Subscribe
Share
Share
Embed
Work Bench for raw audio interviews
Daniel, welcome back.
Daniel Kraft:Great to be back, Moira.
Moira:Now, I have to say, we talk about, whether it's technology or health or gadgets or anything. We say, is it now? Do I have it now? Is it near like it's gonna happen soon? Or is it next?
Moira:Like, we gotta wait for it. It's brewing on the horizon here. And what happened now you tell me is that OpenAI has announced ChatGPT Health. What is it? And is it now near or next?
Daniel Kraft:Well, it used to be near, and now it's now. So many of us, as we've talked about in prior episodes, you know, we used to go to doctor Google, now we go to doctor GPT. And so a lot of folks are already using these LLMs as ways to get questions about diagnosis or symptoms, sometimes with very validated answers. Sometimes across the board, there's still hallucinations and challenges. But this week, January 7, 2026, OpenAI an OpenAI announced that they're launching ChatGPT Health, which is going to be sort of a dedicated environment for health and wellness.
Daniel Kraft:Again, one of the major uses already today, but now it can be built a bit more with privacy in mind, with the ability for you to upload in a much more secure way your medical records. It could be your labs. It could be your x rays. It could be your wearable data, and sort of be a real sort of dedicated space for health and protect your information, but also connect that over to your healthcare systems, because they also introduced OpenAI for healthcare, where it's, you know, gonna be more plugging in the OpenAI architecture for hospitals to manage data, maybe do clinical trials, the workflow of your clinician, your pharmacist. So really kind of exciting opening up of the aperture to do your sort of own self enabled health.
Moira:So let me ask you some questions about this. Number one, right now, I go to the, my computer or my phone, and I go, hey. What's this symptom? What's that? Do I have you know, is this right?
Moira:Is that right? Now I can ask those same questions or very soon in the in the near context. I can ask those same questions, and it will have the information about me to answer in a very relevant way. Is that part of it?
Daniel Kraft:Right. We all wanna have personalized health and medicine, That's and all very contextual. Right? We have all these new contexts. We've talked about measuring, you know, your genome and your microbiome and your digitome and your exposome.
Daniel Kraft:Well, now your GPT and your sort of personal health agent, which will be built on these sort of systems, will be continuously sifting through your data. When you ask a question, about the fever you had, it might know that, wow. You just got back from, Western Africa. Did you take your malaria prophylaxis? Or, if you have a bad cough, is the air pollution level higher?
Daniel Kraft:Are you at risk for asthma based on the weather and the pollen count? So putting these things in context. And then what's also exciting about these sort of systems is that they'll talk to you, you know, more or you've got a, you know, an advanced degree, engineering and beyond. You know, you need different levels of context whether you have an eighth grade education or you have an MD or a PhD in bioinformatics. And so the we're opening up the aperture where health information used to require access to a clinician or a medical school library, and now that's being democratized.
Daniel Kraft:And that's it's a really exciting shift. It's a shift, health for and health as wealth for for everybody.
Moira:And I have to say when you say, well, you just got back from a certain place, not because you told it, but because it's into your travel accounts. And the weather and what's been happening in the weather with pollen count and temperature and everything else, well, it knows where you are. So it's more than just your health data.
Daniel Kraft:Yeah. We talked about artificial intelligence, but what we really we really, really want is contextual intelligence. Knowing, you know, based on your integrated medical records or where your labs are, what those small symptoms might be. The exciting part is I think we're all aware that we don't have a health care system, we have a sick care system. Most of our health care happens when we present with conditions that we're often burbling along for a while.
Daniel Kraft:And now in our era of continuous data from our voice, from our wearables, from our environment, from our refrigerator, which can tell you're cheating on your diet, that integrated into a sort of a health GPT or chat GPT health type environment, which feels a little more constrained and can take that information over time, knowing where you are at your baseline, can really become a tool for all of us, whoever we are, to to really get engaged with our health and hopefully optimize it, not just wait for disease to happen.
Moira:But we still have to talk about privacy. I mean, we've got HIPAA, h I p a a, not h I p p a as many people misspell it, which is the Health Information privacy and protection act. Boy, I don't know. We'd have to go and look that one up, but it has to do with that's what protects that's what protects our data at the hospital, data at our doctor's offices. It's like it gives us that privacy.
Moira:Do we have anything like that now that will protect this information?
Daniel Kraft:Well, HIPAA, which is a well meaning regulation from the predigital era, is often I mean, we're all familiar with filling out the little form.
Moira:Your paper is safe.
Daniel Kraft:Yeah. It it in my view, and many others, it's often sometimes a bit of a barrier to care. I mean, I've seen patients die with their privacy attached because no one had signed the form and faxed it over to get their labs or their medical record from another hospital in time. I think there's a balance between health data privacy and safety and security, with also the ability for each of us to be data donors and to share your data from your particular health condition or medical condition so that we build better insights for all of us. So HIPAA is more regulating what your, let's say, your clinician can share, but you can put anything you want on the on the web.
Daniel Kraft:It's not a HIPAA violation for you to, you know, send me your medical data, but if I'm your practicing physician, I can't share that with others. I think it's a rule that needs to be optimized or or modernized. But with chat GPT type systems or chat GPT health type ones, they are gonna be, and many of them already are HIPAA compliant, meaning they have certain levels of security and safeguards. But what's interesting now is they can start to connect the dots with your own permission. You have control over your health data.
Daniel Kraft:You might wanna be a data donor and be part of a clinical trial or be identified for a clinical trial. You might wanna connect to your Weight Watchers app, and if someone's on a GLP-one, be guided with your AI agent about what's next for your meals or maintain your muscle mass. It might want to be recommending, you know, what hike is around the corner based on where you live and what your goals are for how many steps you want that week. So and tied to your food and your shopping. So health is such a continuum.
Daniel Kraft:I think what's exciting about the AI and contextual health era is it can put it together and make it match you, not just the the average patient.
Moira:And I think what you're also telling me is even though they're coming out with this, you know, ChatGPT Health and ChatGPT HealthCare, the whole idea is that this is all version one point zero. We're gonna find out what works for people, what doesn't, what they can do, and how they're using it that they may not have expected when they designed it.
Daniel Kraft:Right. It's early days for only nearly three years and a bit since GPT was launched into the world. It was existed a bit before that, but now it's being used for almost everything. Health is a great application given the shortage of clinicians, the challenges of getting your doctor's appointment, or, you know, just this last week, the first example of an AI platform is able to prescribe medications in a certain state. We'll see how that works out.
Daniel Kraft:We'll see the advent of robotics get more AI enabled. It's sort of you know, we've talked about this framework of now near next. The the now is really that sick sick care model we talked about. We're entering a bit more of a health care era, and it's shifting to one where you can do more self health and use these sorts of tools to create health. So moving from reacting to a disease to actively building capability, health literacy, and the tools that will match you to manage your more day to day health and and vitality and and, optimize your health span and that for your community.
Moira:Well, thanks for coming in, and stay healthy till I see you next time. Okay?
Daniel Kraft:So you can always text your chatbot if I'm not available. Alright. Thanks, Moira.
Moira:That was a good one. It was a good one. You slowed yourself down.
Daniel Kraft:Short and sweet.
Moira:And you were telling him things, and it was all it was it was still eight minutes.
Daniel Kraft:Okay. It's a good it's a good context. Okay.
Moira:It doesn't seem yeah. The time is, you know, time is relative. You know, it's like it used to be how much are getting in here? You know? Now we got this, and you stayed in the same thing, and you had your nice doctor leave bedside manner.
Moira:Yeah.
Daniel Kraft:I'm coachable. I'm coachable.
Moira:You are coach you're better than coachable because while I'm coaching you or anybody else is talking to you, which is it's it's like you're like, oh, yeah. And you're like, you can see it not just like I'm taking it in, but I'm also folding it over, and I'm coming up with new stuff, and that's the best.
Daniel Kraft:And we reiterated some things. So for a listener, it's it's they send me something they heard twice.
Moira:It's good. It's good. It's not bad. It's not a bad thing because and sometimes you'll hear me strategically say, especially in biotech, you said this, and I go, now let me get this straight.
Daniel Kraft:I do hear that.
Moira:I kinda repeat. Is it this true? You see that? Well, I don't do it because I'm can't think of the next question. I do it because they need to hear it again, but they can't they really wanna rant wind up.
Moira:So they're okay. So how about I start this one next one with hey, Daniel. Great to see you.
Daniel Kraft:Hey, Mara. Good to hear you.
Moira:Don't tell anybody. You can see me too. Okay? Our secret. It'll be our little secret.
Moira:Our secret. Our secret. Now lately, I realized you have been talking about, now, near, and next, and it's a context that in our conversations, both on air and off air, we keep talking about. So let's talk about this now, near, next. I think you've been you've been bubbling on this a while.
Moira:You've been working on this a while. What is it?
Daniel Kraft:Well, sort of a framework we've used at NextMed Health to sort of help people see the arc of the possible. Where are we and where we're going? So I like to say, you know, what do we have now? What's near near term, like next couple years? What's coming next in the, let's say, next five and ten years?
Daniel Kraft:And also what's needed to get these sort of new mindsets, particularly in health care, to emerge. So one one one that might resonate with the listeners is that health care for most of us is now very episodic. Right? You hopefully get your primary care doctor visit in once a year and get your labs tested and your weight and your blood pressure. Often, that happens in the four walls of your clinic or, god forbid, your emergency room.
Daniel Kraft:Where we are moving to the sort of near is continuous data that could enable us to be proactive and preventative, meaning the data from your smartwatch or your labs, which could be picked up from the camera on your laptop, which you can now look at your face and pick up your blood pressure and your heart rate and maybe even soon your blood sugar. That's sort of the episodic to continuous now to near. And where that takes us is moving from sort of snapshots and signals to capturing the trajectory shifts, early that might indicate a problem, before you sort of arrive on diagnostics day. So what's coming next is not just this continuous proactive health care, but one that becomes really predictive and preemptive, meaning we can pick up problems super early and push you back to true true health.
Moira:Now I'm sort of fascinated because I always feel like, you know, the future's coming at me. You know? And we talk about technology frequently in in just ways that it's kinda wishing and hoping. Some people like, oh, yeah. As soon as they get a new piece of technology, it's gonna do this.
Moira:And sometimes I know it's not. You know? We do projection onto technology around us, which we would like it to do, but it can't. Or it can do it, but we wanna ignore it. We got too much going on in our lives.
Moira:So tell me about, in terms of people in now, near, and next, with your experience, where do they get confused? Where do they get alarmed, and where are they really excited about it?
Daniel Kraft:Sometimes people call me a health care futurist. I'm actually a bit more of a nowist. Right? I thought it's also, pretty incredible what's here. I was just, at the consumer electronics show in Vegas, and there's a huge digital health and health section on the consumer side that's really been exploding.
Daniel Kraft:So the art of the possible of what's here now often comes in some of these consumer elements which aren't quite perfect. The dots aren't connected. They're not paid for by your health system. Your doctor or nurse or pharmacist certainly doesn't want to look for, isn't paid to look at these, but sometimes it's pushing the envelope because the consumer, each of us, kinda knows what we need, and that helps drive what's next. And, so I think it's in in many cases, it's not about the technology, but it's aligning the incentives, the payment models, regulatory.
Daniel Kraft:The FDA just actually announced that they're going to sort of be loosening some of the restrictions on, consumer products, meaning your, smart wristband can maybe measure your blood pressure. It doesn't need be perfect anymore, it doesn't need to be equivalent of a medical grade, it can at least give you an indication that, hey Moira, your smartwatch is telling you from your signals that your blood pressure might be a bit high. Might want actually go and get a formal blood pressure check. So it's about connecting those dots, having a bit of a workflow for you as a consumer, as a human, as a patient, for me as a clinician or a health system to start to catch the signals, get rid of the noise and the the alarm fatigue, and really translate that to this era of predictive preemptive health.
Moira:Now give me some examples of of now, near, next, because I happen to know you're working on a lot of these different aspects of now, near, and next. So give us one or get or two. One or two.
Daniel Kraft:Well, now and now is that, you know, ten years ago, it was pretty magic to maybe get a heart rate on your watch or your Fitbit type equivalent.
Moira:What watch?
Daniel Kraft:Now your yeah, what watch? Well, now your watch is really a supercomputer on your wrist. With Moore's Law, we've compressed all this technology into a smart ring or a smart watch, which can measure your steps obviously, can give you a sleep score and analyze your sleep stages, can do your basically blood pressure today, in a rough way. It can soon probably measure your blood sugar. It can do your temperature, etcetera.
Daniel Kraft:When you put those individual data points together, then now is it can give you a context of your health. When I'm wearing my smart ring, my smart watch, it now comes with a little bit of an AI coach and goes, wow, looks like you had a bad night or a bad few days. You're not in your best state. Take it easy today. Or wow, your heart rate availability is up, your resting heart rate's down, you get a great night's sleep.
Daniel Kraft:Today's a really good day to to push it, get your workout in, and maybe stress yourself a little bit and build your capacity. So the now again is the ability to collect these dots. The near is making sense of them, which is starting to happen with AI and contextual intelligence, and the next is gonna be where our sort of physical exam, our sort of primary care element will be agentified. It's always gonna be sniffing through, if you give it permission, into your digital signals. And those can be again from wearables or the ambient sensing in your home from WiFi or what your laptop might pick up on how you're typing or other changes in behavior and enable you to have that sort of preemptive predictive model that's based on you, that idea of a digital twin.
Daniel Kraft:So now we're able to collect amazing amounts of data and commoditize devices. The NIR is next generation sensing, like continuous blood pressure and blood sugar on your smartwatch or Fitbit or wearable or smart ring. And the next, be able to contextualize that and use that continuously in the era of what we talked about of episodically continuous to preemptive, predictive, and participatory health.
Moira:You know, I'm still amazed that I can log on to my, health system, and I can see all the various tests I've had in the last, say, five years. And what were all the scores? Because people say, well, how are you doing on this? It's like, you know what? I don't really I don't really know.
Moira:Cholesterol, I don't really know. And there's an and it's not just cholesterol now. It's all these various things. It's like, well, let's just go and look. You know?
Moira:The it's amazing. We have so much more data about ourselves, much of which we don't even look at. And the idea that we could start to collect it on a continuous basis and have it be in context really speaks to your ability to maintain health and and to see how well, for instance, a recovery is going along. Say you had surgery or you were in an accident. How well are you recovering?
Moira:This changes your attitude about yourself. It's pretty amazing.
Daniel Kraft:You've got not just sort of an attitude ability, but a ability now. You can do this today. We talked about in a prior episode that CHATGPT Health is coming online. But you can go to a regular agent, Gemini, Chat GPT, Grok, and others, and take your labs that you might have gotten at Quest and upload the PDF. It could be the PDF of your labs from last week and the ones from the years prior.
Daniel Kraft:I can say, what do you see in these labs that might be relevant to my health? Some of them, for example, might be your hemoglobin A1C, is sort of a measure of your average blood sugar. Is that going up? Might you becoming pre diabetic? Is there something you want to do about that?
Daniel Kraft:And these sort of agents can start to recommend that yes, you really should start working on your metabolic health, maybe changing your diet and your exercise or getting a deeper set of labs. So you have the ability not just to collect data, but to understand the meaning of that, and not just the meaning that oh, you're on your way to prediabetes, but what is the actionable information and insights that you can do together with your health care team, etcetera, or something that you can start to do based on behavior change and modification of your diet and your exercise and your sleep that can really make a difference? And that engagement factor and the ability to pull that insight together is is really powerful.
Moira:Now give me another example, pretty quick here, of now, near, and next, because that's just one aspect you were talking about.
Daniel Kraft:Sure. So now, if you need a medication, let's say for high cholesterol, your labs come back that your cholesterols are running high, and your physician or your AI agent recommends a statin. Today, it's often just prescribed. You get that atorvastatin at forty milligrams, the standard Lipitor that many folks may have heard of. But that's not really personalized to you.
Daniel Kraft:That's the now. Everyone sort of often starts with the average dose, kind of general. The the the near is getting to be much more personalized. It'll be based on your genetics, your pharmacogenomics, because some people will react and respond very well to a statin at the average normal dose, but some people need twenty milligrams, some need forty, some need eighty. And other folks have, pharmacogenetics, meaning how their genes metabolize those drugs that might give them side effects like sore muscles, they'll fall off their statin and end up with a higher risk for heart disease.
Daniel Kraft:So that's sort of the the next is sort of, sort of the near is sort of using our genomics and information to really pick the right medication for, let's say, something as common as high cholesterol and managing blood pressure. And then the next is gonna become sort of adaptive real time precision. So personalization step one, then the LEAP is dynamic care plans that adjust in real time based on your dosing needs, maybe changes in your behavior. It could be coaching, your dosing of of of meditation. It doesn't need to be a drug.
Daniel Kraft:I've talked in prior episodes about a technology I'm developing on three d printing your personalized pills. So instead of taking, you know, six pills a day for a particular condition, it might be three d printed into one pill that integrates the right doses and the right timings for you. So that's kind of the arc. Now is one size fits all. Near is getting personalized.
Daniel Kraft:In the future, the next is adaptive real time precision where you can imagine your preventative regimen or your medication being three d printed on your kitchen sink each morning.
Moira:So how well anything works is more than just, well, how are you feeling, Joe? How are the side effects? Is that alright? Well, we'll keep you on this. Looks pretty good.
Daniel Kraft:Yeah. The now is just that subjective how you doing. Right? Yeah.
Moira:But for
Daniel Kraft:for mental health, for example, you know, a huge issue, and a lot of folks are on SSRIs, etcetera. How do you measure mental health? It could be done through a survey, which is kind of kludgy, but now you can measure mental health based on movement and sleep and voice as a biomarker that can tell if someone's depressed or manic or having a schizophrenic episode. And those then can be part of that real time adaption to adapt someone's mental health interventions, whether those are a drug or behavioral or therapy or, you know, sending you to your chat a GPT therapist as well.
Moira:Well, Daniel, thanks so much for coming in. I hope I see you in the near term and not in the next term.
Daniel Kraft:I'll see you in the near term. And for for the listeners out there, you know, don't wait for this next to arrive. A lot of the near is here. You can start utilizing it. It may not always be paid for.
Daniel Kraft:It may not always be perfect, but you can go out there and find solutions to help you manage your sleep or your mindfulness or your exercise or help you connect to the right care. So the time to be engaged as a consumer and as a patient is now, not next.
Moira:Great. See you soon.
Daniel Kraft:See you.
Moira:Good. That was a good one. See how good that is? Boom. You're a new man.
Moira:You're a new man. Because I'm gonna have
Daniel Kraft:my coffee. Let me get my coffee now. No
Moira:coffee. No coffee.
Daniel Kraft:Oh, I thought it was hot. It's not. It's empty coffee. I think
Moira:you're You didn't get your coffee?
Daniel Kraft:But your we're all that's fine. We can do about it.
Moira:I'm like let me see
Daniel Kraft:if she's up.
Moira:We'll let you drink coffee. We don't we don't like we don't we're not really really kidding. We've got no coffee.
Daniel Kraft:What is a good drug? Okay. Let's just do the next one, and I'll I'll get you next. Okay. Alright.
Daniel Kraft:I think we'll do hospital to home as a next, but we can still use the the noun here
Moira:next as
Daniel Kraft:a setup, I think.
Moira:Okay. What what about the health care was like a Waymo?
Daniel Kraft:Oh, we can do that. Let's do that after the hospital to home.
Moira:Okay.
Daniel Kraft:I think.
Moira:Okay. So that's good. Alright. So let me just put my my notes here. This is gonna be three.
Moira:Well, Daniel, good to see you.
Daniel Kraft:Great to be back.
Moira:Daniel, I have to say that in the last, oh, I don't know, twenty years, ten years, something like that. It used to be that if you had an operation or you know? And sometimes a significant operation, you could be in the hospital for a week to two weeks. And now all these things are same day surgery. Same day surgery.
Moira:You go into the morning. You're out by sometime during the day, not even late in the day. And, and these are major surgeries. And when you go home or you get a knee replacement or something like that, a hip replacement, they send you right home. Well, you still need care.
Moira:And you mentioned something earlier. You said hospital at home. What is that?
Daniel Kraft:Right. The term sort of hospital to home or hospital to homespital is really becoming an exciting part of the now near next of health care. Right? You know, if you think way back, the idea of having a computer in any home was a crazy idea. And now we're starting to think about can you almost have, almost on demand a hospital at home?
Daniel Kraft:And the the need is significant. Right? We have an aging population. No one likes to be in the hospital, as you mentioned, things like, total knee replacements, even hip replacements, and both my parents have had a total knee and a total hip in the last couple years. They got sent home the next the same day or the next day because we know folks recover better at home.
Daniel Kraft:It's safer. We don't get some of those what are called nosocomial infections, some of those bad bugs that move around hospitals when you're at home. People mobilize more, they're happier, and it's far less expensive. Know, the the the price of even a mid tier hospital bed is, you know, much more than a five star hotel and an exotic resort. So the ability to change what we do and where we do it is part of this connected age of health medicine and technology.
Daniel Kraft:And so just a couple of examples of what's shifting. A lot of this work is being pioneered at my old, training hospital, Massachusetts General Massachusetts General Hospital. And at MGH, there's a colleague of mine named Jared Conley. He he's a ER doctor, but also really focusing on this new era with his colleagues of how do you bring care, used to require the hospital bed and nurses and monitoring to your home? That's becoming reality, not for an intensive care unit, but diseases where it might be an antibiotic for a common pneumonia, or treating a skin infection like cellulitis, or someone who needs oxygen with a little nebulizer for asthma or COPD exacerbation.
Daniel Kraft:And they're learning how to transition. Sometimes you discharge someone from the hospital, not to a skilled nursing facility, but to home, or you never admit them to the hospital in the first place. You admit them into their own home connected to clinicians and nursing care and technology that can really, in many cases, give them superior outcomes to what used to happen in a hospital at much lower prices.
Moira:Oh, so that last part there. So you're at home, but you're somehow connected to the capabilities of a hospital remotely so they can be watching you and say, woah. Woah. We got a problem here. Or I mean, is do I have that right?
Daniel Kraft:Yeah. A lot of what really happens on a sort of general hospital ward is sort of monitoring. I can't tell you how many times when I was a resident and fellow and attending in a hospital, we'd wait, let's one more day to track the temperature or the EKG, and now in the sort of sensor revolution, you know, your over the counter or prescribed wearable device does a really good job of tracking your heart rate. It can measure your EKG. It can tell your position.
Daniel Kraft:Have you had a risk for a fall? These kind of wearables also can now start to do blood pressure and blood sugar. And so you don't need to be monitored in a four walls of a very expensive hospital room. You can be sort of monitored at home. The other term for that is RPM, remote patient monitoring.
Daniel Kraft:And as new technologies get smaller, better, cheaper, and we have high speed connectivity, and that's opening up around the planet with five g and satellite technology, it really gives you ability to measure things anywhere that doesn't have to be in a clinical setting.
Moira:Okay. So how does this all get together? Is a doctor prescribing it? Is are you just part of a plan? How does it work?
Daniel Kraft:Well, places like Mass General Hospital and others are starting to understand, first of all, who qualifies? Right? How do you select and be proactive about who's able to go home and not have a high risk of deterioration? Because you know commonly, you know, people in the hospital sometimes end up in the intensive care unit. How do you sort of reduce the risk of you know, admitting someone to home that really might be on their way to the ICU?
Daniel Kraft:So first of all, it's sort of a bit of a triage question. And then it's a bit of a training and staffing and workflow element. How often do they need to be measured and monitored? Do they need a visiting nurse full time or a couple times a day? How does the clinician virtually visit or sometimes go back to the doctor Welby days where they actually do house calls?
Daniel Kraft:And I think the work from MGH and others is showing that you get better outcomes, less complications, people mobilize better and recover and don't need to go to skilled nursing facilities as often as they do after a hospital stay, that you can start to, as we think about the world of Uber and Amazon, put the supply chain together so the pharmacy can be delivered on time. The labs can be processed from a home lab draw. If you can deliver radiology and ultrasound to the home, you can bring in home consultants like physical therapy and nutritionists. So it's about connecting a very sort of I like the aviation example, you know, how do you connect all the dots on an airline schedule and delivery and packages and gates and who boards the plane when. It's sometimes a coordination issue less than a technology issue.
Daniel Kraft:And then of course aligning how these things get paid for by insurance payers, etcetera, and how do you change the culture of clinical care and patients themselves to start to expect and sometimes even demand a hospital to home level care.
Moira:You know, I do find this is really interesting, and a lot of it is has to do with your situation. If you live alone, then you're like, well, can I really go home? And somebody's gotta I I can't have somebody pick me up. I can't have the taxi driver or the Uber or Lyft driver come and take me into my house. But the point is once I got to my house, you know, if I live alone, then, the idea of you could have anything delivered.
Moira:You could be hooked up to anything. I mean, it's just these are the days that, you know, you don't have to go shopping. Everything comes to you. And so in a sense, there's this perfect convergence of technology that enables you to do this the way you might want to and and truly be in the comfort of your own home.
Daniel Kraft:Right. It's blending together of telemedicine. We've all all probably most experienced during the pandemic. Our first maybe clinical experience on Zoom. There's the remote patient monitoring technology we talked about.
Daniel Kraft:There's the ability to coordinate teams like on the SWACS and the GPT coordination for medical management. And then that supply chain, which can now literally look up in the sky, the future's coming faster than you think by drone. Companies like Zipline are in Texas and in Michigan and companies like Google, Wing are delivering packages from your pharmacy to your front door, you know, in minutes. In Africa, they're delivering blood products that way, and organs are being transmitted in some cases by drone. But now with sort of just in time, just like you could do Uber Eats or DoorDash, you can get a medical test picked up or delivered or potentially a clinician to your home, that kind of waymo uberization of health care is coming together.
Daniel Kraft:And that's some of the dots that really enable hospital to home spital to really take off.
Moira:Well, I got a really cute doctor, one of part of my huge care team, of course. I'd love to DoorDash him to my front door. That would be I think you're onto something.
Daniel Kraft:Well, you know, now you can advertise him and turn him into an AI agent and have him available 247. He might show up on your Alexa or on your screen.
Moira:Fake is about all I got. That's the only option I really have, so I'm with you.
Daniel Kraft:Well, don't forget the robots are coming, and we won't go dystopian here. But, given the shortage of especially home health care workers who aren't paid very well, and the need, in some cases, have to not live at home and be in nursing homes and such because they can't pick up their socks or need some assistance, we're gonna see in the next decade a real explosion of this convergence of AI and robotics that is gonna help with hospital at home, but also just help with living at home and aging in place and can teleport in your acute doctor on demand.
Moira:Well, I'll tell you, as a former robotics engineer, I'm all over this. This is gonna be fun. Now we're now we're having some fun. Well, thank you so much, Daniel. I'll see you next time.
Daniel Kraft:Thanks, Mara.
Moira:Okay. Oh, good.
Daniel Kraft:We And a couple gaps here, but I think that's a good one. Right? It covered, like, in the
Moira:Oh, no. They'll they'll that's Danny will clean that right up. That's a that's the thing. You can always say, hey. Do you want me to say it this way or that way?
Moira:And he just just knocks it right out. So we get it cleaned up pretty quickly.
Daniel Kraft:Great.
Moira:Okay. So what time is it? Okay. So we've time to
Daniel Kraft:do one more.
Moira:We can
Daniel Kraft:do two more at least. Let's do the Waymo one. I'm just trying to find my little post about that because I made it funny.
Moira:Okay. Look at yourself to make sure you got all the references. Yeah.
Daniel Kraft:Promise, if you look at a reference and you fixate on that, I'm just trying to find my post on that. It was a while ago. It was a really popular one, because it everyone everyone kinda gets it. Come on. My Internet is so slow.
Daniel Kraft:You can still hear me. Right?
Moira:Yeah. Spend millions on your house, and your Internet is slow. Got new
Daniel Kraft:Internet too. I think it's more my laptop, but it's time for a new computer.
Moira:Time for a new computer's right. Yeah.
Daniel Kraft:Well, they get some sort of software bugs, and then sometimes you just it's not doing it. Sorry. I'm gonna go into the I can probably do it without, but it's it's helpful to have it.
Moira:It's better to have it with. It's better to have it with. Spend a minute or two getting it. Yeah.
Daniel Kraft:But for that last one, was that the right pacing? I think people can get get the idea.
Moira:You're all of your all of your ones. It's like I'm talking to a different man. And so
Daniel Kraft:If I hadn't had my coffee yet, I hadn't had my coffee.
Moira:No coffee. No coffee for you. It's like the soup nazi.
Daniel Kraft:Coffee for closers.
Moira:Seinfeld. No. Not yet. No soup for you. Oh god.
Moira:Too messy. No. You're doing good. But it's like you can it's like they're gonna they're gonna be able to to get it.
Daniel Kraft:So maybe on this one, I'll flip it with you a little bit. I said, you live in San Francisco. Have you taken a Waymo yet? And and Yeah. Used the robotics thing.
Daniel Kraft:It's a what's what's been impressive about that? Of course, my my connection is not
Moira:Right. I'd give you some good ones and some whoops. Oops. Couldn't figure that one out, but happy to talk about that.
Daniel Kraft:I mean, sort of like it's an it just has a sort of the experience, it's a what if Yeah.
Moira:Fears of Waymo in your town. I can do that, and say, okay, let's convert all
Daniel Kraft:that we can talk about how it makes it so much safer. Oh gosh. I cannot there's something really wrong. I'm going to close Safari. It's just one second.
Daniel Kraft:Going to quit something here, but not you. I think that's the problem. I've got too many windows open. Give me a second. Come on.
Daniel Kraft:Force quit. Come on. Seem to
Moira:mean it. I'm
Daniel Kraft:sorry. This is really frustrating. You can still hear me. Right? Because I'm off screen.
Moira:Oh, yeah. Yeah. Yeah.
Daniel Kraft:You can't still hear me?
Moira:I can. I can. Can you hear me?
Daniel Kraft:I can't hear you.
Moira:Get back to your screen.
Daniel Kraft:Hold on a second.
Moira:Back to your screen.
Daniel Kraft:Something weird happened. Oh. Yeah. First quit. I can't hear my I don't see myself showing.
Daniel Kraft:I show I but I can't can you say something?
Moira:+1 234-1234? What happened? Can are you back you're back? Alright.
Daniel Kraft:You can hear me. I just can't hear you, which makes this
Moira:I'll stop.
Daniel Kraft:Since you're typing.
Moira:And reload. Okay.
Daniel Kraft:I don't I didn't change any settings.
Moira:Yeah. Stop. Daniel, good to see you again.
Daniel Kraft:Great to see you.
Moira:Now I gotta tell you, people have been telling me, you wrote a post on what if health care was like a Waymo. They're like, oh, you gotta hear this. You gotta read this. And I'm like, I think I'll just ask him. What if health care was like a Waymo?
Daniel Kraft:Well, it's pretty incredible. We both live in the San Francisco area now. You can open your Waymo app, press a button, and a robocar will come up and pick you up. And it got me thinking, you know, what if healthcare was more like a self driving car. Right?
Daniel Kraft:There's a few things that we probably all want in healthcare. Number one, that it comes on demand. Right? When you order a Waymo, it comes with your name flashing on it or at least your initials. So it comes on demand.
Daniel Kraft:It has transparent pricing. Shocker. Right? You know how much you're paying before you sign up for the ride. Right?
Daniel Kraft:You can personalize that ride. You can say, I want my Beatles playlist or I want this temperature, And then it takes you to your destination, and you don't have to talk to anybody. So there's a few qualities of a Waymo or or sort of robo driving that I think could be lessons for for health care.
Moira:I think what's also fun is that you can send people, hey. I'm coming to you on the Waymo and track me track the map through the city as I come. You know? And so you you're you say, hey. I'm over here doing my thing, or I've got my I'm starting to do this tracking of something on my body or my my for your friends who may all have the same condition.
Moira:Like, I'm doing this. You can track me. I'll track you. I mean, that's that's kind of, the sharing business is is a big deal.
Daniel Kraft:Right. The idea of sharing your health care journey. Right? It might be a weight loss journey or managing your diabetes or a a friend or family member with cancer, god forbid. So we're all hopefully can cross connect because health is social.
Daniel Kraft:We all feel sometimes much more supported by our community and friends, and that can be done in a more virtual way. But, you know, sort of back to the self driving car analogy, you know, it took a while for them to evolve and be regulated to the point where you didn't need to have a test driver at the wheel. So there's they're super AI enabled. Right? Think of all the sensors that come together, the LiDAR, the crowdsource mapping, the the GPS.
Daniel Kraft:So they're AI enabled, but there's still a human in loop. When you need one, you can press a button or make a call, and someone will sort of teleport the driving of your Waymo. So that's one. So they're not necessarily keeping completely autonomous and you can pull and get help when you need it. Number two, they're sort of on demand, meaning you don't have to even own a car anymore.
Daniel Kraft:It can come to you. You don't have to go to it, and we often have to wait a long time to to to sit in the doctor's office to see a clinician. It's also contextual. It knows your Waymo app history and maybe where you like to be picked up and what your next destinations might be. That transparent pricing, think, is super key because that can give you control of do you really wanna take that trip.
Daniel Kraft:And what's more exciting about when we pull these dots together, these Waymo robotaxi type examples are continuously learning. They're learning from your patterns and the traffic of thousands and millions of folks around you, so they're adaptive. And hopefully, when we go as we've talked about from now near next, our health care isn't reactive. It becomes adaptive and can reroute you as you need based on a traffic jam or your labs or your behavior or your environment. And, you know, it's always on as opposed to sort of our health care experiences.
Daniel Kraft:We're often very fragmented. So I think the world of autonomous vehicles, which are solving for mobility with real time adaptive navigation, has some lessons for health care where we wanna sort of guide our health care journeys to be more personalized, contextually aware, and always being optimized for sort of the next best step or next best turn in our health journeys.
Moira:Well, Waymo doesn't care if it's the middle of the night or the weekend or really early in the morning or that has been on shift for, you know, how many hours. It's just ready to go like any good robot. You know? Doesn't get tired. Just keeps going.
Moira:It just keeps going.
Daniel Kraft:And we know that fifty percent of doctors are below average.
Moira:They always will be. You know, your math skills here, they'll always be below average.
Daniel Kraft:And especially when they're under slept or stressed out or been up all night.
Moira:Those are the ones definitely below average.
Daniel Kraft:But that happens. And we as you mentioned, you know, robo drivers don't get tired, and we're seeing in the data that, at least with the way most self driving cars, they've been involved in ninety one percent fewer serious injury or worse crashes, and 80% fewer crashes causing any injury than human drivers, so they're safer. And that has huge public health implications because we spend a lot of money and trauma and suffering from our sort of car culture where crashes and deaths occur on the roads at relatively high numbers. With a self driving car element, that's a public health element. It's not model for healthcare, but when everyone is in a self driving car or more of us are, we'll have less accidents and less needs for emergency room visits and, fracture repair and and beyond.
Moira:Now, of course, everything is quirky. You never you you have to keep finding situations to see how to solve it if you can solve it. I was going through Golden Gate Park the other day, and, there was a van which had broken down on a two lane road, which was sort of filled with cars in both directions because it was like five ish kind of thing, and they were trying to cut through the park. Well, this van had broken down in a one of those huge tow trucks, not just the little ones, but the big ones that, like, you have to kind of crank a chain to get it up on that had come. And so what had happened is the fellow who was driving the van had gotten out and had sometimes had cars come in one direction, and then he had other the cars coming the other direction drive around it.
Moira:It was just a two lane road, except right behind this was a Waymo, and it had no idea what to do. And everybody I have to say everybody had a big smile on their face going, uh-uh. I'm driving around. That's it. So that there will be times when it doesn't quite work, and we have to work it out.
Daniel Kraft:100%. It's it's just an example of we have to keep learning. None of these systems are ever gonna be perfect. We're moving to sort of the self driving operating room for the future where we'll have more autonomous robotic surgeons, but all these things take time to evolve, but on an exponential, I mean, we remember twenty years ago, the first self driving cars, it was a DARPA competition. They had a competition in, I think, 2005 or so.
Daniel Kraft:In the first year, almost all the cars fell off the road, and I think one only finished. The next year, more. And then, you know, we started to see some Google versions of the self driving cars, the early prototypes on the roads with safety drivers. And now we're almost taking it for granted that there are self driving cars coming and picking us up. So, you know, you need some time for these to evolve, particularly the regulation, the safety, and those edge cases like the tow truck one example.
Daniel Kraft:We'll learn from those to build better self driving experiences. And in health care, we can start to learn to have bet better sort of self health guided experiences as well.
Moira:Now I don't know this if this will always be true or if it will be true for health care, if it's more like a Waymo, is that the first time you're in one and the first couple of times, you're like, this is really exciting. There's nobody here, and it's driving me home. I was like, this is really the first year of the ledge, and they're like, this is really fun. Do you know that after a while you don't get tired of it? There's still this little buzz of there's nobody there driving you.
Moira:Now that might be generational. You know, the next generation will be like a driver. What do you need that for? You know? But it's it's really very interesting that the idea that it's doing what I want it to do.
Moira:I have confidence it's gonna do what I want it to do. And I was very precise about it. You know? It's like, I don't have to talk to the driver. Don't I have to look at the driver.
Moira:They don't have to look at me. You know? And it's interesting. That ability to control your environment is of whatever you can do that as a human, it's a really good thing.
Daniel Kraft:Right. It's hyper contextualized, it's personalized, and in some cases, you know, we won't have the self driving car coming to you. It might be bringing inside it having healthcare sensors or be a little mobile clinic. You know, you can imagine for emergency care, there's folks developing drone ambulances that will pick you up on a highway or if you're lost in the mountains and and fly you out to safety. You can think about all these sort of things coming together and sort of this automation, robotics, self driving components that will sort of bring us care anywhere and will save lives, and in some cases, just getting us to our destination, whether it's a health destination or a physical one, in a in a more smooth, fun, and economical way.
Moira:And you'll get way more health care. Sorry.
Daniel Kraft:That's a good pun.
Moira:Punishing, but we'll take it. We'll take it. We'll take it. Hey. We'll see you soon.
Moira:Thanks for coming in.
Daniel Kraft:Hopefully, didn't drive you crazy, but I'm bummed. See you soon. Bye.
Moira:That was good. Okay.
Daniel Kraft:Yeah. The little zingers at the end. Yeah. Let's do
Moira:one more. The little zingers. Little zingers. I like I like the zingers. You wanna do the the the imaging CT?
Daniel Kraft:Yeah. It was sort of a based on this stuff, I'm gonna pull the notes up on this one because there's some more detail to share.
Moira:I have GRAIL handout.
Daniel Kraft:Early cancer early cancer I do have early cancer detection.
Moira:Early cancer detection. Okay. Okay. So you use GRAIL a lot. A lot of people don't know what it is.
Moira:Don't forget to say what it is.
Daniel Kraft:And I'm assuming grail. And there's gardens, and there's what is it? Clue. I don't see the one for I should look up the one that does.
Moira:Yeah. I
Daniel Kraft:think it's called ColorGenomics. Just wanna mention it. Call Cologuard.
Moira:Cologuard. Yeah. Cologuard.
Daniel Kraft:Yeah. Those are all molecular. That's the yeah. We can just just paint the picture of molecular genetic testing at the beginning and then blending that with imaging and other other modalities. Okay.
Daniel Kraft:Ready to rock. And we can do some now near next too if you wanna frame it that way at the top.
Moira:Yeah. You just just bring that in wherever you it is. It's like oh, okay. I could just let me reframe it in that. Daniel, welcome back.
Daniel Kraft:Great to be back, Moira.
Moira:Cancer is always a bad thing. That's just always bad news, and we want the diagnosis of cancer to come as early as possible. Where are we with, cancer diagnostics? And, of course, I'm gonna ask you in the context of now, near, and next.
Daniel Kraft:Well Mora, I think, you know, the now of oncology or cancer experience for too many folks is that they get diagnosed at stage three or stage four, meaning that the cancer is not small or it's spread beyond its point of origin, like a colon cancer is metastatic, it might end up in the lungs or the liver or the brain. Of course cancer is a very heterogeneous description, many subsets, many types. So that's the now. The near, what's starting to become common is the ability to screen for cancers. We all are familiar with mammograms and colonoscopies, but to do that in a much more, let's say, specific way, at molecular screening or the idea of a blood biopsy is becoming available to do a basically a blood biopsy to look for the molecular and genetic signs of a cancer, before it becomes evident clinically.
Daniel Kraft:So there are a few companies doing that. One is called GRAIL, which is a spin off from Illumina, where they look at sort of the epigenetics of circulating DNA and can see if there's signs of a tumor or a cancer, you know, sort of leaking some of that genetic material and its subtypes. There's companies like Guardant doing that at scale now, Clue. Cologuard is one that can, look at molecular elements in your stool to to replace a colonoscopy in some settings. So we're really entering an era of sort of molecular cancer screening that, again, particularly for folks at high risk, are really great ways to pick up cancers early.
Daniel Kraft:And when you have an early stage cancer, it's much easier to treat and hopefully cure than than late stage.
Moira:Well, what's next, though? What's next
Daniel Kraft:is sort of this idea of sort of multimodal health because these tests are perfect. They don't have a 100% sensitivity or specificity, meaning that some cases you'll have false positives and false negatives, which can cause a lot of stress and expensive workups. The next step is sort of blending modalities, and sometimes imaging is one way to do that. Some of the listeners out there might be familiar with companies, like Ezra or Pernuva where you can pay out of pocket today, but spend a couple thousand dollars and get a full body MRI done in about forty five minutes or sometimes faster that can screen for things that don't show up early like some some forms of brain cancers or pancreatic cancer or ovarian or prostate that can show up on imaging. So that's one modality.
Daniel Kraft:When you combine an imaging modality with the genomic type screenings or the blood screens, that gives you another higher level of ability to go, maybe every year I get a blood screening and a full body imaging screen to make sure I'm not gonna have something surprise me later. And many folks get scans for other indications. It might be a chest x-ray for a cough or abdominal pain, and what's getting exciting is it used to require sort of special imaging to pick up a cancer, often done with something called contrast, like a contrast CT, but a group out of China has released a platform called PANDA, short for pancreatic cancer detection with artificial intelligence, where they found when they analyzed like over almost 200,000 abdominal or chest CTs that they could detect dozens of cases of pancreatic cancer at very early stages using AI to find those very subtle changes around the pancreas that would indicate this person has early pancreatic cancer. We should intervene, we could do that surgery, we could save their lives just from a chest or abdominal study that was done for totally other purposes.
Moira:I think what's so interesting in the now, near, next of it all is if you're looking for cancer, you know, this is what's now, this is what's near, this is what's next. And if you just move that slightly over, it's like, what would a general health program be to make sure you're in good health? Not even looking for cancer, looking for anything that every year you would go in. I just say every year because that's what we used to do. And you would go in and take a series of tests and it would look for all kinds of things.
Moira:And that's what you you would do every year. I don't know if they'd call that preventative or just a question of maintaining your health.
Daniel Kraft:Yeah. Preventative screening. Sometimes the guidelines change. Do women get a mammogram at age 40 or colonoscopy use at age 50 or 55 or earlier? Sometimes we can think about hyperpersonalizing.
Daniel Kraft:Many folks do have genetic risks for cancer, like the the many folks have heard of BRCA one and two, which gives many women and even men higher risk for breast cancer or women ovarian cancer. So we might wanna screen those folks differently. And sometimes the signs of an early malignancy are gonna be quite different. They're now AI enabled little cameras you can put in your toilet that look at your stool. Not a sexy idea.
Moira:No. No. Sorry. No. Guess just try another technology.
Daniel Kraft:Right. Well, that technology can see changes in, let's say, bowel habits that might indicate an early GI issue or even a colon cancer. So that's my main point is the idea is that we can start to hyperpersonalize screening based on your risk factors and your exposures. If you lived in, let's say, near a pollution plant and you have high risk of exposure to certain toxinogens or you grew up in Beijing with higher levels of lung exposures to pollution and you might need screening for lung cancer even if you're not a smoker. Those things will kind of shift our personalized screening and then those screenings can be not just the fancy MRI and lab tests, but come from subtle changes, again from bowel habits, maybe how you're typing, changes in your physiology that are suddenly changed from your baseline over time.
Daniel Kraft:So we're ending this era of putting little small dots together and putting them into context that will not just be for cancer, but for many other diseases that hopefully can really change the game and not surprise us with these stage three and stage four diagnoses.
Moira:Well, thank you so much for coming in, Daniel. I I look forward to seeing you next time.
Daniel Kraft:Thanks, Maria. Stay healthy.
Moira:We got it. Got a five. Whoo. Whoo. Whoo.
Moira:Whoo. Now don't hang up. I