.jpg)
The MD+ Podcast
MD+ is a community of aspiring physician-innovators who are exploring unique career paths with the ultimate goal of positively transforming healthcare.
Created by Geoff Bocobo & The MD+ Executive Team
_____________
(Podcast Music from: https://pixabay.com/music/corporate-wake-up-to-the-renaissance-135540)
(MD+ Logo Image by: Heath-Rutledge Jukes)
(Buzzsprout Background Photo courtesy of: Tippapatt)
The MD+ Podcast
“Plus One,” Ep. 1: Who Certifies AI in Healthcare? A Conversation with Dr. Brian Anderson
Welcome to Plus One, MD+'s new video series featuring insightful 1:1 conversations with physician-trailblazers who took their medical degree somewhere both unexpected and impactful.
In this first conversation, Geoff sits down with Dr. Brian Anderson, CEO of the Coalition for Health AI (CHAI). They discuss his path into AI, how CHAI helps build consensus across 3,000+ member organizations, its partnership with The Joint Commission, and the Applied Model Card (a "nutrition label" for health AI). Brian also shares considerations that health systems should think about when interfacing with AI vendors, and how trainees can get involved in AI and digital health today.
Originally recorded at Big Night Studios on June 15, 2026 in Boston, MA.
Special thanks to Brian and the entire CHAI team for making this episode possible!
Video edited by Matt Dunlap.
🔗 https://www.chai.org/
https://mdplus.community/
So, hey everyone, welcome back to the MD Plus Podcast. I'm your host, Jeff Bacobo, and I'm very excited to be joined today by Dr. Brian Anderson, the CEO and co-founder of the Coalition for Health AI, which is a nonprofit that he helped build from scratch that now connects many organizations shaping how AI is deployed responsibly in healthcare. So before Chai, Brian was Chief Digital Health Physician at Mitre, where he worked directly with the White House COVID task force and Operation Warpsfeed. He trained at Harvard Medical School, has advised government and industry alike, and is one of the most important voices in health AI government today. So, Brian, welcome to the MDP Plus Podcast. It's so great to have you here today.
SPEAKER_01Well, Jeff, thanks so much for the warm welcome. It's clear that you got the note that my mother sent you for the introductions there. Um, but yeah, I'm looking forward to our conversation. Awesome. Thanks so much, Brian.
SPEAKER_00Yeah, so first off, I always love to start with a career path question. So would love to hear about your path from medicine into co-founding one of the most influential AI organizations in the country.
SPEAKER_01Yeah, no, so you know, I think anybody who's in digital health, who's also a physician, has an interesting story on kind of how their career uh path uh took. So starting off at Harvard Medical School, you know, didn't have any AI courses, certainly there, didn't have any digital health courses. There were maybe the beginnings of a few informatics courses. I was lucky enough uh to have taken a course, uh, the Dr. John Holumpka, who was then I think the CIO at Beth Israel Deaconess, and I was really impressed with kind of everything that he was accomplishing seeing here, an emergency room physician who's teaching, who's also a physician, uh, that sort of kind of opened my eye my uh mind to the potential. Fast forward, uh did my pediatric residency here at MGH and then began working up north of Boston in the Lawrence Andover area. Um, and it was in that place that I started working at a federally qualified health center in the Greater Lawrence area and began appreciating the challenges of EHRs. Um and so my wife and I were recently married, we wanted to start a family. I found I wasn't getting home until like 8 30 each night, and it was principally because the EHR that we had sucked. And I thought, naively, um, that I could solve all these problems if I went solo and opened up my own private practice and I chose a good EHR. And so I started looking at the time, this might date me somewhat, um, there was only one EHR that was on the internet. Being on the internet as an EHR was a big deal uh at the time. And so Athena Health was the name of the company, and so I was about to start working for them, uh, excuse me, uh, sign with them as my as my EHR. Um, but a buddy of mine who was working there said, you know, you should take a look at the company. They're looking to hire and grow um their clinical teams to build out the suite of tools and services. Um, like any EHR, Athena started as a revenue cycle management, but it was growing into the electronic health records space. Took an interview, hit it off really well with the CMO and the CEO, and you know, from there I, you know, started working, um leading the clinical services team, helped building out the EHR product, and that's really where I cut my teeth in digital health. Started leading an AI team there. Um, and then subsequent to that, a very persistent recruiter at a government contractor called MITRE reached out saying, you know, great that you're working for a, you know, um a publicly traded company, but you know, don't you want to do something bigger for your nation? And that sort of kind of pulled at my heartstrings. Um, then took a job at MITRE, uh, became the chief digital health officer there. And it was really in that spot that um when the pandemic hit, um, the FDA and the NIH asked me to help lead um some of their data analytic efforts associated with monoclonal antibodies. That if you remember during that time, that was one of those first medical countermeasures that we had before the vaccines came out. Um so did that, and really it was in that space, working at the intersection of public and private sector companies, um, that we launched Chai in the middle of the pandemic.
SPEAKER_00That's fantastic. Really great journey, and it really informed kind of the work that you're doing now as well as how you think about the next generation of trainees too. So curious, um, in general, what do you think the value is of having clinicians actually working um on the ground with AI?
SPEAKER_01Oh, it's huge. Um, I think so a couple thoughts on that. It is incredibly important for frontline clinicians to have the kind of skills they need to think critically about AI. I think so many of us, when we see AI and the outputs that it generates, it's almost like it's magic. These things are conversational, they're insightful, they're very confident. Um, and it's really important in high-risk situations for clinicians to have the training and understanding about how these AI tools work, um, the limitations of these tools, the critical questions you should be asking about these tools, like was the model that I'm using on a patient in front of me, was it trained on patients like the one in front of me? Because if it's not, you know, there are questions or additional questions you should ask. And so I think from an end-user standpoint, it's really important. I think also importantly, particularly in development of AI tools that help for clinical clinical decision making, it's really important to have physicians or nurses, whoever the end user community is, to really participate in those building conversations with technology companies. So you're not just having situations where it's just AI engineers are building this. It's important to have the nurses, the physicians, and the patients part of those development conversations.
SPEAKER_00Absolutely, yeah. It's super important to have all stakeholders involved. I I completely agree. Um, and so now Chai, what you've built um and what you've you know helped build over the the last few years, is really living proof of that thesis, right? So would love to hear for folks that uh don't know in depth what Chai is. What is Chai and what is it not?
SPEAKER_01Yeah, no, so again, back to the middle of the pandemic. So uh we're now in March of 2021, and a lot of the private sector companies have been meeting alongside with the public sector, so along with the U.S. government and the state governments that we were all working with. Obviously, we were all focused on the pandemic, but I still remember in many of those conversations, um, in many of those meetings, we were using AI all the time. Um AI was being used to design new molecules, some of which would become monoclonals, some of which would become some of the basis for vaccines and other antivirals. We were using AI to predict where the next emerging variant was gonna come up and what was it gonna look like? Um we were using AI to predict all sorts of different things. Um and one of the things that we quickly realized was if you fed an AI model, two different AI models, the same data, those two models were gonna come up with very different recommendations or outcomes. And so that begged the question: do we have a common definition of what good, responsible AI looks like? And we quickly asked um the folks that we were working with in Operation Warp Speed, and the answer we quickly came to is no, we don't. We have really good work happening in different health systems and in different tech companies, but not a technically specific definition of what good AI looks like. And so that was really the animating start of the Coalition for Health AI is bringing together the private sector alongside the public sector, building technically specific best practices or industry standards on how we should think about building AI or developing it, deploying it if I'm a health system or a payer, and how to govern it longitudinally, how to maintain it. What are the metrics that we should be monitoring? I mean, a lot of these are questions we still haven't definitively answered yet because AI is moving so quickly. Um, but that's the basis of it is bringing together, sharing insights, building consensus, coming up with technical levels of specificity, because it's with that technical level of specificity that's really important because without it, you don't have a specific metric on how do I measure the performance, right? Because like you want to generate the evidence to build the trust. When we first launched Chai, it was interesting. One of the things I think grabbed folks' attention, um, and this was like right around the time that Chat GPT was coming out in November, I think, of 2021, was um the level of trust. And if you look at like different surveys, there was one I think from Pew at the time, and it showed that you know maybe 50%, 60% of people were excited or trusted AI. But it wasn't 80%, it wasn't 90%, it was like maybe half of the population. And so we recognized that trust was a big issue. Fast forward to now, right? Like we even see now the trust is even more of an issue, right? Like you look at some of the polls that have recently come out, less than 20% of the population trust AI. And so, I mean, you know, lots of different reasons for that. But what we're trying to do in Chai is bring together those clinician leaders, bring together those technology leaders, and building that common definition of how we can build AI, deploy it, to generate the evidence to build the trust that these tools can't can be used in high-risk scenarios, right? And where the limitations need to be and what those guardrails should be, right? Like I think I would assert that you know we can encourage innovation and develop the kinds of policies to you know run this race for AI as quickly as possible. We can do that at the same time as appropriately understanding what the limitations are and what the guardrails are to keep our patients safe. And that's the mission of Chai, and that's what we've been doing. We started with eight organizations, we now have over 3,000. I think we're the largest health AI coalition on the planet. We have 220 health systems, a lot of exciting things going on. We were just talking about the cyber work that we're going to be launching, so it's never a dull moment.
SPEAKER_00Definitely. No, that's that's really amazing stuff, and I'm glad you co-founded an organization um such as Chai because with technology that's so powerful, um we we really need to be able to address it properly and utilize it properly, right? And leverage it to you know improve patient outcomes and and every everybody improve life for everybody involved. So, with you know, the over 3,000 organizations, uh, many of which probably have slightly or very different incentives, how are you able to reach kind of consensus when you are providing the recommendations with the use of AI?
SPEAKER_01It's hard. Um, in a word, it's not easy. It's a lot of iteration and working together to identify where we have commonalities. I think you know, part of the magic in Chai is bringing together the builders and their customers, right? And so by builders, I mean, you know, the technology companies that are building these. And by customers, I, you know, broadly am saying like health systems and the payer community and the life science companies that you know buy these models and deploy them and use them. Um, and then in addition to that, bringing the patients together. And so when you bring those three groups together, there is an alignment of incentives, right? You can imagine if you're a builder, you want to understand what your customer's definition of good is, right? And if you're one of these customers, like a health system, you want to be seen by your patients as trustworthy, as safe, as like the health system that you go to when you're when you need your health to be taken care of. And so I think when when we establish that kind of aligned incentives, it's easier to find the consensus points at a technically specific level. It can be challenging. Like, don't get me wrong, like there are a lot of really hard conversations that we have. Like, yeah, I can talk to you a little bit about some of the work we did with our model cards and the challenges in building consensus around those. But um I think there's a lot of a reminder that like what we're here to do is to one, understand what the common definition is, right? Because that's how you accelerate innovation, right? That's how technology companies can sell into their customer base is when they have that common definition. And that's how patients build trust with their providers that and provider organizations are trying to reinvent themselves, right, in this age of AI. They need to. And so how do you do that and not lose trust? You do that by building this common definition.
SPEAKER_00Definitely. No, that that's super helpful and it dovetails nicely into my next few questions here. So I guess a two-part question is you know, you mentioned that applied model card. Um, how is that I guess actually sorry, you described the applied model card as a nutritional label for health AI. Um so what's on that label, what problems is it solved, and then how does it help organizations that are maybe making some overlooking some aspects of AI governance?
SPEAKER_01I think one of the first things we appreciated when we launched Chai was that the point for maximal leverage for the customers is in the procurement phase. So if I'm a health system or a payer or a life science company and I've put out an RFP, um, one of the things that we in earlier uh uh days of Chai, we appreciated is that a lot of these procurement decisions are being made without the kind of data that tech that that health systems or payers want to have, right? Questions like how does the model perform on patients like those that my health system has? Or what sort of training data was used to train this model? Or great that you're saying it works well on these populations, are there any limitations? And what are those limitations? Like these are very basic questions, right? And what many health systems were appreciating is they were getting equivocal answers, they were getting sometimes no answers, or unclear answers. And so what we wanted to do is again, building that common definition of what good looks like, is to have a conversation with the builders and their customers to establish what are the minimal set of disclosures that the tech companies would be comfortable sharing, that the health systems really need to make that buying decision, and to do it in a way that doesn't, you know, reveal the IP of the tech company, um, but builds the kind of data points that the health system or the payer might need to actually spend millions of dollars to procure something. And so that was the basis of launching this AI nutrition label. And so it has 31 or 32 categories of different things. Um of the most important, if you ask me, are what sort of data was used to train your model, right? Again, if you're particularly in the clinical decision-making space as a use case, you want to have that model trained on sets of data like those that it's going to be used on. Of course. And then so that then begs the next set of disclosures. What are the indications for use? And importantly, what are the limitations for use? I'm a provider, I can readily tell you lots of providers use drugs in an off-label way. Similarly, lots of providers and nurses use AI models in an off-label way. Oftentimes that's safe, sometimes it's not. And you really need to be able to understand which category the use you're about to use falls into. And that can be informed by what are the limitations, right? Like if there's an AI model out there that doesn't have any limitations listed, you should start asking questions, right? Yeah, you should be a little bit suspicious about that. Like, did the tech company not do sufficient, robust enough subpopulation testing to understand what subpopulations of people or types of scenarios where that model shouldn't be used? Because nothing can be used on every single person and every single use case. And so it's really a kind of robust, intentional, rigorous way of understanding these models, similar to a nutrition label that you and I and everyone listening is familiar with. When you go to the grocery store, you turn a can of soup around, you see a nutrition label. Um, that's how we wanted to make these is make them understandable for the physician leaders at health systems to make their procurement decisions.
SPEAKER_00Absolutely. No, it sounds like it's been super impactful and it will be for years to come. Um so that's fantastic. Um, you know, there's so much great work that Chai is doing, so we'd love to kind of uh change directions a little bit and hear about the recent partnership with JCO and the recent um like press releases publications that have been coming out. Yeah.
SPEAKER_01You know, it's interesting. When when we launched Chai, a lot of the health systems were really excited about the procurement base. That's one of the first things that we developed was the model card. And then the next year, we started hearing from these health systems: oh my gosh, we just spent five, ten million dollars buying all these things. Now what do we do? How do we manage these things? How do we govern? And that became, I think, the existential question that we still have today, which is health systems, many have negative margins, right? They're in the red. And yet they're being, you know, pushed um to make big procurement, big capital expenses uh with AI infrastructure, enterprise platforms, you know, et cetera. Um and then what they're quickly realizing is that it is not a one-and-done capital expense. It is a set of ongoing infrastructure expenses that are very much in a space that we don't have definitive consensus on on the you know, on how to do AI governance across all sorts of use cases. This is an emerging space. How do you govern and maintain AI models, right? As I think probably all of your listeners appreciate, AI models, particularly in the generative, non-deterministic space, are um not uh static, right? They're going to change. Newer versions, you know, come out every week, and you know, breathlessly we hear about emerging capabilities, be it mythos or fable or you know, something like that. And so how you think about monitoring the performance of these models can in some ways seem like a moving target. Like what is the metric one day might not necessarily be the metric that you're focused on the next day because the use cases may change. And so governance is a real important issue. And on top of that, how to do that in a financially sustainable way is a huge challenge for all the health systems in America, at least, that are part of Chai. And so that was what was compelling when we were approached by the Joint Commission to create a partnership. Um, if you're familiar, if your listeners are familiar with the Joint Commission, they know that the Joint Commission since the 1970s has been certifying health systems for patient safety and clinical quality. AI is a tool. AI is going to be very relevant to patient safety and clinical quality. And so I think the Joint Commission rightfully understood that AI is an important consideration as part of the certification process. Um they recognize that Chai has all the experts developing these consensus-driven best practices. Why shouldn't we do one on not necessarily a use case like sepsis or um you know stroke, but something more broad like AI governance? And so that was when we created that partnership, we launched that effort. Um, over the past year and a half, about 150 health systems have partnered together to create a set of eight playbooks. It's like 150 pages long, it's quite long, but it boils down to 47 controls. And so I would encourage your listeners to take a look at the executive summary and look at those 47 controls. They're things that every health system, every payer should understand and be working towards implementing because that's how you implement AI governance in a responsible way. And that's how you set yourself up for the future when we have real challenges like agentic tools, right? That have multiple steps, multiple complex tasks to complete. How do you orchestrate? What are the right entitlements that you should give to agents? How do you make sure that PHI doesn't leak out when your agent goes and says, I want to go access Google documents or some other non-HIPAA compliant tool? These are all real questions that health systems are struggling with today. And how do I do a root cause analysis of an agent that has, you know, six or seven different steps before it completes a task? Um so AI governance is critical, particularly in the world of agentic AI, which is a future where we're all headed.
SPEAKER_00Absolutely. That's that's super powerful stuff. So it sounds like a lot of what Chai is doing is creating tangible shared intelligence, right, um, that addresses a rapidly evolving technology and AI and generative AI as it applies to healthcare. So you have the um applied model card for procurement, which is definitely tangible, it's open source on GitHub. It's deployed at several academic medical systems as we discussed, and then you have the partnership with JCO um for the governance piece um and their playbooks there. So, what does success look like for the use of these in a few years? Will they um undergo, you know, a flywheel by incorporating new best practices? Or do you think you're gonna move on to something else um in the future to after after procurement, after governance? Um is there or both?
SPEAKER_01Yeah, it's a great question. Um, I think governance is always gonna be with us. And I think one of the things that we don't have enough of in the AI space is the level of humility around not being certain um that what we're doing right now is the best thing. And so I think with that level of humility, I certainly remind my team and when I share it with our broader community that these are living documents. Um they're not perfect. I can guarantee you the first versions of these are not perfect. And so what we've told our community is please use them, adopt them, tell us what's not working, tell us what we can do better, right? These are gonna be iterated on. Um, so we're committed in Chai to you know releasing versions of these annually or sooner as needed. Um but for example, the agentics space rapidly evolving, right? Mythos class models coming out. We're gonna need to develop new best practices and considerations for governance. And so I can you know entirely expect next year we're gonna have an updated version of this. Now, what does that mean? mean for health systems? I think health systems, I mean, we actually just had a call with um a group of um uh health systems that are on our advisory board a large number of them said you know we want to start adopting this right away um I think health systems recognize inherently the risk in using any kinds of new things like AI tools and working to establish industry best practices help protect health systems from the kind of inappropriate use right if you have all the smart people in the room saying I think this is the best way let's let's do it in a safe way and and learn from each other that's the vision of the learning healthcare system that you know the Institute of Medicine back in the early 2000s established. We we need to create that space where people can share and iterate and improve on things moving forward.
SPEAKER_00Absolutely um I completely agree and it it's great to hear that there's going to be you know continued updates with with Chaiza, you know shared intelligence as as I called it earlier because it it seems to be doing great work already out there in the health systems. Yeah. So on on the health systems piece we want to try to make this pretty tangible maybe for some health systems leaders. Do you have any tips or um anything that they should be asking vendors about AI that they might procure that they might not be asking at the moment or what's like maybe a theme that you've seen when you're engaging with these um health system leaders?
SPEAKER_01Yeah no great question so we're actually and we have a a a group that is actually building out essentially a um a set of collateral or templates for health systems to use moving forward. You know Jeff uh if you're familiar with the college application process uh maybe I certainly when I applied to college there was something called the common app, right? And you could you could fill it out and you could send it to like as many colleges as you want. Many of the tech companies are frustrated when they go and they try to sell to health systems. Health systems say, you know, one health system says great here's a 50 page set of questions please respond. Another health system says you know here's a 20 page another health system says here's a hundred page and all the questions are different. So it's a real burden for a tech company to respond. What we've done is we've created a group of tech companies and health systems to come together and say okay what goes into our common app that health systems say yep this is sufficient. These are the questions tech companies say great you know we're not revealing IP in responding to these so yeah you can put this on a common app. And so I would encourage to your question all health systems that are interested in understanding kind of what goes into that common app reach out to Chime like this is stuff that we're gonna be publishing for the broader community we hope that it accelerates that kind of builders and customers coming together to innovate and do cool stuff that's going to improve people's lives and it's going to be less of a burden. Things that go into that common app show us your model card. What sort of uh performance metrics can you share on how the model has performed to date have any of those performance characteristics been validated by another health system or a third party entity rather than just coming from the benchmarks that a health that a tech company might establish for themselves? Those are some of the questions that kind of go into that common app.
SPEAKER_00Absolutely no that that sounds great and the analogy to the common app is a great one that I'm sure will resonate with many of our listeners. I'm not sure if it's still around today but um yeah folks will know what that is um definitely so you know a lot of this governance talk um and rightfully so is on the buyer side how does the model card or how is Chai potentially um moving into like recommendations that help startups and maybe physician founders who are building and not on the procurement side if that's an area of that will be addressed or is being addressed.
SPEAKER_01It's a great question. So first let me just say like in Chai I am so thankful with the startups that are already engaged. You would think based on the conversation we've had that you know the most active group is the health systems. They are very active but I would tell you that the most active group are the entrepreneurs at startups. We have over 700 of them and you know I mean they're hungry right they're scrappy. They want to understand what the pain points for their customers are and so they're very actively engaged in the working groups. And so my advice would be similar understand what the problems are that your customers are facing. Meaning have a deep understanding of the workflow challenges. I think there's this challenge in health systems. I mean health systems in some ways are the most complex organism that humans have ever created right you imagine a hospital and the complex workflows and stuff going on it can become quickly sclerotic right like some of these systems and processes that were created for health systems were created back in the 70s before you know computers really got you know widely adopted. With AI we're having a similar challenge right we have existing workflows that have worked fine that have had just humans involved but now we have these amazing new tools that could you know move from what we might think traditionally as just a tool to being a member of a team right like agentic tools can be real members of a team doing you know complex tasks. And so I think in addition to knowing the problems and knowing your customer really well and what their definition of a good tool or solution looks like having the creativity and the curiosity to reimagine what those workflows could look like in an AI first space I think is something that a lot of tech companies and some of the startups that I've talked to have overlooked. Meaning oh that's your problem oh that's your existing workflow we're just gonna build a tool to plug into that. And I think that might solve the near term solution but it's not going to solve um the real challenge which is a reimagined way of um delivering care. And so I would say the the the final thing I would encourage um entrepreneurs and startups in this space to look at is uh you know we love our healthcare delivery organizations in Chai. The growing most exciting space for healthcare delivery if you ask me honestly is not in brick and mortar hospitals and health systems. It is in the you know quote unquote direct to consumer or direct to patient space. I think there's a lot of opportunity for startups to disrupt this whole space by reimagining that AI first kind of healthcare delivery where it's not coming from a big monolithic health system or you know big hospital, but it's coming from a very agile health system that might be employing a few physicians delivering care to a patient at their home, right? Where we have a bunch of different remote monitoring wearable devices that are you know providing levels of insight that we didn't have five, ten years ago. I think that's going to be a really exciting space for startups to get into as well.
SPEAKER_00Absolutely and um a lot of our listeners will be very excited to hear that startups are actively engaged with Chai because I think that um a lot of what we see is you know the the very strong work that's being done with academic medical systems. So a lot of our audience will definitely be excited to hear that. So definitely definitely reach out and you know if you're a high growth startup or early stage startup that wants to do something with Chai. Great. So this kind of dovetails nicely into the closing question which is as you know MD plus is full of aspiring physician innovators and early career physician innovators so medical students, residents, early career attendings is the majority of our um uh listenership and folks that are in the community so for folks that have not yet engaged with digital health, AI, all this great technology that's evolving, do you have any tangible suggestions or anything that you would suggest to yourself back when you were training?
SPEAKER_01Yeah yeah for folks no I came prepared for this question. Okay nice um so a couple things uh first off the bat I think if you're a med student um or an early career physician resident um and you're really not using AI, first thing is start using it. There are so many easy on-ramp ways to start using it in low-risk um you know HIPAA compliant ways and so you know I think there are a number for example a number of companies out there uh they can be very helpful for an early career physician or a medical student um to start asking questions. Open evidence is an example of one but even some of the um uh big frontier labs like ChatGPT or Claude or Gemini have I think the opportunity space for early career physicians or med students um to start learning what they can do but importantly also what their limitations are and I think that's I think an important you know set of sober perspectives that that every early career pay uh uh individual be a physician or a nurse or a med student should take into this is like it's great what can it do but also what can't it do? What are its limitations? What can I trust? What shouldn't I trust? What should I question? What should I have a critical eye towards um other advice I know a lot of um medical schools don't yet have AI courses but finding the opportunity to take some kind of courses to get a formal education of some level in AI or informatics or digital health is so important. I was you know long ago not able to because we didn't have it AI you know wasn't really part of our our work back then. But there are uh growing numbers of courses that people can take either formally as part of their medical school or you know a virtual online set of certifications. I think the other thing for residents and physicians outside of residency is look to your specialty society. You know as a pediatrician when you think about who are the trusted curators of guidelines for the specialty that we practice it's our specialty society. And it those are the societies that also understand the new and emerging tools that we can use right AI is more than an odoscope it's more than a stethoscope right but it's also a tool and as a physician we were trained on how to use these things. Our specialty societies are going to be the spaces where we learn how to use these tools with the context of the clinical scenarios and specialties that we're in. And so in Chai we've made a major bet in working with those clinician leads in those specialty societies to develop that new content. I think one of the you know really interesting questions that many of us have is like when are specialty societies going to start saying no one has said it yet but I'm waiting for the first one to say you know it is the standard of care to use AI in you know X disease. Right. Right? Because right now it's sort of optional it's a nice to have but very soon these are going to be the kinds of tools that physicians are expected to use. We're not there yet but the specialty societies are going to be the ones that are going to chart that course for us and the closer physicians get to them and understand what those guidelines are the better off we'll all be.
SPEAKER_00Awesome. That that's really insightful stuff and the MD plus community and broader audience and listenership will be uh really ecstatic to hear kind of those very tactical tips there. So Brian this this is exactly the conversation that MD Plus was made for so thanks so much for joining today. This was great. Thanks for having me Jeff this was great. Of course well thanks everyone I'm Jeff Bacobo MD Plus's podcast host and creator and we'll see you next time
.png)
