Physician burnout, AI scribes, and the future of medicine, feat. Dr. Ali Okhowat, MD

[00:00:01] Speaker A: I'm Rian Gauvreau, and this is Healing Healthcare, a podcast where we explore the challenges, innovations, and ideas shaping the future of healthcare. Today I'm joined by Dr. Ali Okhowat, a practicing primary care physician and the co-founder of Autochart.ai. Ali sits at a fascinating intersection. He lives the daily realities of frontline medicine, the time pressures, the documentation burden, the cognitive load. And he's also actively building AI tools designed to reduce that burden. He's not theorizing about the future of medicine from the sidelines. He's building it from inside the exam room. Today we'll talk about physician burnout, AI and clinical workflows, what technology gets right and wrong about medicine, and what it would actually take to heal the infrastructure of care. Ali, welcome.

[00:00:50] Speaker B: Thank you.

[00:00:52] Speaker A: I think the first question that I've got is, you know, what personal frustration or professional frustration led you to create Autochart?

[00:01:03] Speaker B: I think it's the feeling of I should not be doing this for the thousandth time. There's a better way to do this, right? And especially the late at night stuff. Like I was awake until, God, 1:00 AM last night, I think, going through the inbox, right?

[00:01:20] Speaker A: Yeah.

[00:01:21] Speaker B: And that feeling of frustration, which leads you to basically be like, this is not for me. Like I chose this as a calling and I felt like I really loved it, but at some point it becomes that death by a thousand cuts. And the frustrations I think for most people who are in the health profession, if you ask them, one of the top 3 is always going to be something to do with charting or documentation, the clicking and the typing that just kind of wears you away.

[00:01:54] Speaker A: Yeah, we hear a lot about that in terms of like, I mean, we call it internally this concept of pajama time, but it feels like almost every doctor that we interact with, you know, they have a regular day that they punch out, you know, 8 hours a day or whatever the number happens to be. And then they go home and they, you know, maybe make dinner for the family and then they're back on the couch charting again. I think that like for us, that seems like that's not a part of the agreement, right? I don't think everybody that heads into medical school assumes that that necessarily means you're gonna have to punch out a 16-hour day for the rest of your career. But I think that that's what a lot of doctors are discovering. And I mean, curious around your opinion, like what do you think the impact of that is?

[00:02:41] Speaker B: Well, I think there's a, a few impacts. One is it drives people or incentivizes them to go towards those areas of care where maybe you don't need so much pajama time, right? Mm-hmm. You carry out your duties and you clock out and your work is done, right? And that I think we had seen, and it's related anywhere I've gone in the world, it's very much related to the, the compensation scheme, right? So if you pay people for that extra time, then you'll see people coming back into the fold. I mean, a good example is in British Columbia, we have this new— relatively new still in the grand scheme of things— longitudinal family practice, you know, the LFP model, where if you're doing what they call indirect care, which is things like checking labs, following up on patients, but not necessarily interacting with them directly, then you still do get compensated for that, which I think is a huge win for people who are seeing patients longitudinally in the clinic, kind of the traditional idea of a family doctor, right? Before that, what you saw was, and it was, it's multifactorial and it remains kind of mired in a lot of different considerations, but you saw a lot of people just moving away from longitudinal family practice. Mm-hmm. Because it was too painful in a way, right?

[00:04:10] Speaker A: Mm-hmm. And like, what do you think the kind of like macro-level impact on, you know, the healthcare system at large is? Do you think like a lot of people are leaving primary care?

[00:04:19] Speaker B: Yeah, I mean, one is people leave the profession. The other one is people stay, but they get burnt out. And the end result is that patients don't have primary care providers that are following them, that know them, that know their family, And you end up with statistics like, you know, 1 million people don't have a family doctor, right? It's not a sustainable or scalable system.

[00:04:48] Speaker C: Mm-hmm. With the longitudinal care model, so you mentioned that now it's a bit more palatable because clinicians are getting paid for that extra work, but that extra work still exists.

[00:05:00] Speaker B: Yeah.

[00:05:01] Speaker C: So, Do you see it as being a part of a bigger solution? Do you see it as just being kind of like a temporary salve for a deeper problem?

[00:05:13] Speaker B: It's not a temporary salve, but it's one part of a multi-pronged solution. And there is— everything's a matter of trade-offs, right? So there are trade-offs in that model, but it's just one part because if you think about you know, a physician owner or a physician who has a stake in a clinic that's being run, you know, outpatient longitudinally, there's major costs to that. And that kind of cost structure determines how many people stay in the game, so to speak, right? So a big part of it is the compensation mechanisms and the incentives. Another part of it, though, equally important, is just how fulfilled people feel undertaking their craft, right, practicing medicine. And, you know, whether it's in hospital or outpatient, that documentation burden is pervasive, right? And now I think with technology, it's been well proven there's some serious areas where that can make a dent and reduce that burden. So, you know, that's why we were really excited to get into it. We didn't know— I think a lot of the things that would happen, but we were hopeful that, you know, the landscape would change and regulation would in a way catch up. And I think that's still a cat and a mouse kind of game in terms of the technology that's evolving, you know, leaps and bounds faster than the regulation and the kind of the soft law and the hard law is evolving.

[00:06:45] Speaker A: Mm-hmm. And so I imagine that that's got to create all sorts of challenges for you in building an AI company when the technology is moving radically faster than the industry itself. And can you maybe talk a little bit about like what you've encountered?

[00:06:59] Speaker B: Yeah, I mean, when we originally started Autochart.ai, it was, you know, a question of, I don't even know if the speech-to-text is good enough, right? If the fidelity or the resolution of that transcription. And then it was, I don't know if even if the resolution is great, if the synthesis of this, you know, story and conversation between a patient and their health professional, or, you know, multiple people speaking together, if it can be pieced together properly. And quickly we saw, not in the order of, you know, years, it was weeks, months, where suddenly a new model would come out, you know, the latency to output would suddenly collapse and, and You'd be like, wow, problem solved. Okay, now what? Right? And so, you know, we've seen that it's just kind of exploded in popularity. And now what we're trying to tackle is, you know, voice agents in healthcare and just more agentic care flows, right? Mm-hmm. And what we're seeing is it's similar to kind of those early days of, you know, the whole AI scribe movement. I think the technology has landed, and now it's expanding from that kind of core area of the health professional and the patient having a conversation through to the back end of, you know, billing and diagnostic codes and charting and the workflows of writing a prescription and, you know, sending off a referral. And even kind of disseminating into what was the patient doing before they came in, right? The pre-visit stuff. What are they doing after they leave, right? I think what a lot of people don't realize is so much of medicine is, I don't want to say guessing, but making educated inferences about what is it that's happening.

[00:08:53] Speaker C: Yeah.

[00:08:54] Speaker B: And I originally started, one of the reasons that I was late at night thinking, I was kind of writing a letter to patients and saying, I just, you know, I don't have enough time to do what I want to do. And as I was doing the very early drafts of writing that letter, I kind of thought, well, I could do this, but then what would I do instead? And I thought, well, I would try to tackle this problem, right?

[00:09:21] Speaker C: Mm-hmm.

[00:09:21] Speaker B: And so I stayed in trying to understand how well could I tackle the problem? Could I tackle it to the point where I would kind of step back from the brink And I'm happy to say that, you know, to a certain extent, yes, certainly from the Autochart.ai side, but there's so much to be done just in terms of, you know, now implementing it in other areas and also doing a lot of the safety testing, which is often not talked about, right? A lot of it is just about, look how cool this technology is, but so much of it still just needs to be rigorously tested.

[00:09:55] Speaker A: Yeah, yeah, well, and like, you know, I think it's, this is one of the few industries where the downside risk is extreme, right? And so you need to make sure that the safety is there. And, you know, to your earlier point, I think, you know, just around like doctors, yeah, you know, you are kind of confronted, like I, you know, I've often likened doctors to almost like, you know, a Jeopardy player with none of the letters turned around. It's just like, you know, can you give me a few more letters so I have something to work with here, right? And, you know, and so, you know, I think like, I've certainly seen your product and I think it's amazing, but, you know, I'd love to— I mean, you're seeing, you know, the inside, you're working with your customers, you're appreciating the value that Autochart is delivering and AI scribing as a, you know, kind of general tool set is delivering to the primary care modality. But like, I'm kind of curious just around like, what's your perspective on things? Like, how is it changing the game? And, And, you know, I think like related to that is what are you most excited about?

[00:10:55] Speaker B: How is it changing the game? I think people were and continue to be very skeptical about it as they should be. How well is this going to actually understand what I'm saying? How faithful to the actual conversation is this output going to be, whether I'm dictating or having a dialogue with someone? And then that's going to be true. Kind of this theme of trust, right? Across every iteration of the technology, whether it's reviewing your inbox and you're like, "Do I really want to give it access?" Yeah. To everything that's coming in and rely on it to maybe tell me what's critical or not, right? So, you know, there's definitely value. There are the early adopters, you know, and then the much more large number of people who, you know, are more conservative to it. And then of course, there's the late adopters. I think we're just now getting into that area where a large majority of health professionals are going to be, by the end of 2026, using Ascribe, far more than— Okay. Will not, right? Yeah. The next question becomes, well, what else is there, right? So I think, you know, voice was certainly in 2025, one of the breakthrough capabilities of AI. And I think in 2026 now we're seeing kind of the verticalization of that, right? So we have developed, you know, Autochart.ai voice and we're testing it right now to see What are all the different things that we have to prove in order to make sure that it's safe, it's transparent, it's reliable, right? If you put a voice agent at a front desk of a clinic, how do you simulate, you know, 999 times out of 1,000, for example, if suicidal ideation is expressed, this escalates appropriately?

[00:12:55] Speaker C: Right.

[00:12:55] Speaker B: If guidance is sought on a critical medication interaction, that, you know, the system responds appropriately, saying that I'm not trained for this, right? And then equally, you know, how many times for those administrative questions where somebody is just wondering about something, can it actually save time from an already overburdened front desk, which I would say 99.9% of clinics, you know, have an overburdened entry point. Yeah. Whether it's inbox, phone, or, you know, people coming in, there's just not enough people to process that in a way that makes people feel taken care of, right? But infinitely patient AI can, but it can also make a lot of mistakes. Yeah. So.

[00:13:47] Speaker C: Yeah. I wanted to ask a question going back to what you said about, you know, doctors are, They just have what they have to work with in terms of what the patient tells them. And to an extent, well, actually in the exact same way, so does AI. So how do you see AI supplementing patient care from a doctor perspective when it only has the information that you have? Does it change the way that you practice?

[00:14:17] Speaker B: Yeah, I'm trying to think of a corollary, you know, the, in the early days of the Apple Watch, let's say, or when there was direct-to-consumer genetic testing, people would come in and they would say, "Here's the data," and they would just kind of dump it in there and be like, "Tell me what to do now." And you'd just kind of be like, "Well, I don't know." But I think what you're seeing now on the kind of consumer front is patient health records are reemerging. There's this movement in the industry to kind of consolidate data on the patient side. And there's some really good intelligence, right? I'm not even talking about like direct-to-consumer, you know, lab testing for certain biomarkers or in the longevity space. I mean, just simply taking the wealth of digital information that's in, you know, your activity monitors and making sense of it. Like if you see what Whoop and Oura, for example, are doing just in the predictive intelligence space in terms of guiding people around some health behaviors, that stuff is huge. And I think that is going to increasingly interface with the clinical realm, right? A good example of how that has happened is these continuous blood glucose monitors, right?

[00:15:37] Speaker C: Mm-hmm.

[00:15:37] Speaker B: That's been a game changer. I mean, yes, you know, people should, where appropriate, do their finger prick testing, but for people who we really need to know the values of how their blood sugar is changing during the day, having a continuous blood glucose monitor is— it completely changes the game because now I see, for example, that every day at 3:00 AM you're becoming really hypoglycemic, right? An alarm is waking you up, or you thought you were doing okay, but, you know, patient I recently spoke with, she said, "Yeah, I thought I was okay, but I'm only 17% of the time in range, like where I should be." And she had shared with me the stream of data, and I called her and I said, "You know, you're running very high." And she said, "Yeah, but this is what I've always done. I thought I was just kind of feeling different because," And I said, "No." Now that was me proactively checking in the system because, you know, I'm a geek and I like to try to figure out if we can kind of connect that data stream into the EMR. But I think more and more it's going to be their digital assistants pinging us to be like, "Hey, you should take a look into this," you know? And it's exciting, but it's also like an information overload. Yeah. Overload on our end as well, right? Like, who has the time to continuously be tracking people's wearables and their continuous blood glucose monitors and all the other stuff that's coming out?

[00:17:10] Speaker A: Slight change in tack, but kind of curious about like, you know, I think you've— you're obviously somebody who sees things that are broken and takes action. And so, you know, I think you've— you know, you're part of the evolving landscape that's kind of changing medicine on the charting front. But What other domains in the medical space are you watching and excited about? Or like, what are you kind of inspired as, you know, kind of a short ways down the road that you're looking forward to that could potentially change healthcare for the positive? 'Cause I think, you know, in these bleak times we live in, you know, I think it's really easy to just dwell on what's broken. But, you know, I think that there's also a lot of, you know, especially us as technologists, I think we get to see some of the stuff that really lights us up and excites us about what's coming down the road, and curious on your perspective on that.

[00:18:02] Speaker B: Yeah, I would say that the 3 big things are, and probably as I speak about them, I'll be like, "Oh, the 4th and 5th thing." One is just technology enabling care to move closer to the home. The US has a lot of statistics on this, but I think Canada is also quite similar. You have an aging population, Hospitals are incredibly expensive. The technology and the infrastructure for care delivery is just coming closer to home. So that's number one. The second big thing— and I'm personally super excited about this— is just personalization. If you just take a theme across health care on the consumer front, provider front, things are becoming hyper-personalized, like from the molecular signature of a drug and asking, is this— Is it, you know, compatible with me? You know, what's the best dose to take? You know, those questions, I think, in 3 to 5 years' time are going to be answered so much more meaningfully than we are answering them right now, which is just based on a large dataset of a study that was done 30 years ago, and we just kind of keep prescribing the same way, right? And that's going to evolve also to a lot of the behavioral stuff with the expansion of wearables and the consolidation of the data on the consumer side. I think the third big area, which probably for organizations is going to be huge, is just a lot of the back office stuff, right? Revenue cycle management and kind of diagnostic labeling. Those are incredibly boring but incredibly important areas that are kind of documentation adjacent, right? They're heavy administrative burden. I think that's going to shrink, and it's going to be by virtue of, you know, these technologies which have, you know, optical character recognition already embedded in them, and they can take these documents and process them, understand them, kind of act on them in a multi-process way. So that's kind of exciting on the organizational front. The thing that, you know, I'm really excited about building, maybe, if I can distinguish it from just kind of general trends, is the— what— it's not my term for it, but I think it's not maybe widespread. So I call it asynchronous care, which is the delivery of care asynchronously. Essentially, you know, you have an assistant that sits between the health professional and the patient. And the interaction or the exchange of information is done in a way that is right for the right person at the right time. So as soon as, for example, I'm starting to feel like, you know, I'm getting a recurrence maybe of, you know, an infection that I've had before, I can start to prompt the system, and it maybe starts to ask me questions. It's clinically intelligent. Mm-hmm. It interfaces maybe with a care team member. And this is where it's fundamentally, I think, different from what we've seen in the past. And there aren't many companies in this space, but this is what we're kind of building towards, which is the care team, the health professionals, and the staff, they also get to interface with this assistant. Which other questions need to be asked? Which other, you know, medications are in the loop? Are there refills? This kind of stuff. Mm-hmm. And there's this exchange by voice, by text. And then at some point, a threshold is passed where a connection needs to be made, a decision needs to be taken, or some kind of action needs to be moved forward. And then the humans are kind of brought in the loop. And so there's this theme of humans on the loop, which I think we're going to be seeing a lot more of as you have this kind of asynchronous care taking place. And I think whether it's scribes or voice agents or EMRs that are AI native, A lot of what we're going to see is kind of semi-autonomous work that's going to mirror maybe what we're seeing in terms of autonomous driving levels, right? I think in healthcare, we're still at like level 1 autonomy, but we're going to see over the next few years kind of moving into level 2 and 3, what that means in healthcare.

[00:22:23] Speaker A: Yeah, yeah, great answer. And, you know, I think like I would probably echo a lot of those things. I think the Anna One Healthcare is super interesting. I mean, I got invited to go to a fundraiser for a rare trait and it was the first moment that I learned that we can actually produce unique medicines for an N of 1. And it's a very expensive process, but the fact that we can even do that now is just so inspiring. And I can't help but think that with AI, we're going to be able to invent these molecules that much faster. And so whatever unique condition people have someday way down the future, we might be able to solve with a unique molecule custom designed for that person, which I think is super interesting. And I think you mentioned about wearables, like literally everybody I know has a wearable. They're collecting all sorts of data. And I think it's really been oriented around the datafication of self, but I think it could be so powerful if you could start pushing that data to your health provider. And now they're able to start providing this asynchronous and proactive medical care, which I think could be— Yeah. That's what I think the future looks like is my watch tells my doctor I'm not sleeping. My doctor sends me some sleeping pills and me and my doctor never talk, right? Like it just shows up in my— It just shows up in my inbox or like my mailbox. And it's like, hey, you're not sleeping. Love ya, buddy. Like, you know, note from my doctor. And, you know, maybe that's a bit of a fantastical future, but I think that that's kind of the care that at least the digital forward people are thinking about, right? Is like, "My watch knows I'm sick. My doctor now knows I'm sick. So all I need is the medicine that's going to make me not sick," right? And I think that that's potentially what the future could look like. And I think why we're so excited to be involved in that equation.

[00:24:08] Speaker B: For anybody who's worried about sleeping pills being prescribed autonomously, we would try cognitive behavioral therapy first for insomnia, but yes.

[00:24:20] Speaker C: And also just throwing it out there, I do not have a wearable, so we'll come back to this. We'll come back to this study in 30 years.

[00:24:28] Speaker A: What kind of technologist are you?

[00:24:29] Speaker C: And we'll assess how—

[00:24:30] Speaker A: Yeah.

[00:24:31] Speaker C: I'm trying to disconnect as much as possible. But I think you said something really interesting that like strikes for me is my family doctor is in Abbotsford. We live in Langley. And when we moved, I— there's just no chance of switching doctors, so you— keep your doctor and now it's like what I'm really excited about is like how can we make it more efficient for him to continue to serve our family and provide that care without that gap being so pronounced. And, you know, there are things that I can book a telephone appointment or an in-person appointment if I need to go in, but what's missing is like sometimes I don't know which I need. Mm-hmm. And there's nothing in between me and him to tell me. And so I'm really excited about, you know, what you were saying, like how can we have something in the middle that's like, "Hey, you actually need to go to the hospital," or, "Hey, this is like, this is important and we can do it over the phone," versus like, "Yeah, we need to see you in person." And like almost like that triage, that automated triage.

[00:25:39] Speaker B: Yeah, yeah. And I think over the past maybe 10 or 20 years, there's been more of this movement from, you know, what they stereotypically might call sick care to well care, right? A big part of the sick care philosophy was the patient will reach out when they are unwell, right? And that marks kind of an episode in your care journey. And when we think about, hey, I don't just want to not feel bad, I want to work on being kind of the best version of myself. You have to have technology that can help those providers who have to be in the loop and make certain decisions to kind of bring a state change in, you know, whatever your care journey, whatever's involved in your care journey. Yeah. But you have to have technology to help that model to scale. And whether it's asynchronous or whether it's synchronous, there has to be some modality in which you can communicate how you're feeling without taking up the time of the limited resource who can help you to do something about it or work with you to do something about it. A lot of it, of course, you can do partly yourself as well.

[00:26:55] Speaker A: Absolutely. So I think like that's actually a topic that comes up for us you know, a fair bit, you know, in terms of like, I think that there is an evolution happening in the general population that, you know, people are increasingly less interested in break-fix medicine and more interested in kind of proactive longitudinal medicine. And, you know, I know, you know, I'm in my 40s now, I'm starting to think about the back half of my life and, you know, staying healthy and, you know, compressing that time between morbidity and mortality.

[00:27:26] Speaker B: Mm-hmm.

[00:27:28] Speaker A: And, and what are the types of things that I can engage in and how can I work with my medical professional? And it feels like the posture of the medical industry, at least in Canada, but perhaps more generally, is that it's really oriented around break-fix. Like if you're sick, you come in, we fix you, we send you on your way. And I think more of what I think consumers nowadays are wanting and asking for is like, I don't want to just talk to my doctor when I'm sick. I want to talk to my doctor about what my goals are and I want my doctor to help me get there. And so I'm kind of curious about like your perspective in terms of like what's missing now and what do you think needs to happen in order for that reality of a truly proactive kind of longitudinal medical system coming to fruition instead of just being oriented around this break-fix modality?

[00:28:21] Speaker B: It's so interesting when you look at like where the incentives are, right, and who you are within that system. If, if you're the steward of a public health system and you're trying to answer the question of how can I provide the best care to the most number of people, um, the majority of the time, you're going to design a system that looks very different from an individual saying I want to I want to live optimally, you know, I want to feel and kind of act in the best way I can every day. And those fundamentally will be in conflict, right? There's the wider kind of harness or guardrails of the society that that healthcare system is in. And again, everything is trade-offs, right? In Canada, We have in many ways implicitly made certain trade-offs in order to make sure that, you know, our health system can provide certain benefits which you won't see in some other areas in the world. And conversely, there's other areas in the world where there's other trade-offs made which maybe benefit more, you know, that, that latter distinction of an individual saying, "I want to feel as great as I can." I'm kind of speaking in the abstract, partly because I know getting into the weeds of this is super nuanced, and I don't want to say something that I'm going to be walking back because I wanted to say, "I mean that." But your question was, you know, what is the— sorry, what was your question again?

[00:30:08] Speaker A: Well, I think it's mainly just around, you know, I think that how the system has been designed is like when people get sick, we make them healthy.

[00:30:17] Speaker B: Mm-hmm.

[00:30:17] Speaker A: But I think what it feels like the system or the society at large is asking for is like we're not just interested in getting healthy when we're sick. We're interested in potentially staving off sickness so that we can be healthier longer, right? Mm-hmm. And I think that's a different orientation, and I get why, you know, the systems have been, built the way they have, but it feels like, you know, we're on an evolutionary path where perhaps we're going to have a future state where proactive medicine is much more central to how care is delivered because of, you know, it's basically if you're doing break-fix medicine, you have to pay that debt later, right? Mm-hmm. You know, like, and so if we're more proactive about our delivery of medicine, can we stave off people getting really sick later in life and where the profound cost to the medical system arrive, right? And so like net-net, can we make a healthier and more sustainable system by orienting the system to be a little bit more proactive? And I think as a physician, you probably have a unique perspective on that.

[00:31:26] Speaker B: Yeah, you know, there's this concept of healthspan which has become more kind of into the mainstream recently. I think a lot of people are focusing more on that and then thinking, what can I do to kind of prevent the things that might be happening down the line? On the other side, you know, the providers I know, as being somebody who has patients coming in and saying, I'm doing this, I'm doing this, I'm doing this, help me to, you know, move these goals forward. And I don't want to speak on behalf of, for example, you know, the family doctors in the clinic where I practice, but I think most of them and— Yeah. Most in general will just say, "I am so overwhelmed with the avalanche of demands. Nowhere do I have the time and sometimes the willingness by virtue of fatigue to kind of try to put into place systems that can help me to address those goals." Because I'm expected to see, you know, one patient or a panel of patients within X minutes, or, you know, this many patients within one year. And the style of medicine that I practice is largely dictated by those constraints, right? Even if I might want to practice a different way. If I do want to practice a different way, it has to become somehow sustainable, right? Mm-hmm. And this is where I think we're gonna see a lot of software and hardware coming together to help those practitioners who, you know, have it deep down saying, "I do want to practice differently." They are gonna be able to do that. Part of it is gonna be the information tools that they're using. Part of it is gonna be the incentive schemes and the compensation. But a big part of it is also them kind of opening up to the fact, and I think you're seeing that in the traditional medical world, I would say, which is always, I think, a little bit more resistant to these kind of evolutions that happen. We're seeing more people kind of stepping up and saying, "I want to participate in this too." Because if they don't, and I think we've seen this, other health professionals will gladly step in.

[00:33:41] Speaker C: Yeah.

[00:33:42] Speaker B: Yeah.

[00:33:42] Speaker A: Yeah, it seems like at least something that I think I've perceived in being in this market for a couple of years is that Everybody knows that things aren't working, but nobody has the time to invest in making them better. And so it's just kind of like in this state of never evolving because nobody ever has the time to invest in that evolution. And I get it. And it's probably a super frustrating place to be. But I've also had some curiosity around like, do some of these new models around compensation like LFP, do you think that that might have a potentially beneficial impact in that it'll allow people more time that is their own to invest in the right areas of their practice that could eke out those efficiencies, that could improve overall care delivery across the board?

[00:34:32] Speaker B: I think it'll definitely improve overall care delivery. If that model of compensation is going to improve kind of proactive or preventative healthcare, I don't think so. Not in and of itself.

[00:34:44] Speaker A: Mm-hmm.

[00:34:45] Speaker B: But if there are compensation schemes where people can invest time, if teams, right, not— it's not just about, you know, the doctor or the nurse, but if the whole healthcare team is incentivized and compensated for investing that time in prevention and proactive care, then yes. I think there's good examples around the world— Singapore, New Zealand, Switzerland— other models where they maybe have a more permissive environment in terms of technology, healthcare interoperability, right, and compensation schemes that allow that to happen. But I think it's by design a little bit more challenging here.

[00:35:26] Speaker A: Ali, I'm kind of curious, you're also a doctor as well as somebody who invests in research, and so kind of curious around like what are your research interests right now? You know, where are you spending your time thinking about the future?

[00:35:38] Speaker B: Yeah, I think it goes back to this theme of trust. And I was at this conference last year where they were talking about AI having to be licensed. And we think about the regulatory frameworks a lot as equivalent to maybe medical hardware where you have medical device regulations. And over the past decade, we've had software as a medical device. Which is okay if the software is static, but in something which is dynamic, is learning, and it's evolving definitively, you need a different framework. And that's essentially a licensing framework, right? And so one of the big research interests, um, as we're doing this safety, uh, work, um, is, is partnering with other researchers and, and kind of designing the studies ourselves to try to answer the question of what does safety look like, transparency look like, the observability of why did a certain interaction, whether it was AI plus human or just pure autonomous AI, why did that turn out the way it did? And then what are the different regulatory frameworks that would maybe enable us to implement this in a safe way?

[00:37:03] Speaker C: Yeah.

[00:37:03] Speaker B: What does a living laboratory look like where you would be able to throw in a care model, right? Whether it's, you know, a visual AI, whether it's a triage robot in emergency, whether it's, you know, a semi-autonomous prescription platform, right? And it would have to meet certain minimum thresholds. What is that? Yeah. It's very similar again to self-driving cars, but it's much more opaque because we don't see necessarily when a wrong turn was taken.

[00:37:38] Speaker A: Right.

[00:37:39] Speaker B: When there was a near miss. And that for me is fascinating because it's something that, as you mentioned, Rian, it has huge downside risk. It's something where if you're 99% accurate, it's still terrible. Because that 1% matters a lot. And so on the research front, that's just something that's fascinating.

[00:38:02] Speaker A: Yeah, I think that's an easy thing to like overlook, 'cause I think in, you know, when you're thinking about the certification of medical products, you know, you kind of think about them like cars, right? Like we're certifying this model of car, everything is always going to be the same. And that doesn't really consider like what does an evolving system look like, right? And so like, What if we were shipping a car and the car could look radically different in 6 months, right? Like, how would that change how we certify things? And I think it's like a really interesting problem space. So it's probably scary, but also really interesting at the same time. Yeah.

[00:38:36] Speaker B: Yeah.

[00:38:37] Speaker A: So I know you're a man who wears many hats and I know you're working on 50 things simultaneously, but for sake of our for listeners, why don't you just tell us a little bit more about like what things you're presently working on and that you think that potentially our audience might be interested to check out?

[00:38:57] Speaker B: Yeah, for sure. I'll put in a plug for our podcast, Feed Forward Health, which is kind of a media initiative that we're starting. It's articles with our chief medical officer, Dr. Tim Foggan, articles, podcasts, videos, courses at the intersection of healthcare, technology, and we like to say say humanity. And we're really excited as we're building out, you know, the team around this initiative, because there's just so many things to discuss. Every day when I wake up, you know, I'm seeing there's a new model, there's a new product, there's a new service, and things are moving really fast. And I think it would be interesting to have, you know, clinicians also kind of commenting on what they're using, what they're seeing their patients use, what they're seeing, the good, the bad, the ugly, that kind of stuff. And that's what we're hoping to cover in the Feedforward Health Initiative. And then the other big exciting thing that we're developing is, you know, we have Autochart.ai Health Assistant, which is the scribe, and we have now Autochart.ai Voice, and we're putting this together in the Aya Health platform, which is the asynchronous care platform that I was talking about. And we're going to be piloting that hopefully in a few weeks. And so it'll be interesting to see, you know, what— how people find that care experience is when there's an AI in the middle.

[00:40:16] Speaker A: That's awesome. Well, thank you very much for being here. We will be sure to become subscribers of Feedforward and be following you. It sounds like we're going to be talking about a lot of the same stuff. And it's a super interesting space to be talking about because it just— it feels like there's an unlimited amount of interesting topics to pick from and chat about. And so, you know, it may be that we're covering a lot of the same stuff, but we would very much like to have you back at some point. And so thank you very much for being here. Thank you for being our inaugural guest. And yeah, we look forward to future sessions together.

[00:40:50] Speaker B: Thank you. Awesome. Thank you. Likewise, on my side, I look forward to welcoming you on the podcast.

[00:40:58] Speaker A: Listening to Healing Healthcare. I'm Rian Gauvreau. Be sure to subscribe wherever you listen to your podcasts.