risk, or a business model risk. And you have to have that mindset in everybody in an ecosystem.
• 39:34
  If you call the senior executive in Germany and say, "Leave your 25-year tenure at Pfizer and join this five-person company," you're not likely to see that answer.
• 39:48
  You are likely to see that answer be yes in Silicon Valley.
• 39:53
  So whether it's the scientist thinking outside the box, the entrepreneurs thinking outside the box, management, or investors saying, "I don't care about these small outcomes. Go for the moonshot," um, that's a culture.
• 40:09
  So Silicon Valley is a culture, not a place. It can be duplicated in other places, but it's been hard because almost all these axes of investors, management,
• 40:22
  employees wanting to join a startup instead of working at a big company, um, entrepreneurs, scientists, all of them have to have this mindset towards more risk and the willingness to fail.
• 40:38
  I like to say the principal reason I'm successful is my willingness to fail gives me the ability to succeed, and without that, I, I probably would be at a corporate job at IBM.
• 40:54
  You know, if the- Or whatever computer programmers who join IBM and stay there do. Yeah.
• 41:00
  Y- y- you know, h- how do you manage that inside your fund, eh, your fund investments and portfolio companies where you see an entrepreneur's mindset, uh, or resiliency wavering as they get, you know, the resistance from the industry, investors, and otherwise?
• 41:16
  Well, look, it's very, very important that an entrepreneur find the right kind of investor. Look, when an entrepreneur can't afford to fail, they want a low-risk strategy, and it is the appropriate strategy sometimes.
• 41:33
  I'm not saying moonshots or large innovation is the only correct strategy. You can do totally incremental things and build a business over time and do it much more safely than if you're doing a moonshot.
• 41:50
  And there are people who can't afford to take that risk, even if he wanted to. There's people whose personality is geared towards less risk.
• 41:58
  And those are perfectly good ways and good things to do, and you want an investor who matches that goal. Now, we don't like small outcomes.
• 42:11
  Making three times our money over five years may be a good IRR, but totally uninteresting to me. I consider that a failure, right? But I don't mind complete flame-outs, which would be bad for the entrepreneur.
• 42:28
  Um, but there's enough entrepreneurs that absolutely want to do that and are willing to take that risk and try it.
• 42:38
  And so it's very important an entrepreneur pick their investor based on the personality of the entrepreneur and what's a good match in an investor.And there's probably a flavor of investor, all good investors, for every class of risk-taking than entrepreneur or goal or personal ambition the goal wants to do.
• 43:02
  You know, it's very easy for an entrepreneur to be very successful putting a physician visit on a video, call it TeleCare, and be successful. It has been successful. But it is not innovation.
• 43:18
  It's a little bit of business model risk, a lot of operational risk in getting costs in line, but it's no radical change to the supply of physician talent to the planet.
• 43:31
  If that's your goal, then you'd approach it very differently. For example, anybody who says they're doing video visits and AI, it's just not possible. It's just not po- You can't claim both.
• 43:45
  It is mathematically wrong, uh, unless you're doing unusual things like capturing every conversation, translating it, analyzing the dialogue, uh, with AI and that kind of, uh, thing. But
• 44:02
  set your goals and find an investor to match you. I like, I like changing healthcare fundamentally. I wanted- That's why I, I'm generally not a fan of entrepreneurs.
• 44:16
  The best way to be a safe entrepreneur is start with a CPT code you wanna bill to. [laughs] I personally hate that, and I'd like to say, "Tell me what you're gonna do for the patient first.
• 44:27
  What radical innovation are you gonna make useful for a patient? And then along the way, we'll figure out billing codes and revenue models and that kind of thing."
• 44:38
  This is where I think the verticalized healthcare system where verticals like di- chronic care can be done outside is gonna be very useful for radical innovation.
• 44:49
  So picking up on that theme, mo- most of the world is, is outside of the realm of, of billing codes when it comes to health. Um, why do you think
• 45:00
  more investors aren't focused on the, the global opportunity to transform health with billions- Yeah... of people in the world, health being universal.
• 45:11
  Um, even in Silicon Valley, there's not many investors focusing on, on global investment. Yeah, no, it... And I think that's sad.
• 45:20
  We are starting to see it happen, but it is sad that more investors aren't willing to do this, but I can see why. I fully understand. There's large chunks of market
• 45:36
  documented by various analysts that say, "If you do this in cardiac or this drug or this that, the US healthcare system, being the largest number of dollars in the world, will do this."
• 45:51
  But realizing that maybe 5% of the world's population is subject to the FDA, um, the other 95% is very, very interesting, especially for large innovation because they have no alternatives. They know, you know.
• 46:08
  Now, Europeans is different because there's, uh, there's universal payer kind of models in many parts of the world. Um, but finding the first MVP niche segment
• 46:26
  can allow for radical innovation. You know, the people are experimenting with mental health globally.
• 46:35
  Um, interestingly, mental health, there's a lot of self-pay in even in the US, but that's an area where people are experimenting with and the need is large. But, uh, we will, I hope, see more of that.
• 46:50
  But I, I understand why investors are focusing on billing codes and market segments and fitting into traditional clinical practice or provider systems. Uh, I do think it's very limiting to large innovation.
• 47:09
  We'd love to- I mentioned a NASH test, you know. We started with thousands and thousands of biomarkers per person. That can fit into a billing code, but we didn't start with the billing code.
• 47:23
  We said, "What if you had 10,000 data points on every blood sample and medical records? What would you discover?" That's where that investigation started.
• 47:35
  And then we said, "Okay, now, shit, we need to make a business out of this. What can we build? Where can we get our first traction?" So we've got just about 10 minutes left. There's so many extraordinary questions.
• 47:49
  Um, we wanted to make sure to, to cover- I'll try and keep it short. [laughs] Yeah. Well, this is great. Um, so much wisdom to share. Thank you. Um, would love to learn a little more about you as a, as a person.
• 48:03
  Um, for example, where, who inspires you? Where do you look for inspiration when you think about innovation?
• 48:10
  Are there, there are books or articles or things that you read regularly that, that really cause you to expand your thinking or think differently? So I've been doing innovation and innovation only for 40 years. Um,
• 48:28
  I like to say I've probably failed at more things than any other person I've ever met in my life. Literally, I probably had more failures than anybody I know.
• 48:42
  But the few successes, um, establish, gives you the confidence, makes it worthwhile, um, maybe establishes your reputation where other entrepreneurs that are like-minded call you.Um,
• 49:00
  I would say there's reading to do. I, uh, Loonshots is a very good book to read. Eric Topol has a couple of fairly good books in, specifically in healthcare. I think Eric and I think alike a lot. Um,
• 49:18
  but, uh, I wouldn't say in general there's a lot of role models. Look, I knew Larry, uh, Page and Sergey Brin when they were, uh, graduate students at Stanford.
• 49:36
  They imagined this other world, and then made it happen. Elon Musk has done that in space and in electric cars. Um, Jeff Bezos has done it in retailing and now in computing. Um,
• 49:56
  in fact, [laughs] uh, uh, this is a sidelight. When I looked at the last 40 years and said, "Has one, one large innovation, broad scale major innovation,
• 50:11
  come from somebody who's an expert in this area?" And deep f- pharma may be one exception, but I've not seen one large innovation that's has societal impact
• 50:27
  im- implemented by anybody who knew the area they were innovating. Think about it. No expert has innovated broadly in an area to have large impact. Plenty of incremental innovation.
• 50:43
  Intel's very good at 16 nanometer, na- nanometer silicon to nine nanometer to seven nanometer, but it's not radical. They haven't said, "Hey, something other than silicon could really work."
• 50:59
  So whether it's retailing, as in it wasn't Walmart. In media, it wasn't Fox or CBS or Rupert Murdoch. It was, uh, Twitter and Facebook and, and Netflix and YouTube. Uh,
• 51:15
  i- in space it wasn't Boeing and Lockheed. So there's no innovation I could come up with
• 51:22
  that came from somebody who is an expert in their area, and this is where I go back to the more you know in an area, the more limited you are, you are in thinking what you can do, as opposed to what you can really do.
• 51:36
  And I think this is the constraint. Let me make one other point since we are running out of time, uh, that's very, very important. I like to say
• 51:48
  there's a huge difference between a zero million dollar company and a zero billion dollar company. These are, uh, and the difference is not the business plan for next three years.
• 52:03
  The difference in how you think about what you're doing. You know, your business plan can be to get to Mount, uh, to the base camp of Mount Everest. But do you pick
• 52:16
  any base camp where you can get revenue and profitability, or do you pick one that leads to Mount Everest, that then leads to camp one and then to camp two? Do you have that long vision?
• 52:28
  And the teams you assemble, can we just hack something together or build for the long run?
• 52:35
  So build the human resource for your vision while you build the tactics in the next 12 months plan for your base camp, where you get revenue and stability and can convince investors to join you.
• 52:49
  So these companies are very different. This mindset is very different, and I've just discovered this along the way. Looking at why can't experts or really experienced people innovate in an area.
• 53:02
  Because they come with a lot of biases about how things are done or not done. Well, this is good news, 'cause Steve, neither Steve nor I came from healthcare, and we're trying [laughs] to, uh, innovate in healthcare.
• 53:16
  So, um, Steve, I'll turn it back to you. And, and, and, and, and let me say one other thing then. You know, it's important to assemble teams.
• 53:23
  Becomes the most in- important ingredient of whether you're building a zero million dollar company or a zero billion dollar company. And I'm not saying don't get some of the experts on your team,
• 53:38
  but the guiding light, you as the entrepreneur, you should have a vision to disrupt the world. And when the experts say, "It can't be done that way," you ask, "Why not?"
• 53:51
  And ask for first principles thinking, not experienced-based thinking. So the people who tell you why things can't be done or how they're done can be extremely important in building that billion dollar company.
• 54:07
  Elon Musk did hire a lot of experts from the auto business, but he never let it change his vision. He just used them to do the things they knew how to do well when he needed that done.
• 54:20
  And I think that's a very good model of how to use expertise. Now, deep science may or may not be different. I've thought a lot about it and don't have a strong view.
• 54:30
  Like, uh, how do you do immunotherapy, a new kind of immunotherapy? Um, I, yes, it's possible. But again, the entrepreneurial mindset I think eventually rules.
• 54:47
  Well, that is a wonderful bookend on this conversation. Uh, Vinod, there's 50 unanswered questions that we're gonna get back, uh, to, because we're gonna invite you back m- m- much sooner than we usually do.
• 55:00
  We're coming back this year for another episode. I wanna thank you.
• 55:04
  I also wanna thank all of the people who tuned in, who submitted questions, and I really, you know, can't thank you enough for the torch that you hold for all entrepreneurs and investors to believe that the future of healthcare is incredibly bright and exciting.
• 55:22
  And the innovation that is gonna be possible within the next 10 years is gonna be as a result of investors like you and entrepreneurs like the ones that are on today's call.
• 55:34
  And most importantly, um, I want you to have a happy, healthy new year, and stay safe. Um, it, I wanna, uh, as we wrap up, thank everybody, especially for joining into all of our fireside chats.
• 55:47
  These are weekly events that we hold. Our next event is coming up this Thursday with our Health Transformer showcase on our nutrition and fitness moonshot.
• 55:57
  And, uh, I wanna make sure everybody's signed up for not just that event, but all of our events at startuphealth.com/events. So again, Vinod, thank you so much. Thanks everybody for attending.
• 56:07
  Unity, another episode under the wraps. Take care, everybody. [upbeat music] Startup Health invests in health transformers from around the world who are committed to achieving audacious health moonshots.
• 56:24
  If you wanna learn how you can join this community of entrepreneurs, or if you wanna connect with one of our 330 companies, go to startuphealth.com. Thanks for listening to Startup Health NOW. We'll be back next week.
• 56:37
  [upbeat music]