Most conversations about AI ask what the technology can do. Dr. Matthew Strehlow and Dr. Rana Kabeer ask a different question: how should physicians think? From shadowing computer scientists to looking beyond diagnostics, they explore why curiosity, not technical expertise, may become medicine's most valuable clinical skill. And they discuss practical ways clinicians can begin using AI today, why the greatest opportunities may lie in improving everyday workflows rather than replacing clinical judgment, and how medical education must evolve to prepare physicians for a future where AI is always available. At the heart of the conversation is a reminder that while technology will continue to transform medicine, the qualities that define exceptional physicians, namely curiosity, critical thinking, and meaningful human connection, remain irreplaceable.
Most conversations about AI ask what the technology can do. Dr. Matthew Strehlow and Dr. Rana Kabeer ask a different question: how should physicians think? From shadowing computer scientists to looking beyond diagnostics, they explore why curiosity, not technical expertise, may become medicine's most valuable clinical skill. And they discuss practical ways clinicians can begin using AI today, why the greatest opportunities may lie in improving everyday workflows rather than replacing clinical judgment, and how medical education must evolve to prepare physicians for a future where AI is always available. At the heart of the conversation is a reminder that while technology will continue to transform medicine, the qualities that define exceptional physicians, namely curiosity, critical thinking, and meaningful human connection, remain irreplaceable.
Learn more about our guest Dr. Rana Kabeer
Learn more about our host Dr. Matthew Strehlow
Learn more about the Stanford Department of Emergency Medicine
(Edited lightly for clarity.)
Dr. Matthew Strehlow: Every generation of emergency physicians enters a different specialty. Some trained before bedside ultrasound, others before electronic health records. Today, a new generation is learning and practicing medicine as AI rapidly becomes part of everyday life.
The big questions aren't really about the technology. They're about us. How will physicians develop expertise when AI is always available? What skills will matter most? And what parts of medicine should never change?
Hello, everyone, and welcome back to the Stanford Emergency Medicine Podcast. I'm your host, Dr. Matthew Strehlow, Vice Chair of the Department of Emergency Medicine at Stanford.
Today's guest is Dr. Rana Kabeer, Assistant Professor of Emergency Medicine at Stanford. Dr. Kabeer was a resident at Stanford when generative AI first entered the mainstream conversation. Training in the heart of Silicon Valley, he witnessed firsthand one of the most significant technological shifts of our time.
With an MPH in epidemiology, a fellowship in innovation and design, and soon an MBA, he brings a perspective that spans clinical care, public health, and the challenge of translating emerging technologies into real-world healthcare impact.
Today, we'll talk about what he's learned from that vantage point, what medicine may be getting right and wrong about AI, and what this moment could mean for the next generation of emergency physicians.
Dr. Kabeer, Rana, welcome.
Dr. Rana Kabeer: Thank you, Matt. It's truly my pleasure to be here.
Dr. Matthew Strehlow: I really appreciate you being here. Today is graduation day for our residency program, and you're speaking to the graduates. Start us off by telling us about the advice you're planning to share with these emerging emergency medicine physicians.
Dr. Rana Kabeer: It's always interesting when you suddenly find yourself in the position of giving advice you didn't realize you had.
It's full circle today. Our new intern class just started as we're sending another class out into the world. It naturally makes me reflect on my own journey, and I'm sure it does the same for you.
Graduation is one of my favorite days of the year. A few years ago, I started a tradition where senior residents are sent off by their junior residents through speeches. I've always believed that building culture and community is what makes this place special.
You work alongside people every day. You learn from the senior residents who seem to know everything, but you don't always hear the gratitude or recognize the legacy you leave behind. I love watching junior residents stand up and speak about a senior resident who profoundly impacted them. Seeing those connections reminds graduating residents that they're leaving this place, but a part of them stays behind through the people they've taught.
I always encourage our graduates to reflect on where they've been, where they're going, and the connections they've built. To me, that's truly the meaning of life.
Dr. Matthew Strehlow: You have such a warm personality. You're a community builder, yet you're also deeply interested in AI. For many of us, AI feels like a cold technology.
When you were a resident and generative AI entered the mainstream conversation, what stood out to you? Why did it capture your interest so quickly?
Dr. Rana Kabeer: That's a good question. AI almost feels paradoxical. There's a lot of excitement, but also a lot of doom and gloom surrounding it.
I'm probably on both sides, depending on which side of the bed I wake up on. Is this technology going to save us, or could it lead us astray?
I've always been a tinkerer. I was the indoor kid growing up. I played sports occasionally, but I spent most of my time on the computer, playing games, experimenting with programming, and figuring out how things worked.
When AI emerged, I didn't really know what it was. I kept hearing about GPT, Claude, and these new models, so I started experimenting. I kept asking myself, What is this? I wanted to understand what was good, what wasn't, and how the technology was evolving.
GPT-4 came out in late 2022, toward the end of my residency, so it was all brand new. We were all asking the same questions: What is GPT? Is this the new Google? What exactly is this?
For me, it meant reading about it, studying it, and then watching it evolve over three months, six months, and a year. Now, four years later, the landscape is dramatically different from where it was when I started. I can't think of another technology that changes this quickly or advances so dramatically that you're essentially operating in a different world than the one where you first learned it.
Dr. Matthew Strehlow: It really is remarkable—and a little scary—to watch that pace of change, because medicine typically doesn't evolve that quickly. We still rely on evidence generated decades ago.
Now we're seeing technologies entering our work environment and changing the way we care for patients every six months.
You've spent a lot of time engaging with AI. You've studied it, built with it, and evaluated AI applications. Where do you see the greatest opportunities in the near future?
Dr. Rana Kabeer: The biggest takeaway I share with people—whether they're enthusiastic or skeptical—is simply to start using it. Experiment with it. Figure out what you want from it.
Take even routine, everyday tasks and ask yourself how these technologies might fit into your workflow.
For me, it starts with a simple belief: education is the greatest equalizer on the planet. When everyone has access to the same knowledge—or at least the opportunity to learn—it has the power to lift people up across generations and bridge longstanding divides.
That belief is personal. I grew up in a family that didn't have much, and I relied heavily on my family's encouragement to do well in school. Those values helped create opportunities that ultimately brought me here.
My story isn't unique. We know so many people with similar experiences. What captivates me about AI is that it has the potential to raise everyone's baseline, as long as people have access to it. Anyone, anywhere, can learn from these tools.
You do a great deal of global health work in low-resource settings, and my background is in international health epidemiology, so I've spent time in many developing countries. When I think about AI, I think about people who suddenly have access to educational resources that previously weren't available to them. That's incredibly powerful. It gives people the opportunity to learn, grow, and compete on a much more level playing field.
To me, that's the most exciting aspect of this technology: its potential to educate everyone.
Dr. Matthew Strehlow: I agree, especially from a global health perspective. In lower-resource settings, AI has tremendous potential to expand access.
We've been working with another country to launch updated clinical care protocols, and for more than a decade we've struggled to disseminate that knowledge despite extensive training efforts involving tens of thousands—if not hundreds of thousands—of clinicians.
Now, if AI can deliver decision support directly at the bedside while clinicians are caring for patients, it has the potential to be transformative.
I see a lot of the positives. Every once in a while, though, I wake up on the other side of the bed and think, Okay, we're all going to have AI overlords, and none of us are going to have jobs.
Dr. Rana Kabeer: Or we'll be measuring productivity versus slop.
Dr. Matthew Strehlow: There's certainly a lot of slop out there. Sometimes I come across AI-generated content and think, You couldn't even run this through another AI program to clean it up?
Dr. Rana Kabeer: We're just generating content, baby. That's it.
Dr. Matthew Strehlow: Exactly. Lots of content.
With all the AI work you've been doing, where do you think the hype is outpacing what's actually useful?
Dr. Rana Kabeer: Can I give you a hot take?
I think we're so focused on questions like: Can an LLM pass the MCAT? The LSAT? Medical boards? Step exams? Can it diagnose patients?
Diagnostics are an especially attractive area because that's what medicine has always centered on. Since the beginning of human history, we've been trying to answer the same question: How do we diagnose and care for people?
But AI isn't the first technology we've used for that purpose. Before AI, there was Google. Before that, there was WebMD. As the internet evolved, people increasingly turned to online resources for health information. To me, large language models are simply the latest iteration of that.
The real challenge is making sure we're evaluating that information appropriately. How do we critically assess what these tools produce instead of simply accepting it?
Is diagnosis the exciting part? Maybe. I think it's interesting. But I always come back to a story from one of my MBA professors.
He was consulting for a major retailer a few years ago, when AI was still in its early stages. After spending a couple of weeks with the company, he presented a plan and said, "Using these AI tools, I can save your company $75 million a year simply by optimizing your shipping and logistics."
Those aren't the kinds of problems people typically think about solving with AI.
The response he got was, "That's great, but we're actually trying to come up with new flavors for our drinks. Can AI help us with that?"
My professor looked at them and said, "I can save your company $75 million a year, and that's not what you're focused on?"
That story has always stuck with me. We tend to chase the flashy, headline-grabbing applications of AI. I'm all for people experimenting in different areas, but I think we're often drawn to the idea of an all-powerful technology where you wave a magic wand, plug yourself into a computer, and instantly know exactly what's wrong.
Instead, my approach has been to build small side projects.
Whenever I encounter something in my daily work—reading papers, evaluating residents, or doing tasks we've always done the same way—I ask myself: How could AI improve this? How could it turn something that takes hours into something that takes 30 minutes?
It may take time to build the solution upfront, but the payoff over time can be tremendous.
If I could offer one piece of advice—not just to graduating residents, but to anyone—it would be this: look at the tasks you perform every day and ask why you're doing them that way. Is there a more efficient approach? Could one of these tools help you accomplish the same work more effectively?
If the answer is yes, you'll be surprised by how much time you get back.
Dr. Matthew Strehlow: There's a lot there, but I want to come back to something you mentioned about practicing and experimenting with AI, because I think that's where many of us are today.
The Raise Health Symposium was held recently at Stanford, and the vice president of health at OpenAI made a comment that really stuck with me. He said that more than 99% of users are only tapping into about 1% of what today's AI models are capable of.
Beyond simply experimenting, where do you think the average clinician can begin accessing that other 99%? Right now, it feels like it's hidden behind a curtain. I don't even know what I don't know.
Dr. Rana Kabeer: I think that's probably true. Can I prove it? No. It's anecdotal, and it's a great sound bite. But I do think there's some truth to it.
In emergency medicine, I think about AI in three broad categories.
The first is clinical decision support. Are we using it to help generate our medical decision-making? To check our work? To brainstorm differentials? Or simply as a thought partner?
The second category is the patient encounter itself. That includes ambient scribes, preparing dispositions, assisting with prior authorizations, and other tools that reduce administrative burden.
The third is operations. There are countless companies working on triage, patient flow, and making healthcare systems more efficient.
Those are the major buckets where I see people applying these technologies.
The question I always come back to is this: learning to use AI is like learning a new language. Do you only know a handful of words, or do you actually know how to communicate? Are you willing to explore, experiment, and gradually build fluency?
The other thing I think about all the time—and encourage everyone around me to do—is to ask people how they're actually using these tools. One of the great advantages of being at Stanford is that we're surrounded by people experimenting with AI in different ways.
I often ask, Can I just watch your workflow?
One of my best friends is a computer scientist, and so is my brother. Sometimes I'll call one of them and ask, "What are you working on today? Can I just sit with you and watch? I want to understand how you think."
We don't do that very often. Maybe we do it during residency, when we're constantly learning from senior physicians, coaches, and mentors. But why don't we apply that same approach to learning new technologies? If we have access to people doing interesting work, why not observe how they think and solve problems?
At the same time, these AI tools can become teachers themselves.
I tell residents this all the time. I learned a little JavaScript back in eighth grade, building simple websites on GeoCities. My coding knowledge isn't particularly advanced—I probably know about 1% of what a professional software engineer knows.
But now I can open GPT or Claude and say, "Teach me how to code. Let's build a simple RPG game together."
Every day I spend a little time doing that, I learn something new. Over time, I've learned far more than I would have otherwise simply because I have a tool that can teach me as I go.
Dr. Matthew Strehlow: This idea of coding in plain language—is that something you think the average emergency physician should experiment with?
I'll be honest: I've never coded anything in my life. I think all of my kids know at least a little coding, but I've never written a line of code, whether in plain language or a programming language. Is that something I should just start playing with?
Dr. Rana Kabeer: I think so. And it doesn't have to be coding for coding's sake. Start with something you're genuinely interested in.
Lately, I've been fascinated by wearables. I've been using continuous glucose monitors, an Oura Ring, and a few other devices. No sponsorships here—I wish I were getting paid, but I'm not.
I started building a simple database of my own data and asking AI to create a dashboard for me. I wanted to track things like sleep, activity, and glucose levels, then look for patterns and correlations I might not notice on my own.
That led me down all kinds of rabbit holes, including financial planning. I started thinking about buying a house. Of course, there are professionals who can help with that, and we have access to those resources. But I enjoy exploring these ideas myself and seeing what I can build.
That's what I mean when I say, just sit down and play.
Part of the reason GPT took off when it became widely available was that the user experience became so approachable. You no longer needed to know how to code. You could simply say, "I want to do this," and it would help you get started.
To me, that's one of the greatest advantages of AI. People have ideas they've wanted to pursue for years but never knew how to begin. Now they can start with a single sentence or a single question and immediately have something tangible to work from.
It's almost like sketching an imaginary character with crayons and then, moments later, seeing a fully realized version come to life.
Dr. Matthew Strehlow: I actually use one AI tool to build prompts for another because I struggle with prompt writing. I'll have one model generate the prompt, then go back and refine it. I've found that approach really helpful.
We've talked about AI helping with things like financial planning and diagnosis. Naturally, a lot of people wonder, Does this mean I'm eventually not going to have a job? Will my role look fundamentally different in five or ten years because AI can do parts of my job better than I can?
Personally, I don't think we'll have robots intubating patients anytime soon, and I haven't seen any AI system that can diagnose patients as well as a trained emergency physician. But the technology is improving every six months.
Five or ten years from now, what do you think our role as physicians will look like?
Dr. Rana Kabeer: You caught me on an optimistic day. I'm a techno-optimist today.
I think this goes back to our earlier discussion about hype. Let me tell you a story.
I've always liked the distinction between science fiction and science fantasy. Star Trek is science fiction. Star Wars is science fantasy.
Could lightsabers exist someday? Maybe. I don't know.
But Star Trek imagined technologies that eventually became real. Forgive me if I get some of the details wrong—I enjoy Star Trek, but I'm far from the world's leading expert. I still put on the original series after a shift every now and then.
One of the devices they carried was a handheld communicator. Decades later, Motorola built a phone that looked remarkably similar because it was directly inspired by that vision of the future.
Star Trek inspired more than just communicators. Have you ever seen the pressurized injection devices anesthesiologists sometimes use to deliver medication, particularly for children? Instead of a syringe, they use a burst of air to deliver the medication. That idea came from Star Trek.
That's what fascinates me. These imagined technologies inspired people to build real tools.
I think AI has the potential to follow a similar path. We imagine a robot physician, the perfect diagnostician, or the master surgeon. Will we get there? I don't know.
What seems much more realistic is using AI to help us triage patients, improve patient flow, or make an overcrowded emergency department run more efficiently. I keep coming back to the idea of AI as a partner—something that helps us operate more effectively rather than replacing us.
Even though the technology is advancing rapidly, I think there's still a distinction between science fantasy and science fiction. Some ideas remain aspirational, while others are becoming practical realities.
My biggest concern is when these systems become autonomous and remove the human from the loop. I believe there should always be human oversight in how these tools are built, deployed, and used.
At the end of the day, AI is simply another technology. Medicine has always adapted to new technologies that didn't replace physicians but did change how we practice and the skills we expect clinicians to have.
Here's a simple example. You hire far more physicians than I do, but imagine you're choosing between two candidates after bedside ultrasound became standard practice. One physician is proficient in ultrasound—performing FAST exams, bedside echocardiography, and ultrasound-guided IV placement. The other has never learned to use it.
Which physician would you hire? Someone who understands how to use bedside ultrasound—to perform a FAST exam, a bedside echocardiogram, or place an ultrasound-guided IV—or someone who's never learned to use it?
Ultrasound didn't replace physicians. It changed the skill set we expected them to have.
To me, AI is similar. You don't have to be an expert, but I do think there's a future where every physician should understand how to use these tools. If you're building a team, you may naturally prefer someone who knows how to incorporate AI into their practice over someone who doesn't.
Everyone has different strengths, and I completely understand that. But I do think there will be certain technologies—and perhaps even certifications—that become standard expectations as medicine continues to evolve.
Dr. Matthew Strehlow: There's an old saying that radiologists won't disappear—they'll be replaced by radiologists who use AI. I suspect the same will eventually be true in emergency medicine.
Is there anything about what we do as physicians that you think is uniquely human? Something AI won't replace and may actually become more important as technology advances?
Dr. Rana Kabeer: Do you want to come to my graduation speech?
Dr. Matthew Strehlow: No, just give it to me now. You can have the evening off.
Dr. Rana Kabeer: To me, medicine is something deeply human. It's rooted in our connection with other people.
We've all read books by physicians like Atul Gawande and others who write about the importance of the physical exam—the physician's hands, the act of examining a patient, feeling for a gallbladder or another physical finding, and truly being present with someone.
There's a reason we have entire fields of medicine, like psychiatry, that rely less on laboratory testing and more on communication. At their core, they're about one person connecting with another.
I think about that every day in the emergency department.
How many patients have you seen who come in with a problem, and medically, you haven't really changed anything? You order tests, maybe a few scans, and five hours later they're leaving with the same condition they arrived with—but they feel better.
Why?
Because someone listened to them. Someone sat down with them, heard their story, and was present with them in that moment.
As the world becomes busier and, in many ways, more disconnected, we talk more about loneliness, screen time, and the loss of genuine human interaction.
I think medicine will always offer something different. Human connection is ultimately what people are seeking. Maybe I'm saying that because I woke up on the optimistic side of the bed today, but I truly believe we're all looking for meaningful connection with one another.
Dr. Matthew Strehlow: I agree with you, but let me challenge your techno-optimism.
Researchers have been comparing responses written by AI with responses written by humans. They ask people which response feels more empathetic, which they'd rather receive, or which would better support their mental health. In many of those studies, participants actually rate the AI responses more highly.
Some people take that to mean we're not as unique as we think—that human-to-human communication can be replicated.
What do you make of that?
Dr. Rana Kabeer: It reminds me of a study about a young chimp that was raised in isolation without its mother. Researchers gave it a surrogate with a maternal appearance, and it immediately formed an attachment.
To me, that gets at something fundamental: we're wired for connection.
Maybe that's part of why people become attached to their phones or constantly seek information that validates their beliefs or experiences. We're looking for connection, even if we're not always finding it in the healthiest ways.
That doesn't mean human connection isn't special. I still believe people can recognize the genuine article. Even in a world where it's becoming harder to distinguish what's real from what's artificial, I think people ultimately know the difference. That's what we'll continue to seek.
Let me turn the question around. What if we think about this as two companies competing to make the same product: genuine human connection?
Competition usually makes both companies better. If AI pushes us to become more thoughtful, more present, and more empathetic, then that's a positive outcome.
It may sound strange to think we could learn empathy from an AI system, but if those tools ultimately help real people become better listeners and more compassionate clinicians, that's a future I'd be happy to see.
I don't believe AI is going to replace us. I think there will be a period of uncertainty as we learn how to use these tools effectively, but ultimately they should augment what we do, not replace it.
Dr. Matthew Strehlow: I really like that perspective. AI has the potential to push us in a positive direction, and I don't think that's limited to individual clinicians.
Our healthcare system—and emergency care in particular—often makes it difficult to focus on human connection. As AI and other technologies continue to evolve, we'll need to think not only about our own behaviors and who we hire, but also about the systems we're placing those clinicians into. If we redesign those systems thoughtfully, I think we can create environments that allow people to do their best work.
That brings me to education. We'll be hiring these physicians in the near future—and in my case, we're already doing it.
You care deeply about education and you're one of the best bedside educators I've worked with. What do we need to teach differently so the next generation of clinicians is prepared for this future?
Dr. Rana Kabeer: For me, the answer starts with literacy.
I've compared AI to learning a language because I think that's exactly what it is. If we're going to teach residents, students, and new physicians in an environment where these tools already exist, then the educators themselves need to understand how they work.
Today, as we're graduating one class of residents, we're also welcoming a new class of interns. If those learners enter a system where AI is becoming part of everyday practice, but the people teaching them don't know how to use it or navigate it, how are they supposed to learn what's appropriate? How will they know what's reliable, what's not, and how to use these tools responsibly?
For those of us in academic medicine and other teaching roles, I think it's our responsibility to invest the time to become literate in these technologies ourselves.
We're fortunate at Stanford because our learners are exposed to AI in a way that's still uncommon. I often tell our residents, "You may feel like you don't know much about AI, but you probably know more than 99% of trainees around the country simply because you're having these conversations and attending these lectures."
We offer sessions for residents and faculty alike, and a big part of that is making sure everyone shares the same vocabulary. That's why I believe literacy is the first priority.
The second is appraisal.
Medicine has always been about critically evaluating information. When we read a study, we naturally ask where it was published, which journal accepted it, and whether the source is credible. We rely, at least in part, on reputation and experience to judge the quality of evidence.
We do the same thing clinically. Every day we review care that was delivered elsewhere and ask ourselves whether we agree with the assessment or whether we'd approach the case differently.
AI shouldn't change that mindset.
We have to learn how to critically evaluate AI-generated information. Where did it come from? Is it supported by evidence? Does it make sense? We can't assume something is correct simply because it's well written or convincingly presented.
There's a concept called citation theater—information that appears authoritative because it's polished, formatted well, and includes references, even if the underlying content isn't reliable.
If someone told you a patient's leg swelling was caused because "the four chambers of the liver aren't working properly," you'd immediately stop and ask, "Where did that come from?" You'd know something wasn't right.
We need to apply that same level of skepticism to AI-generated content.
Dr. Matthew Strehlow: It's been a while since medical school, but I think I'd catch that one.
Dr. Rana Kabeer: Exactly. You immediately recognized that something wasn't right because you appraised the information. That's why we'll always need a strong foundation in the basics—physiology, anatomy, and clinical reasoning.
I'll give you an example.
I recently cared for a patient with a suspected intra-abdominal bleed related to a prior vascular stent. The team looked up the workup using a large language model, which recommended a particular imaging study. They ordered it.
I asked them to show me the prompt and the AI's response. It looked great. The explanation was polished, the citations were convincing, and everything seemed well supported.
But the recommended study wasn't the right one. It wasn't an angiographic phase study, which is what the patient actually needed.
The AI wasn't enough to catch that mistake. I didn't need a large language model—I just needed to open Netter's Atlas of Human Anatomy and review the anatomy. The underlying physiology explained why angiographic imaging was necessary.
That's why I think these tools are incredibly valuable, but they don't replace a strong understanding of basic medicine. They build on it.
Dr. Matthew Strehlow: There are two important points there.
The first is that we need AI models designed for the high-risk environments we work in—models that minimize hallucinations and are reliable enough for clinical care.
The second is education. It's difficult to maintain the discipline required to learn information that an AI can produce almost instantly and that may be correct most of the time.
That means the responsibility falls on us as educators. We need to design educational systems that ensure trainees still master those foundational concepts rather than leaving it up to individual motivation—especially when they're working long hours and learning under constant time pressure.
That's a lot to ask of individual learners, even when they have the best intentions. I think we need to challenge educational leaders, program directors, and our graduate medical education offices to think differently.
How do we design training programs that build the structures needed for people to develop true expertise? It's not enough to tell learners, "Don't use an LLM," or "Work through the differential before asking AI." I'm not sure that's realistic without intentional educational design.
Dr. Rana Kabeer: I agree. Right now, we're in the Wild West.
Over time, the pendulum will swing, and we'll figure out how to use these tools more effectively. Five or ten years from now, I think we'll be in a very different place.
I think about the transition to the internet. I grew up in a world before it was part of everyday life, and then suddenly it was everywhere. We all remember the advice: Don't believe everything you read on the internet. You don't always know where the information came from.
Those conversations were everywhere, and eventually we learned how to navigate that environment. Today, we sometimes trust information simply because it's polished and presented well, but we've adapted before, and I think we'll adapt again.
Dr. Matthew Strehlow: I already see that with younger generations. They seem much more comfortable judging what they trust and what they don't. It's almost as though they've developed an instinct from growing up immersed in digital information. They form an initial impression quickly and then layer additional information on top of it in a way that's more intuitive than it is for many of us who didn't grow up in that environment.
You're involved in so many different projects, and you clearly enjoy exploring new ideas. What's something you're especially excited about working on over the next one to three years?
Dr. Rana Kabeer: It's interesting because I'm most excited about improving things we don't usually think about. In emergency medicine, there are so many processes we've been doing the same way for years.
I think about resident evaluations, feedback, and how we assess learner growth over time. Can we build better dashboards? Can we track progress more effectively? Wherever possible, I'd like to reduce unnecessary subjectivity and create more meaningful objective measures.
That doesn't mean removing the art of medicine. Medicine—and life—will always involve human judgment. But I do think we can improve the way we evaluate and support learners.
One thing I tell every resident is that medicine is fundamentally an apprenticeship.
Surgery may be the classic example, where trainees learn directly from an attending in the operating room. But emergency medicine is unique because, as a resident, you're almost always working alongside someone more experienced. There's nearly always another physician available to supervise, teach, and help you think through difficult decisions.
My wife is a surgery resident here at Stanford, and she's graduating this year.
Dr. Matthew Strehlow: That's a big milestone.
Dr. Rana Kabeer: It is. Honestly, that's what I'm most excited about right now.
Watching her training has reminded me how different our specialties are. In the operating room she's always working with an attending. But when she's on the floor, she's often making decisions independently—answering consults and managing patients before discussing them with the team.
In emergency medicine, we rarely practice that way. For the most part, there's always another physician nearby whom I can turn to for advice or guidance.
At the same time, every evaluation is still shaped by an individual perspective. We naturally give more weight to some opinions than others based on experience, specialty, or the setting in which someone works.
The projects I'm most excited about are the ones that help us standardize those assessments. Can we better understand the experiences each resident is having? Are some residents seeing certain types of cases far more often than others? If so, how do we identify those differences and make training more equitable?
This is where AI can be incredibly valuable. We have enormous amounts of educational data that we've never really been able to use effectively.
Imagine looking at four years of residency data and discovering that, Matt, you're exceptionally strong in STEMI care.
Dr. Matthew Strehlow: I'd like to think so.
Dr. Rana Kabeer: Then we ask why. Maybe the answer isn't simply that you're naturally better at it. Maybe you've seen 25% more acute coronary syndrome and STEMI cases than your peers. Greater exposure leads to greater expertise.
That's useful information. But the more important question is: what do we do for the residents who haven't had those experiences?
Now we know exactly where to focus their learning. We can create targeted educational opportunities so every resident has the chance to develop the same level of competence.
Those are the kinds of projects that excite me. I don't know if they're flashy or particularly glamorous, but I think they're meaningful.
Dr. Matthew Strehlow: It sounds like precision emergency medicine applied to education.
Dr. Rana Kabeer: Exactly. Precision education.
Dr. Matthew Strehlow: We like to end each episode on a lighter note. You mentioned that your wife is finishing her surgical residency. How are you going to celebrate? What's next for the two of you?
Dr. Rana Kabeer: Honestly, I think sleep is high on the list.
We met as residents. I was the day resident on the trauma service, and I signed out to her on nights. That was seven years ago.
Dr. Matthew Strehlow: Good thing you were on the day shift.
Dr. Rana Kabeer: Exactly. I got lucky.
When you're both residents, that schedule just feels normal because you're both so busy. But once you become a fellow or an attending, you suddenly realize you have time again, and you start thinking, we can actually take a trip and spend time together.
That's what I'm looking forward to most. I'm excited to travel, spend time together, and enjoy life without the constant stress of being sleep-deprived or on call.
Dr. Matthew Strehlow: A relaxing beach vacation sounds like the perfect plan.
Rana, thank you again for joining us and sharing your perspective.
If you enjoyed today's episode, don't forget to subscribe, leave a review, and share it with a colleague. We'd also love to hear from you, so send us your questions, ideas, or feedback using the link in the description.
Thanks for listening. We'll see you next time. Until then, keep taking care of anyone, anything, at any time.