Doctors aren't "resistant to AI" because they're technophobes. They're navigating a fundamental shift in what it means to practice medicine when patients arrive having already diagnosed themselves via ChatGPT. The path forward isn't "adopt more AI tools." It's building a new operating model for clinical practice that balances ambient documentation, patient empowerment, and the irreplaceable human elements of care. In this episode, Dr. Spencer Dorn breaks down how AI is reshaping medicine: from eliminating "pajama time" for some clinicians to democratizing medical knowledge in ways that fray the profession's grand bargain with society. He explores why trust remains the currency that matters most, and what it means to be a physician when AI tools change faster than the relationships they're meant to serve. For benefits leaders and consultants, the implications are just as significant. As members increasingly turn to AI before ever engaging the healthcare system, employers have a real opportunity — and responsibility — to help them navigate that information and make better care decisions. Understanding how AI is reshaping clinical practice isn't just a healthcare story. It's a cost, strategy, and member experience story too.
Doctors aren't "resistant to AI" because they're technophobes. They're navigating a fundamental shift in what it means to practice medicine when patients arrive having already diagnosed themselves via ChatGPT. The path forward isn't "adopt more AI tools." It's building a new operating model for clinical practice that balances ambient documentation, patient empowerment, and the irreplaceable human elements of care.
In this episode, Dr. Spencer Dorn breaks down how AI is reshaping medicine: from eliminating "pajama time" for some clinicians to democratizing medical knowledge in ways that fray the profession's grand bargain with society. He explores why trust remains the currency that matters most, and what it means to be a physician when AI tools change faster than the relationships they're meant to serve.
For benefits leaders and consultants, the implications are just as significant. As members increasingly turn to AI before ever engaging the healthcare system, employers have a real opportunity — and responsibility — to help them navigate that information and make better care decisions. Understanding how AI is reshaping clinical practice isn't just a healthcare story. It's a cost, strategy, and member experience story too.
-------
“ The question that really motivates me most right now is what does it mean to be a physician in this new world? And even more so, what does it mean to be a human being as things are changing so quickly? How do we keep our eyes open and make the best decisions that we can so that we don't go down the wrong paths, which inevitably we will, but hopefully we'll have the awareness to recognize we're going down the wrong path so we don't keep going and we'll have the agency and autonomy to shift gears.” – Dr. Spencer Dorn
-------
Episode Timestamps:
*(01:01): Dr. Spencer Dorn's journey from gastroenterology to clinical informatics
*(04:31): The hidden work of medicine before, during, and after patient visits
*(06:54): How AI is transforming clinical workflows
*(12:12): The democratization of medical knowledge
*(15:47): AI-first care and the healthcare math problem
*(25:00): What it means to be a physician in an AI-powered world
-------
Links:
Connect with Dr. Spencer Dorn on LinkedIn
Connect with Kirk McConnell on LinkedIn
Learn more about Collective Health
Learn more about Caspian Studios
[00:00:00] Kirk McConnell: Healthcare costs are rising. Benefits are confusing and the system doesn't always work for the people who need it most, but it doesn't have to be that way. Smart employers and their brokers are flipping the script, cutting costs, making things simpler, and creating a health benefits experience their people love.
[00:00:20] Kirk McConnell: Welcome to The Benefits Playbook. I'm Kirk McConnell, and in each episode we uncover the bold strategies that are rewriting the rules of self-funded health benefits. Today, I am really excited to be joined by Dr. Spencer Dorn, who's the Vice Chair and Professor of Medicine at the University of North Carolina at Chapel Hill.
[00:00:40] Kirk McConnell: Dr. Spencer, thank you so much for joining us today.
[00:00:43] Spencer Dorn: All right. Thanks for having me.
[00:00:45] Kirk McConnell: So Spencer, we're gonna spend a lot of this conversation talking about ai, but we're gonna talk about it from a couple of different angles, how you see it impacting how doctors work on a day-to-day basis, but then how also it's changing how members interact with doctors.
[00:01:01] Kirk McConnell: Before we get into that topic though, I'd love for you to walk through, kind of walk us through your career journey and where you got to where you are today.
[00:01:09] Spencer Dorn: Sure. So I am a native New Yorker. I. Trained in New York and then Boston. I came to Chapel Hill where I am now, way back in 2005 to complete my training in gastroenterology and just started out with a clinical interest in, in kind of mind-body medicine, which I still, I.
[00:01:28] Spencer Dorn: Oriented around that. Clinically, I started out primarily as a researcher and then just got interested in why healthcare is so hard to deliver and got more involved in helping to run clinics and procedure units and different types of programs, and it became very clear to me that technology was about to reshape how we practice medicine, how we run practices, and how people find us, and all those things.
[00:01:51] Spencer Dorn: So kind of got interested in clinical informatics. And yeah, I've been doing this kind kind of work now for a while in different capacities, but I continue seeing patients and I work for our health system doing things related to digital technologies. I help with our school of medicine and clinical programs, and I do a bit of writing and just like to learn, like to share ideas.
[00:02:17] Kirk McConnell: You mentioned running a practice in practice management. For most folks, they go to the doctor. Maybe they watch the pit on tv, but they don't know what it's like to run a practice. What do you think most people maybe don't understand about taking notes, staying up to speed on the kind of the latest research?
[00:02:35] Kirk McConnell: What does it take to actually kind of run a practice? If you're a doctor these days?
[00:02:40] Spencer Dorn: There's the, so there's the running the practice, the orchestration of care, which most doctors maybe they participate in somewhat, but usually there are people who do more of that than others. So, you know, it's kind of, it's kind of like, to make an analogy, I dunno if it's a good one, like a restaurant, right?
[00:02:53] Spencer Dorn: There's gonna be a manager or the restaurant who's kind of looking at all the aspects of what's happening at the front of the house, what's happening at the back of the house, how are, you know, customers being served, uh, and then there are chefs and then there are, you know, waiters and hosts, et cetera.
[00:03:07] Spencer Dorn: So. For running a practice, it's more about just figuring out what's the configuration of our team? Like, who are the people that we need involved, what does care look like for individual patients, members, people, whatever, however you want to refer to the people that we care for. What does it look like from start to finish?
[00:03:27] Spencer Dorn: So that's kind of what, what running a practice looks more like. And you know, a lot of it hasn't changed. In many ways, probably over decades, you know, being a, a good clinician means, you know, caring for people, treating them with empathy and respect, and being a good listener and understanding, you know, the science and the state of the art and keeping up with that and just, you know, doing your best each day for each person you serve.
[00:03:52] Spencer Dorn: So a lot of, a lot of that is constant. I think what's changed more recently is that people tend to have more complicated. Needs than before, in part because we're doing a good job helping people manage chronic diseases and helping people live longer. So we're seeing people with, in general, more either more complicated aspects of illness or multiple illnesses combined.
[00:04:14] Spencer Dorn: So I think that's one challenge. And then that brings a related challenge with information. And how do you manage all the information about the person you're serving, about the information and the literature that may be related to the people that you're serving.
[00:04:28] Kirk McConnell: Well, so let's focus on, I, I loved your restaurant analogy.
[00:04:31] Kirk McConnell: Let's focus on the back of the house. So the doctors, the PAs. You mentioned that those interactions in a lot of ways are becoming more complicated 'cause you're spending time with patients that have more complex conditions. What do you think? People don't necessarily understand of when the patient leaves.
[00:04:47] Kirk McConnell: Walk me through what charting looks like, how staying up to speed on all that information looks like.
[00:04:52] Spencer Dorn: Yeah, so I'd say probably the clinicians are more the middle of the house. The front of the house is kind of, you know, your front door, how people discover you. Maybe you could think of things like check-in and making an appointment and, you know, the back of the house would be things like a lot of the administrative activities or maybe scheduling, processing, referrals, um, things like that.
[00:05:09] Spencer Dorn: I think what you're getting at is the actual, you know, care that happens, that kind of all this is, is created to support. That itself is a little more than maybe meets the eye. Uh, I'm a specialist, so I take a different lens at this, but I think it applies to primary care and different specialties. But in general, there's a, like, how do I prepare to see a person who's coming in, figure out what their needs are, what's happened before, what maybe.
[00:05:34] Spencer Dorn: We're thinking like in our mind, maybe we want to do like what are the, what are the types of problems we're likely going to address? Maybe pending some orders, getting some things in place, starting to write the note before someone comes in. And then there's the actual visit where someone comes into the office, you're seeing them online, and it's about, you know.
[00:05:52] Spencer Dorn: Getting, collecting more data, trying to go a little deeper on certain things, clarifying things, and then trying to come up with like, you know, what's the right plan for this condition based on what the sign says, but also what, what makes sense for the individual, what they prefer. And then there's the, after the visit and someone leaves and there's often, you know, a note to finish up.
[00:06:15] Spencer Dorn: Or there are results to process or coordination to happen or messages from patients afterwards. So that's kind of like a gen, I mean, it's, it's, it's a little oversimplified and it's not always as logical, but there's the before and there's work that has to happen before, which largely is about getting to know what you're about to see, who you're about, to see what the issues may be.
[00:06:36] Spencer Dorn: And then there's the work. In the middle, which is really about collecting information, interacting, making good decisions, hopefully starting to write notes, writing orders, pulling up literature, et cetera. And then there's the after piece, which is usually more cleaning things up, following up on things, coordinating things.
[00:06:54] Kirk McConnell: And when you think about both the opportunity for ai, but some of the impact you're already seeing, is it spread across that before, during, and aftercare, or is it especially relevant in one of those phases?
[00:07:07] Spencer Dorn: I think it's spread across all, and I'd argue that it's even, you know, it comes from when people are looking for care and even before they're look looking for care and necessarily know what they need.
[00:07:17] Spencer Dorn: Almost all the people I see have seen other gastroenterologists. Sometimes they've seen many other gastroenterologists. So you can imagine they have a lot of records and those records may be locked up in different information systems in both structured and unstructured formats. So a lot of the work I do is the night before I see I go to clinic, I review all these records, and I start writing a note.
[00:07:35] Spencer Dorn: Well, there's AI software that's starting to help summarize patient records. For us, and it hasn't really hit a tipping point yet, but I believe it's going too soon. And then when we're actually working with someone in person or online. There's a conversation that's happening and often there's a typing happening because right, we want to capture what's being discussed and we wanna write a note and we wanna start writing orders, and maybe we want to look up some things in the record, or maybe we want to look at the literature and there are AI tools addressing all of those aspects.
[00:08:08] Spencer Dorn: Probably the widest used and the best known as ambient dictation or ambient documentation, which, you know, it's, your phone is listening to the conversation you and I would be having. Generating a note as it. It happens. So I don't, I could look up and I'm freed from the keyboard so we can have a conversation without my being anxious or distracted by that.
[00:08:27] Spencer Dorn: I have to write a notes and some of these tools are going further to start, you know, adheres that I'm saying, oh, well we should prescribe this medication. It can start pending some of those orders. So AI's having a big impact there. AI's having a big impact on looking up the medical literature and figuring out what's the right.
[00:08:44] Spencer Dorn: What's the right medication to use in this situation? Or people who present with these symptoms, what's the proper differential diagnosis? So there a whole host of, of really effective AI tools There. And then when people leave the office and then there's all this other work to do, like, you know, processing messages or dealing with prior authorizations and various other things.
[00:09:05] Spencer Dorn: There are AI tools being applied there. So, long story short, yes, AI is starting to touch various aspects of care, both the care that people, you know, the processes that happen before a clinician's involved, as well as what happens when a clinician's involved as well.
[00:09:22] Kirk McConnell: I think we've all had those doctor's appointments lately where you almost feel like, Hey, I'm over here.
[00:09:27] Kirk McConnell: I'm over here. Because they're so focused on kind of typing notes. How is it changing your ability to be present during a conversation with the pre, with a patient? If you're not having to actively take notes? Do you feel more engaged?
[00:09:43] Spencer Dorn: I think it depends on who you are. I like to, so I think. I'm a very good typist.
[00:09:48] Spencer Dorn: I have a lot of kind of well-designed templates and macros, and I do a lot of work the evening before to get to know who I'm seeing the next day. So I feel like often I can just usually. I usually feel very comfortable typing without being distracted, and I don't feel like I need some help and I, I sign all my notes within a few minutes of my clinic ending.
[00:10:10] Spencer Dorn: However, other people feel differently. Other people don't like typing and they're not as good at typing. They may not be as efficient using the EHR, and I think for those people, those clinicians. It's very liberating to not feel that pressure of, I've gotta jot this down. I have to write a note. And not only during the visit, but after the visit.
[00:10:29] Spencer Dorn: 'cause a lot of clinicians go home and they, you know, they have a bunch of charts they need to sign off on. And there's this, there's this concept of something called pajama time, which is kind of, you know. After dinner when family's gonna sleep, doctors staying up late at night writing, finishing up their notes and touching them up.
[00:10:45] Spencer Dorn: So I think it can be li very liberating, but I think the point is that it's not necessarily for everyone. If you look at the data, these AI scribes in general tend to be used by about one in three used heavily routinely by about one in three outpatient clinicians. So that means two and three are like me.
[00:11:03] Spencer Dorn: They're either. Using it more selectively, which I do, or they're not using it at all. So I think it usually, my answers to these things are, it depends. It's hard to come up with like, it's always this way.
[00:11:15] Kirk McConnell: Hmm. Yeah. I don't think a lot of people understand that pajama time notion that you were saying of you get the kids to bed, but then you still have two hours of charting to do.
[00:11:24] Kirk McConnell: What impact would reducing that pajama time? Have to just quality of life for doctors?
[00:11:29] Spencer Dorn: I mean, it's, it's easy to imagine, right? Just like any, and first of all, clinicians, physicians are not unique. A lot of professions take their work home. Mm-hmm. So I wouldn't, you know, I wouldn't, wouldn't wanna pretend like we're the only ones doing this, but of course it's liberating.
[00:11:44] Spencer Dorn: Of course it's freeing. Of course. It's makes a difference, right? I mean, if you can. Gain an extra hour to your day to go to the gym, to get an extra hour of rest, to read a book, to have dinner with your family. I mean, of course these are, these are meaningful benefits that I think will help sustain our profession because the work's gotten very difficult for many people in many different ways.
[00:12:06] Spencer Dorn: So I think we need these, these types of supports to help us sustain ourselves.
[00:12:12] Kirk McConnell: Let's flip to look at the other side of the doctor patient relationship and more the patient. Now, one of the things you've written about is go back, you know, a decade, maybe a little bit more. When patients started showing up with printed Google results, how did that change kinda the doctor patient relationship?
[00:12:27] Kirk McConnell: When the internet started to give members more access to information?
[00:12:31] Spencer Dorn: Yeah, we, I remember, 'cause I was a, I was a resident at the time in the early two thousands and I wrote about this. I remember that people would show up with printouts. This was pre right, pre smartphones. You couldn't just look at your phone.
[00:12:43] Spencer Dorn: They would show up with printouts and you know, Hey doc, I read this, or Hey, I. You know, I was thinking about this and a lot of my older teachers recoiled at that notion. It was like, whoa, what's going on here? And there was this like popular coffee mug that showed up that said something along the lines of like, your Googling doesn't equate with my medical degree, or something like that.
[00:13:06] Spencer Dorn: So there was a lot of pushback initially because this was really one of the first times that medical knowledge was starting to. Democratize beyond the profession. You know, our profession has long existed under what many have called a grand bargain, which was, you know, doctors held all the knowledge and the public trusted us to use it for their benefit.
[00:13:32] Spencer Dorn: Right? So we had the privilege of, of practicing and regulating ourselves 'cause we were the gatekeepers of knowledge and the general public. Just trusted. And this was kind of a, a bit of a fraying of that grand bargain in many ways that the, the internet allowed people to look up and find information that was hard to access before.
[00:13:52] Spencer Dorn: So yeah, I think most of us got used to that notion. Many of us welcome that, right? There's, there's a lot of benefit to people being more informed. There's obviously a cost to that as well, because more information can mean more. Disinformation or more empowerment can also sometimes mean more confusion.
[00:14:11] Spencer Dorn: But I think many of us learn that this is just the new reality and how do we embrace that and what's changing now. Is that search is being replaced by language models, right? So people are asking queries and they're getting back direct answers, right paragraphs of answers and not, not only on text there, as you know, there's an opportunity to speak as well.
[00:14:35] Spencer Dorn: So it's just, it's, it's, it's kind of taking this preexisting trend and it's supercharging it and it's bringing it to a new level and it's providing answers, yet the answers. Are very convincing. They may be out of context, but more than just answers, it's providing a type of clinical reasoning that helps people make sense of otherwise raw information from websites.
[00:14:59] Spencer Dorn: So it's, you know, people, it's not just answers, it's, you can ask questions, well, what if I'm experiencing this? Or, what would you do in this situation? So it's not just information from a generic source, it's information plus. More engagement, customization, reasoning, et cetera. So it's, it's a, it's a whole new world in many respects.
[00:15:18] Spencer Dorn: And I think this trend is just starting. I think we're on, it's a dawn of a new era that kind of continues the last generation of, of changes.
[00:15:28] Kirk McConnell: Yeah. One of the things we think about on this podcast a lot is employers who are managing benefits for their people. And so if you were an employer right now. I think you're trying to figure out how is AI changing how my members or my employees interact with the healthcare system?
[00:15:44] Kirk McConnell: So I wanna talk about the future next, but kind of where we are today. What do you think are the positives of this first kind of big AI push, and then what are some of the downsides you're seeing and how members are interacting with their providers?
[00:15:58] Spencer Dorn: Yeah, I think, well, for, as we mentioned, I think democratizing information is a major positive overall.
[00:16:04] Spencer Dorn: The flip side of that, of course, is that there may be information outta context can be confusing, can even be harmful, right? And there's also potential for disinformation. So I think number one I would think about is how are people accessing information before they've even decided to have care? Right? To receive care, right?
[00:16:21] Spencer Dorn: People most symptoms, most. Health concerns don't necessarily lead to someone actually seeking healthcare. So, so I think that's the first. And then if they are seeking healthcare, how are they using tools to help guide them to the right type of care? So we're, we're starting to see a lot of kind of care navigation tools being, and I'm sure I, I imagine that your company is developing some of this, but even within traditional health systems or kind of standalone companies that are helping people make sense of.
[00:16:51] Spencer Dorn: All right. These symptoms, this is maybe what you would consider doing next, and these are the types of places. And then lastly, there are new types of healthcare practices popping up that are AI first. So instead of the traditional doctor's office or the health system, they're actually AI first medical practices that are emerging that just like there were virtual first medical practices that emerged during the pandemic, that some people may.
[00:17:14] Spencer Dorn: Find very attractive and, and lean towards. So I think those are what I would be thinking about. You know, how are people accessing information? How are people navigating care? And then finding the places and the, and, and the venues and channels that they wish to receive care for. Again, generally positive, but the negative is it could be overwhelming, it could be confusing, it could be misinforming.
[00:17:36] Spencer Dorn: People could, could go down the wrong rabbit hole
[00:17:39] Kirk McConnell: as someone who thinks about this a lot. That AI first care, I think is really interesting. What are you watching in that space to understand? Is it gonna be widely adopted or are people gonna be uncomfortable with that?
[00:17:52] Spencer Dorn: I think things take time, right?
[00:17:54] Spencer Dorn: Healthcare changes slowly. We are seeing a crop of startups that are orienting around AI first care. So these are completely new practices that are built with, you know, new ratios of, of clinicians to patients because they're using ai. To do a lot of the upfront work and then only passing to a clinician after a lot of that's been done.
[00:18:14] Spencer Dorn: So like completely different types of models, you know, different types of capacity, really capacity building. So I think that's one thing, but I'd go further. I think even beyond these kind of upstarts we're seeing. And we will increasingly see people interact with AI before seeing a traditional clinician too in a, in a health system, in a medical practice.
[00:18:37] Spencer Dorn: We'll start seeing, you know, rather than the old school clinical intake, which was like a clipboard in the waiting room that you know. That people were filling out, checking boxes or writing things that the Dr. May or may not have, have actually would actually read. We're going to see people when they come into a traditional practice where they're actually interacting with AI first to kind of gather information, go through a discussion, and then clinicians coming in with.
[00:19:05] Spencer Dorn: You know, access to that information.
[00:19:07] Kirk McConnell: That's a really interesting idea. Do you think that providers and health systems are going to try to own that first AI interaction as you go into the system versus being prepared for someone having talked to chat GPT before they walked in the door?
[00:19:23] Spencer Dorn: Yeah, I mean, I think that's a great, that that's, that's a big unknown.
[00:19:26] Spencer Dorn: And you know, there's a question I think a lot about aggregation theory, which you or some of your listeners may have, you know, may know about. You know, is there just going to be a, a funnel, a top of the funnel, that there's going to be a race to control that top of the funnel, which will then kind of determine what happens downstream and has that affect existing health.
[00:19:46] Spencer Dorn: Practices, health systems, et cetera, or is healthcare a little different? And maybe there's only so much you can do at the top of the funnel? Because I, I believe people don't just want information. They want actual care. They want answers to their problems. Not only like, you know, in an, in an information format, but actually treatment results.
[00:20:07] Spencer Dorn: And there's a limit to how much you can do at that top of the funnel. So it's unclear. It's unclear where, you know. AI kind of re shifts and reshuffles how industries work and where power accumulates and, and how things shift. It's unclear to what extent that will ha happen in healthcare compared to other industries, but it certainly will happen on some level.
[00:20:31] Kirk McConnell: Another idea, you hear a lot in the AI delivered care is when it's just tough to get care. For instance, I was talking to someone last week who said, I don't love the idea of taking a picture of a mole and asking Chad GBT if it's cancerous, but I'm more open to the idea if it's gonna take me six months to get in to see a dermatologist.
[00:20:48] Kirk McConnell: So where do you see opportunities for AI to extend care when you just, when there's not enough provider availability?
[00:20:55] Spencer Dorn: Yeah. I mean, I like, I like the, the phrase, I didn't coin it, but I, I adopted it. You know, healthcare has a math problem. There's just too much need and not enough supply. Right. We don't.
[00:21:06] Spencer Dorn: Turn out physic. It takes a long time to become a physician. It takes a shorter time to become a PA or nurse practitioner or certain types of therapist, but it still takes a considerable amount of time. And then there's a lot of time to train up once, once schooling is completed in those areas. So it's just like we have this mismatch of tremendous demand, not enough supply, inability to ramp up supply through human healthcare workers fast enough so.
[00:21:34] Spencer Dorn: I think what you know, what you're alluding to is that sometimes AI is the best option, and because it can meet needs more immediately, and if a more significant issue is detected, maybe it can bounce it up and prioritize your hypothetical patient who needs to see a dermatologist being seen faster. And so I think that's kind of.
[00:21:58] Spencer Dorn: Unavoidable. And I'll also say sometimes AI alone is better, right? I mean, I think there's a challenge that, you know, we often imagine, you know, going back to the, it depends. We often imagine that this hypothetical patient, like ideally they'd see a world class dermatologist who's just like, you know, the best, does the best skin exam, and it's so kind that compassionate, and their office is lovely and it smells like roses and all that.
[00:22:21] Spencer Dorn: No, like there are dermatologists. First of all, they can't get in to see a dermatologist. Second of all, if they can, not every dermatologist is world class and great at what they do. So, you know, I think in some cases AI alone may be better than what's available. In other cases, AI is really the only good option and the challenge there is how do we integrate it in ways that make sense to the rest of the system.
[00:22:47] Kirk McConnell: Do you think there's gonna be a dynamic then where. In areas of medicine where to use airline. I'll say the math doesn't math. You know, you're a world geography, you're in Medicaid, and there's not enough primary care physicians out of necessity. Do you think some of the AI innovation's gonna happen there first, or maybe it would be a more interesting laboratory on what's possible?
[00:23:09] Spencer Dorn: Yeah. Or maybe not in the US right. Countries outside of the us You know, I think, I think a lot about how certain countries in, in Africa, they skipped landline. Like there was there a lot of regions of Africa, like they just didn't have. Telephones in, you know, traditional telephones in homes or, or, or, or workplaces and they just skipped straight to mobile, right?
[00:23:32] Spencer Dorn: It's kind of a leapfrogging effect. So I think there are some parts of the world where just there's such an undersupply of healthcare that AI type tools may kind of. Leapfrog building out, you know, a broad healthcare infrastructure. So I think that's, that's likely, and I'm aware of some efforts to make that happen.
[00:23:51] Spencer Dorn: And then, yes, I think within the country, you know, the question is. Will certain underserved areas, is AI a patch that's more likely to be adopted there? I think in some ways, yes, of course. There's the concerns of the opposite happening that AI won't get to rural areas or won't get to underserved areas because of, you know, digital divide issues.
[00:24:12] Spencer Dorn: And then there's also the worry that certain. Populations may be taken advantage of, or AI that wasn't designed to serve them will be applied to them and will somehow be either inaccurate or biased. But yeah, that goes back to the, it depends. All these, all these discussions are nuanced and you know, I think it's important just to be open-minded and approach these things with a sense of.
[00:24:34] Spencer Dorn: Of, of wonder, of wanting to understand because usually like the, the brief soundbites that people gravitate towards, usually they're either overly simplified or just frankly wrong.
[00:24:47] Kirk McConnell: Well let, let's build off of that because one of the things you often hear with new technologies is that as a society, we overestimate the impact they're gonna make in the near term, but then we underestimate the long-term impact they're gonna have.
[00:25:00] Kirk McConnell: How would you kind of critique your view on how AI is impacting medicine right now? Are some of the civics now going to be proved wrong 10 years from now?
[00:25:10] Spencer Dorn: I always, I often quote that am Omar's law, that famous law that it takes longer than it thinks, but we can't imagine like the, the, the eventual output.
[00:25:19] Spencer Dorn: So I, I, I, I fully subscribe to that idea. I also fully subscribe to the idea that a guy named Melvin Kranzberg coined a, a, a historian that. Technology is neither good nor bad, nor is it neutral. So like it's not predestined what the technology does, but it will have an impact. So those are kind of two guiding principles that I follow both Kranzberg Law and Ammar Amar's Law.
[00:25:41] Spencer Dorn: I think we're starting to see some early traction in certain pockets. I mean, we talked about ambient scribes, or we're talking about consumer grade AI tools to learn more about. Themselves and their conditions. So I think things in some respects are happening quickly, but on other respects, things are happening quite slowly.
[00:26:04] Spencer Dorn: I mean, medicine today is not that different than it was. Four years ago before chat, GBT turned everyone onto ai. Things still happen slowly. We, you know, by and large visits still look the same way they did 10 years ago. There's a little sprinkling of some different. Aspects to it, but by and large it's generally the same, and I think that's okay.
[00:26:27] Spencer Dorn: I think that's okay. While understandably, you know, we need to move quickly and we, people are in many ways desperate for newer solutions. There's still a lot at stake and there's still a lot that we don't know.
[00:26:39] Kirk McConnell: I love what you said, that it's neither good nor bad, nor neutral. When you think about kind of the evolution here, what are the ethical or kind of governance elements of this that you think are gonna be most important?
[00:26:55] Spencer Dorn: I mean, the, the notion of neither good nor bad, nor neutral is that the technology has no destiny of its own, right? It has no, these are tools, right? These are tools. It's ultimately up to us as a society, as organizations, as individuals to figure out how we utilize these tools for, to maximize benefit and mini minimize downsides.
[00:27:16] Spencer Dorn: Um, so organizations like mine have created governance. Around how we use these tools to make sure that they're being harnessed safely and effectively without, you know, disadvantaging particular groups. So governance is super important and it's something that all good organizations. Are doing now and we'll continue doing.
[00:27:39] Spencer Dorn: This is not just like a one-time project, this is like an ongoing, we're looking at a new tool. How do we evaluate its potential benefits, its potential drawbacks so that we can make the best decision to serve those that we're responsible for caring for.
[00:27:55] Kirk McConnell: How do you think if you turn that governance and ethical question more to the consumer side?
[00:28:01] Kirk McConnell: How do you think or should, are you worried about how members are under, or patients are managing that as well? Do they understand that if they submit their medical records to an AI model, it's not the same as, you know, talking to your doctor about something in terms of governance and data protection? I,
[00:28:20] Spencer Dorn: I think it all goes down to, it all goes back to trust.
[00:28:23] Spencer Dorn: And if you look at national and international surveys, people, Americans. But also those outside the us, they're pretty distrustful of the government that are also fairly distrustful of big tech companies. What. People trust that, at least right now at this point in time, is they trust their friends and people they know personally and they trust their clinicians.
[00:28:46] Spencer Dorn: So I think that's a big role that we must double down on, is that like forming the trusting relationship so we can help individuals make sense of information that they're getting from, you know. Consumer facing tools, which may be really great and may be really accurate. It also may be too much, it may be overwhelming, it may be wrong, but I think that's, I think that's the role that our profession will increasingly play over time.
[00:29:11] Spencer Dorn: And I imagine likewise with in the benefit space a lot, there are great opportunities, I think, to help members make better sense of the information that they're receiving and help them make better decisions. But again, I think it's all about the, those trusting relationships.
[00:29:28] Kirk McConnell: What do you think is gonna be the force pushing innovation forward?
[00:29:33] Kirk McConnell: Is your hypothesis that some of the big tech companies are gonna make the biggest gains? Is it going to be a bunch of kind of one-off smaller startups? Is it gonna be a combination? What do you think?
[00:29:44] Spencer Dorn: I think the force pushing it forward is the market, right? The healthcare is about 20% of the GDP. So most large companies want to be involved in healthcare just because it's hard to ignore and market forces, whether it's through, you know, startups and venture funding wanting to recoup, recoup investments or publicly traded companies, I think it's.
[00:30:04] Spencer Dorn: Unclear whether it will be big tech companies leading the charge, or smaller startups. Both obviously have advantages and disadvantages, and then they're in there. You know, they're incumbent companies like electronic health record vendors that really sit in the middle of all this and in some ways have the strongest position because all the workflows flow through their product and they have the distribution.
[00:30:26] Spencer Dorn: So it's unclear and I think it would be hard to say, oh, it's going to be one specific aspect of the industry that will. That will ultimately thrive. It'll probably depend, again, it'll probably depend on, on the problem space, on the potential solution, and we'll probably see a mixture of, of, of, of winners and losers in those areas.
[00:30:46] Kirk McConnell: What are the questions in this space that you think are most interesting right now? Let's look over the next decade. If you could go and talk to yourself in 10 years, what questions would you love to ask about? What's gonna happen over the next 10 years?
[00:30:59] Spencer Dorn: The question that really motivates me most right now is just what does it mean to be a physician in this new world?
[00:31:07] Spencer Dorn: And, and even more so, what does it mean to be a human being as, as things are changing so quickly. So, you know, that's maybe too high level, but to me it's just trying to figure out. What, how do our roles and responsibilities change? How do our relationships change? What do we gain? What do we lose with these new digital technologies?
[00:31:27] Spencer Dorn: You know, what are the things that we're not thinking about but need to be thinking about more? What are things we're holding onto too tightly that it's time to let go of? So those are the, the general questions that I'd love to ask my future self, or, you know, anyone in, you know. Looking at into the future is like, you know, what do things look like and, and, and, and, and how do we, how do we keep our eyes open and make the best decisions that we can so that we don't go down the wrong paths, which inevitably we will, but hopefully we'll have the awareness to recognize we're going down the wrong path so we don't keep going, and we'll have the agency and autonomy to, to, to, to shift gears.
[00:32:04] Spencer Dorn: So yeah, those, those are, those are some of the questions that, that really interest me now and I think will continue to interest me moving forward.
[00:32:11] Kirk McConnell: Well, you're right, those are big questions, but I think they're super interesting ones and actually a pretty good note to end on. So, Dr. Dorn, if folks want to continue this conversation with you, how can they reach out to you?
[00:32:23] Spencer Dorn: They could find me on LinkedIn. They could find me on Forbes, they could find me. I'm not really in other social, so social networks, I'm in Chapel Hill. They could find me at, they, they could learn a lot about the, the great work that's happening at at UNC Healthcare and UNC School of Medicine. So.
[00:32:37] Kirk McConnell: Well, Dr. Dorn, thank you so much.
[00:32:39] Kirk McConnell: I wish we could do this conversation in 10 years to see how some of these things have changed, but for now, thank you so much for joining me today in your thoughtful perspective.
[00:32:47] Spencer Dorn: Thanks so much, Kirk.
[00:32:49] Kirk McConnell: If you're enjoying the Benefits playbook, we'd love your support. Take a moment to rate interview the show wherever you listen.
[00:32:55] Kirk McConnell: It really helps others discover us and join the conversation. And while you're there, don't forget to hit Subscribe, so you never miss an episode. If you know a colleague or friend who'd enjoy the conversation, share it with them too. Thanks so much for listening, and we'll see you next time.