Can Medicine Survive Without AI?

As a futurist, I expect AI to replace me in about 10 years. >> What's AI really going to change in medicine? >> It's not a choice. It's a must. We will not be able to keep on providing healthcare. We are living in the Dr. AI era. Physicians used to be the key holders to the ivory tower of medicine. But there is no tower anymore. There is no gate. There is no key. Patients and physicians will keep on using tools which are not regulated, validated, safe and you know simply proved yet. But they seem to be very efficient. What I'm curious about is what am I missing? And it always nine out of 10 times it comes up with something I literally missed.

It's not that I cannot work without AI anymore. I don't want to work without AI [music] anymore because what if I miss something huge that I was not aware of. We have a direct impact on the future. >> The next show >> today Dr. Meshko I want to ask an easy question first. What's AI really going to change in medicine? >> That's an easy question. All right. Okay. That's you might think it's an easy question, but it is not. Usually the simplest questions are the hardest and the most complex to answer. First of all, thank you so much for having me on the show. I was looking forward to it. So, yeah, we are I I can say we are diving right into the middle of it.

um because it's it's a bigger question than one might think because it's not just that of course repetitive and database tasks might be prone to automation and then that's the role AI could play in the future of medicine or healthcare that's not about that it's I think the issue behind or the reason behind AI coming to the stage is much much bigger and I think it's based on a simple math mathematical problem that there are about 6 million healthcare workers missing worldwide today according to the world health organization and they estimate that This number would be about 11 million by 2030. So simply we will never train as many healthare professionals as we need while the number of patients requiring medical health will keep on increasing because we are getting better at diagnosing them monitoring for them and treating them on the long term.

Just imagine you know we had to treat the diabetes patients for like 10 15 years before and now we treat them for 30 40 50 years uh in the best possible cases. So there is a mathematical gap between these two numbers that I think cannot be filled in unless we reach out to quite an advanced breakthrough technology. That was the reason I think be behind AI coming to the stage and then there is a great technological progress in the field that made it possible to be used in a range of medical issues and for medical purposes. And I'm not even talking about yet how generative AI and large language models came to the picture in 2022 2023.

So if I to give you a brief answer, I would say that every single repetitive and database task will eventually have to get replaced because we will we will not have enough healthare professionals to take care of those and still about 1 billion patients worldwide today has zero access to healthcare. So that gap will even get bigger by the time uh we give them access to healthcare services worldwide without even geographical limitations. And what we've seen actually on the field that how what kind of AI technologies have been becoming regulated cleared and accessible mostly in the US because the US regulator the food and drug administration I think is the most like forward-looking regulatory body is that we see a range of different tasks being prone to automation.

One huge group is in radiology, images, analyzing videos, um, radiology, ultrasound images, CT scans, X-rays. I think the the biggest and best examples have been coming out of that medical specialty. And then we see a bunch of others in that in that from that perspective, smaller specialties like oncology, cardiology, dermatology, opthalmology coming out in which we find very special use cases of how AI could be utilized in the healthcare field. But in general usually even from and that's my last part of my answer even from press announcements or studies and articles I always feel that those talking about what role AI could play in healthcare they imply that we have a choice that we human beings who are like nerds and geeks like myself we choose to have a doctor patient discussion with an AI and not with a human being.

But that couldn't be further from the case. It's not a choice. It's a must that because of that basic mathematical gap unless we reach out to AI and we learn how to implement it into a very sensible cultural relationship between doctors and patients. Simply I think the the easiest thing to say is that we will not be able to keep on providing healthcare. So that's that's the the simple answer for your simple question about why AI has come to the stage. Yeah, I mean I knew that would be a very uh complicated answer. The it's such a big topic and that's an interesting framing that we really have no choice here as a society but to lean into these tools because of the gap in the number of health care workers.

How are things changing right now? And then we can jump into a little bit where do you think they are headed down the line. Okay. I think right now especially in the last like three four years because of the rise of generative AI and large language models we are in a constant state of like shock. Alvin Toler futurist called it the future shock back in the 1980s. I think the vast majority of physicians are in a constant state of future shock because there are technologies coming to their practices to their daily jobs that are not like you know something a new gadget they have to learn how to use or a new electronic medical record software something like that but it's a vastly different kind of creature if I may say so and the billion dollar question is will AI replace physicians and that's I think a general fear among healthare professionals word right?

That if this thing can do so many things so efficiently and so fast and the performance can even be comparable to the performance that we put on the table as medical professionals then what is my future? Will AI replace me in my job? Will it replace medical specialties and all these? And these are big and important questions. I think where we are now is that there are maybe more than a thousand um regulatory approved real life use cases for very specific medical purposes where these tools are being used in practice like there is a there are there's I think at least two dozen tools for analyzing CT scans in the case of stroke to to um provide the diagnosis as soon as possible.

There are AI based tools to um being used as AI scribes to listen into the conversation between patients and physicians and transform that conversation into the exact electronic medical record format the given hospital or institution is using. There are very specific use cases and there are about 1,400ish uh FD cleared AI based medical technologies out there. We at the medical futurist we were the first ones to publish a study back in 2020 uh after we we had to dive into the database the quite huge database of the FDA looking for these AI based technologies because back then the FDA didn't even have like a different process for approving AI based or other types of digital and technological products and we found 64 imagine 64 in 2020 almost 1500 right now and you can imagine that it will pass the threshold of 2,000 very soon.

So it's not an exponential kind of uh increase but it but the increase and the progress has been extraordinary. So I think what's happening now is more like um a cultural transformation as part of that future shock that every stakeholder not just physicians but nurses patients government officials policy makers even the researchers and developers themselves who make and develop those technologies all of us have to learn to find our new ways in this new status quo because already before the AI revolution really you know jumped into our faces in the last couple of years there was a digital health transformation going on from the beginning of the 21st century very briefly.

Meaning that gradually with all the influx of advanced technologies from variable sensors to smartphone apps and all these patients started becoming the point of care and a big conclusion of that is that patients are also becoming members of their medical teams. And it was quite a new again cultural shock for all of us to learn to use the the most underused resources of healthcare patients themselves in the ivory tower of medicine which has been breaking down with physicians who now has to answer questions that patients bring from search engines and websites and social media channels and now what's happening now is that the this this AI entity is sitting at that same table.

So while experiencing that doctor patient culture shock now all of them have to start learn new ways about welcoming a a technological creature in this very cautious and sensitive kind of relationship that's based on trust empathy and compassion so all of us are learning along the way medical associations are coming up with guidelines about using AI but you know never as fast as uh they should have done it before some healthcare governments have been publishing the right regul ations and and laws and policies to ensure that we can implement this technology safely and safely and efficiently enough but every stakeholder is just have a learning path or learning curve and I think we struggle more than how much we enjoy or benefit from using this technology today but of course the the long-term goal is to change that balance and benefit as much as possible from the technology without losing control without harming patients safety or or efficiency of predisy medicine.

What does that mean for the doctor patient relationship now that that patient might be showing up with all this personal information from the wearables and technology that they might be using on a regular basis and then when they show up and they have chatted through some of their problems with an LLM and maybe have some insights into their own care that normally that or I should say in the past that patient never would have had that information before, never would have had those insights before because there are many examples of patients who have uploaded their labs into their LLM of choice and asked it to do an analysis and then that has turned out to be helpful for their care.

So how does that change that doctor patient relationship? I think the biggest change here is not just accessing studies or information or insights. I I like as a futurist I always look at the evolution of how something changes or progresses regarding patients that the evolution's first part was about I call it the doctor neighbor era before the 20 21st century when patients essentially went home after getting a prescription and they discussed their issues their medications their diagnosis with people they knew maybe their neighbors that was more about sharing their experience and authority with others and then came the 21st century and that's the era of Dr.

Google when patients started accessing the insights that you mentioned the same studies the same books and information websites social media channels second opinion peer support and experience from other patients worldwide that was the era about accessing finally the same information within the without the ivory tower of medicine only physicians used to be able to access what's different now is that we are living in the doctor AI era when patients ask very specific questions and get better at prompt engineering to to benefit benefit from using large language models. So what they can access now is not just the insights but clinical reasoning.

That's unprecedented. It's, you know, it's very different from reading something on a website and a medical website and then bringing that piece of information to the next doctor patient meeting compared to getting questions I should specifically ask my physician at that meeting because of based on the medical records I shared and the data from my variable sensors and all these and my symptoms and the experience and the threads we've been having uh with that large language model, it will come up with insights about reasoning and that might really surprise if not frighten and medical professionals themselves.

So I think what's happening or what's changing in that relationship is that there are two sides of that story to that story. One side or one perspective is coming from physicians that they are the authorities. There is a reason why someone studies for years if not decades to become an expert of a medical specialty, a therapeutic area, a sub field of medicine. So they want to keep that authority because they have it for a reason. But when patients can access the same insights and reasoning tools, then they have to not fight but compete with the AI's questions and answers and all that and it it creates like a new I think communicational gap between patients and physicians.

From the other perspective, patient finally enjoys getting all that information because you can have a conversation with a large language model 50 times a day, not just once every month or so when you actually meet a healthare professional. And with Dr. shortages. I think it's going to be a luxury to be able to meet with a physician with every health problem we might have. So they might enjoy that, but it's very easy to get lost and feel doomed in the jungle of data and information. So again being very brief about the answer. I'm sorry about that. the roles are changing and that's the biggest biggest shift that the patient used to be a passive stakeholder you know waiting for a symptom and only after that asking for medical help and now they are more like so-called empowered patients they are equipped with technologies they are enabled they are engaged in their care they want to be equal with their healthare professionals physicians used to be the key holders to the ivory tower of medicine but there is no tower anymore there is no gate there is no key so they are becoming more like guides for their patients in the jungle of data and information.

So that's I think the biggest change or shift in that relationship. >> Yeah. And a lot of it feels really positive to me as someone who wants to know about my own personal health. And I'm not facing a medical issue right now, which is almost a different topic altogether, but just in in general wanting to know more about what's what's going on and how do I stay healthy and be proactive in medicine rather than just reactive to problems. And an example of that recently is that I know I have friends who had their testosterone tested and found that they were low on testosterone. Uh started supplementing and it's been very positive in their lives.

And I don't feel that I am currently in in need of supplementation on that level. Um, but somebody mentioned the other day, they said, "You know what is great to have though is to get the baseline right now. You're in your 40s and then you'll know if there's a change in the future. And then you're not just comparing yourself to a broad metric of your age range. You're actually comparing yourself to your your previous self, so you'd really know if there was a drop. and the uh medical insurance provider that I have um they were really reluctant to to run a test like that. That was just they said, "Well, if you don't have any symptoms that there's a problem that is just something extra.

We don't really recommend that." I still wanted this information for myself though. And then someone mentioned, hey, they said there's an app um it's called HIMS. They said it's $90 and they will send you a home test kit. You can do it and then find out where you're at. And the the the kit is on its way right now. So, I haven't actually taken the the test yet. Um, but I thought, how fantastic is that that I want this information for myself. This will only be good for my self-nowledge and long-term health. And that's now available to me. and I don't even have to go into a medical facility to get it done, but I can still get accurate results with that.

You sound to be very involved in your health management and you can imagine I do the same. You know, I'm I'm I would I dedicate so much effort, time, money, privacy into trying having a chance just to live to 100 and board a spaceship to Mars tourist then. So, of course, it's great news for us, for people like us, but the to be to be object objective here um and impartial for the vast majority of patients and physicians, I think it's a pain. I think it's a threat. It sounds like a danger and it sounds awfully complicated because actually, technically, it is. So, for the vast majority of people, it's a it has been a struggle and a big challenge to learn, you know, their own ways through it, to learn how to use technologies in health and disease management.

Otherwise I'm not even sure what you can do uh properly not without you know getting immediate feedback about how you perform but whether you make good decisions or not on the other side you find physicians who now have patients coming with such data from variables from at home lab test and genetic services and all these so I think for the vast majority it's not so enjoyable as it seems to be for people like us just just I wanted to put this on the table but what you mentioned is is crucial because that's one of the major I think consequ quences of the digital health paradigm shift of healthcare currently is that healthcare is gradually becoming like globalized I guess that's the right word to use here meaning that and it might be a bit controversial but meaning that um what I mean by that is that that someone's access individually to services at home lab tests technologies AI based services and all these I think will have a bigger impact on their lifespan and health span than which country's healthcare system they belong to.

And it might be a very bold statement, but that's what that's how digital health has been simply, you know, re reducing and removing all the geographical boundaries because then you could say I I'm based in Budapest, Hungary, but my healthcare, my quality of of health care doesn't depend on my country's healthcare system alone. My genetic test was done in Italy. The at home lab test I use for lab markers I think is was based in the United Kingdom. I I'm using so many services from the United States and then you know I can go on my ser my health care service is globalized enough because as an empowered patient I want to take active part in that I want to lead my health or disease management and I'm happy to devote attention time money all these things resources to that but that's not common that's not so widespread and as long as we cannot solve these major challenges for people like you and I how can we expect that that digital health and AI paradigm shift to work for billions of patients and physicians even though they might not even be on the same page where we are.

So first we have to make it happen for those who are ready to get engaged and and take active part and play an active role in their care or on the other side be open to such patients and become their guides in this jungle of data and information and only I think after that we can start helping the other billions of patients uh with the same goal but all purpose but I think still the discussion we are having is different from having the same discussion five years ago when AI was still a niche tool. maybe a few maybe hundreds of thousands of people could use an actual artificial intelligence based technology you know mostly developers and researchers around the world but now about one and a half billion people have used the large language model and according to data coming from um code and um Gemini and CHP about 60 70% of them have used those models for health related purposes so imagine one billion people just having access to a healthbased AI out of the blue when it used to be a very much niche tool for just a few people.

That that's the kind of major shape you're experiencing now. >> Yeah. Well, clearly it's on a a radical uptake. I mean, you've got a uh 100,000 subscriber plaque there in your background and doing a very niche topic or at least it previously was, which is all this personalized medicine and now people are really there's there's a large group of people really tuning into that and I think it drives at a need in the medical system which is that so much of western medicine as I understand it from Dr. Friends is based on a reactive model to a problem arises, how do we address the problem? And then you've got people that want to be proactive about it and go, well, I'd rather not have the heart attack in the first place.

So, what do I need to do to monitor that so I don't get to that place and have to react to it because that has long-term consequences just having that big medical uh negative effect in the interventions and that's a real shift in how people are personally uh moving into medicine on the on the doctor hospital side of it. What are some of the issues you see in the uptake of these new tools from the systems point of view? >> Absolutely. Just a comment to your your first part of your question. Uh I think the system of medicine reacted to that by introducing lifestyle medicine as a new medical specialty which is basically primary care but focusing on prevention and health management rather than treating a disease or providing disease monitoring for the long term.

And I have to defend the system of healthcare in general because for centuries if not millennia the idea was that if you are sick come to us and we will help you as much as we can and we cannot change such a you know system with a long history and background in a few years because that's the kind of uh attention uh the period we are talking about now. In a few years it should be you know upside down and everything should focus on prevention not sick care. We are moving towards that but it cannot be changed overnight. Now about the uptake of those technologies there is I I I think I'm I can say I'm lucky enough to witness quite an evolution there because I've been going to give speeches at conferences for medical events and found my events at for about 15 years.

So 15 years ago, the idea was that yeah, there might be some variables on the horizon and maybe a few people will use smartphone apps for health purposes and you know maybe a few dedicated hardware kinds of products could come to the market but that's it and that's not for the masses and literally back then maybe a million people or so used a variable sensor. usually, you know, geeky, nerdy people like us who wanted to measure something about their health and they wanted to get bio feedback so they can make better decisions. But it was a really niche topic. But today and we checked that the medical futurist published an infographic about that.

About 1 billion people today have a variable sensor that can measure at least one vital sign or health parameter. So that's not for the not for the niche populations anymore. It's for the masses. And of course, it's becoming even bigger than that and billions of people. we measure something otherwise how can you make a good decision I mean I absolutely respect and understand those patients decisions who are rejecting the idea of using such devices to measure something about their health I I understand that but also have to acknowledge and understand that then you are not benefiting from these technology if it's all right for you then we should be off we should be all right but you have to know that you are not benefiting from these sometimes evidence-based technologies so the reason why there was a rejection a general rejection And I think a general fear from medical professionals about many of these technologies even before the AI era which with the rise of um large language models I think just made the whole situation worse.

But there is one like catch here evidence-based medicine works in a relatively simple way. If you or your company or you yourself come up with an invention, it can be a biomedical product, a drug, a treatment, a diagnostic procedure, a variable sensor, an at home lab test or an AI based service, whatever you come up with. If you can prove through peer-reviewed studies and clinical trials that your product is as efficient and safe as anything else on the market before it or if there was nothing before it on the similar kind then and it's efficient and safe then physicians will be able to prescribe that and use that while treating or caring for patients.

That's how simple it is. Even if um if OpenAI comes up with the world's best healthcare invention ever today that could save millions of lives, but they don't have the proof yet through those studies and clinical trials, simply nobody is allowed to use it in medicine. And as long as evidence-based medicine applies, and I think there's a very big reason why we keep on sticking to that rule because that's how medicine works and that's how we can keep patients safe. then um I think physicians will have to gradually understand and accept that there are some new technologies I have to use because those have just become evidence-based enough.

So it's not about whether my opinion matters as a physician. If it's evidence-based as a medication or a variable sensor on AI based large language model, I will have to learn to use it because evidence-based is part of medicine. >> This episode of the Nick Stanley Show is brought to you by Zapier. If you've ever felt buried in repetitive work, copying data, moving files, sending follow-ups, you know it's like death by a thousand mouse clicks. Zapier has always been the tool that fixes that. It connects over [music] 8,000 apps. Google Drive, Slack, Notion, Gmail, MySpace, you name it. So your tools can finally play nice together.

But here's the big shift. Zapier now lets you create AI agents with their chat GPT integration. Think of them as tireless teammates who never complain, never take lunch, and never get bored of doing the boring stuff. I've actually made several of these little agents myself. No Python, no JavaScript, just pure vibe coding, which is my favorite kind of coding because even though I have no idea how to code, I just talk to the AI, tell it what I want, and almost magically works. For example, when a podcast guest books a time, Zapier can send them a personalized email from me preparing them for the show, create a draft of show notes, update my calendar, and even prep a social media post automatically.

All of which frees me up to focus on the important stuff, like taking credit for the amazing work my agents did. And now, Zapier just took it up another level for developers. Zapier MCP is available right inside Chat GPT, which means you can connect those 8,000 plus apps and trigger workflows just [music] by writing what you want to happen. Literally, you tell ChatGpt, "Send this file to my team in Slack, update the spreadsheet, and draft an email, and Zapier plus ChatgPT figure out the right tools do it for you." Getting started is simple. Head to the Zapier Chat GPT MCP server and add the tools you want.

ChatGPT to access. Follow the steps in the connect tab. And if you're an admin on a chat GPT enterprise account, you can enable MCP across your entire workspace. So if you're ready to stop wasting time on busy work, join the AI revolution and make a little automation magic of your own. Try the Zapier Chat GPT integration using the link below. Now you're in the business of predicting what's going to happen next, right? The medical futurist. It's never easy to predict anything in the future. And I'm interested in what are some of the things that you thought would happen and then turned out not to or surprises along the way because you clearly are ingesting as much information as you possibly can to make good predictions about the future but it is an impossible task to accurately get everything right.

So what are some of those surprises? uh I'm happy to mention some even some success stories if you don't mind just to give a you know objective impartial image here and then if you don't mind I will try to destroy uh predictions in themselves because they have no no place I think it makes no sense to have predictions like that but 15 years agoish I started talking about empowered patients and the ivory tower of medicine breaking down um the rise of the variable revolution patients becoming the point of um healthcare becoming globalized. But the things I missed I think was the the reasons for those and I will tell you real life examples about those were more about expecting the public to adopt them faster which simply didn't h or haven't happened yet.

One is definitely uh 3D printed cast or actually 3D printing in general. It it seemed like five 10 years ago it it seemed like a no-brainer that we can print out customized splints, casts, parts of medical equipment um for the you know the fraction of a cost compared to how factories have been making them. I had a 3D printed cast before. I could have a shower by having that on my arm. It was customized to my arm and it was a vastly different experience from having a plaster cast on. But you might find I think literally two three places in the world where they can actually print out 3D printed casts today.

It has not become routine part of traumatology practice or or think about 3D printed drugs, medications in customized dosages, even customized forms um for maybe kids who didn't want to take their medications but they would take it if it was in a very customized form. There are two or three existing examples worldwide in the US and a few in China when the regulators approved such drugs but it's not part of routine care because of how expensive it is and how maybe it can be used in very specialized rare conditions rare diseases but not as part of you know pharmaceutical manufacturing. So I was definitely wrong about the hype about those regarding AI.

I think my hype was not hyped enough or not high enough uh because of the rise of large language models. when CHP came out then I I uh started working with professor Eric Toppo one of the biggest minds in the digital health and AI field and we published a paper in nature's NPJ digital medicine uh journal about regulating the future versions of large language model not just the ones that can deal with text but the ones that can that deal with sound image video even full publications or or full documents so then when these tools would become accessible and they will or they have screen some of those for image for example then the regulators would know what to do and there would be no you know lag of years between when a technology becomes physically accessible and when regulations would allow that to come to the market when it was on the market already that's where I think the the power of prediction lies but it makes no sense to make predictions and literally my full-time job for the last 15 years has been making predictions and being as efficient as possible uh about those but I still come to the conclusion that it makes no sense to make a prediction.

And I describe briefly why. If I make 10 predictions and nine out of those 10 would actually turn out to be right, which is quite a good track record, right? Then how about my 11th? What what how would my 11th prediction inform your decision- making today if I made a mistake once? And of course, everyone will make a mistake. And I think what predictions do is that they imply that there is one future, one singular timeline. And as a professional futurist, I know it better than you do what exactly is going to happen and when. But that couldn't be further from the truth. The truth is that there are multiple futures ahead of us.

And what futures methods can help do which we futurist use as part of the daily jobs is to analyze, observe and discover those futures to create scenarios and visions and forecast certain probabilities of the outcomes of events and create empathy for future characters or people in the future. What we can do in general is to analyze those futures, maybe choose one we like the most and then backcast towards today. Well, if this is our desired future out of the many possibilities, here are the steps we can start doing from today. Legal, technological, cultural changes we can start implementing to have a higher chance to reach that desired future.

I think that's what can inform decision making, not blind predictions. I still make predictions as part of my job, but I I every single time I have a sort of disclaimer saying that these are not predictions that I think this or that is going to happen, but these are like pinpointing or highlighting trends and directions and cultural shifts or technological progress. For example, I'm much more interested in the evolution of certain technologies. I analyzed the evolution of blood pressure cuffs, ECGs, sleep trackers, fitness trackers, steoscopes because the evolution giving people the context about how these technologies change and shape over time.

I think it's it's more impactful than telling them that the ECG will be this big only and it will be thin and a analyzed in five years because what if not? So, I'm I'm much more interested in the business of giving people the reasons, the context behind all these changes than telling them what I think is going to happen, which should be more relevant than what they think because I'm a professional futurist. So, it might sound a bit controversial, but that's how I've been fighting against this idea that the future is predictable because and I I always play this game. At every single event I get a question, what is the future of oncology or the future of a therapeutic area?

I always get that and I always ask back tell me which future and I will predict it for you. >> Right? >> You know which future? The future of oncology where every single oncologist and patient are on the same page about the progress of technologies or the future where you reject the the implementation of those technologies. So tell me which future and then I will predict it for you. Of course a bit jokingly saying that. So all in all I think the best we could do all of us in healthcare is to just acknowledge that multiple futures exist. There are very simple methods to learn to use called futures methods that help us analyze those potential futures.

And even more importantly if we do that then we we actually come to the conclusion that we have a direct impact on those futures. If I choose to be, you know, open-minded, if I choose to have a good mindset about this, then we will come to the desired future simply faster, then we would have been able to if we were not on the same page about this. So, we have a direct impact on our future. >> That's an interesting frame. Again, it reminds me I've had several conversations with people at the uh machine intelligence research institute and they talk about predictions in a similar way because they are concerned about some of the negative impacts AI could have for all of society.

So then they get cast as as doomers or that they're anti- AI. And when you actually talk to them, they're like, "No, no, no, no. I'm not negative on these technologies. I use them every day. These are power users of these technologies. They just said there are different paths that the future could take. Much like you just said, where we want to identify what those negative effects could be down in the future and then take steps today to not go down that negative path, to go down a positive path where these emerging technologies have a huge positive net effect on society. And rather than going down a negative path where we're just moving blindly and just see where the technology goes, what are some of those steps we need to take today then to have the positive medical future that you referenced earlier where you have billions of people across the globe and their health care system within their country isn't necessarily the roadblock to getting worldclass medical care.

Um very simple question to ask very simple answer to provide uh I will try to take the role of you know someone as a futurist I'm my job is to do like have a zoomed out perspective and look at healthcare from this this bird's eye view um because I don't want to tell people what to do but I have ideas about that trends that seem to be like very much high should be very much highlighted and it will affect lives of billions of patients and physicians. One is definitely that we cannot expect physicians to learn to use all kinds of AI based technologies because you have to become almost like a developer to learn to use them and it's almost impossible to to put that knowledge and skill set into medical curricula.

It's more than enough to just teach them the very basic characteristics of AI and what machine learning is and what kind of levels AI might AI might have risks advantages but just a simple minimal package would be enough. Um but just to give you a real life example about that I think every single physician learns how the MRI machine works the magnetic resonance imaging machine in in studies of uh biohysics but without hurting my peers 99% of physicians cannot describe it in technical detail. We studied it back then but you know it's very complicated very complex kind of technology but we understand enough or we have a scientific understanding and a trust large or big enough that we know that we won't harm our patients sending them to that machine.

I think we need a a similar level of understanding and trust into AI based systems. So one of these I think long-term trends or conclusions is that physicians will have to learn prompt engineering and context engineering because they will not learn a range of AI based technologies one by one. It's physically impossible but we can expect them to learn to use one AI technology with which they can have humanlike conversations based on text. So the I think large language models or generative AI will act like the the middleman the ultimate interface between physicians or nurses and nurses and a range of AI based technologies and it will we will be able to use them to translate this or that like there is a uh family history from my patient with a radiology report in this or that format.

I need you to translate that into the exact format of radiology my department is using at my hospital. I cannot do that myself and I cannot ask radiologist to do it for me but the AI can take care of the test quite easily. That's one. The second is that we cannot we when 15 years ago so many physicians at so many medical events kept telling me that we should not allow patients to use search engines and look at websites to find medical information. I kept on telling them that they don't need our permission to do that. Of course they are authorized and engaged and empowered to do whatever they want.

Our job as medical professionals is to keep on providing enough trust and medical knowledge so then they won't be you know um they won't feel doomed in that amount of information. Now we have to do the same but with insights and clinical reasoning coming from AI. We have to somehow provide the kind of patient education through which patients will still turn to us for the ultimate decision but we will encourage them to keep on using all these tools. We will be their guides in the journey of learning to use all these tools because at the end of the day if we do it in a proper way then it will save us time, effort and resources.

So we can spend even more precious time of ours on patients or on develop developing even better technologies that all of us can use for the benefit and sake of patients. But that requires so many new things regulations where the FDA has been like leading the leader example worldwide but others should just follow the lead. But even the FDA cannot deal with adaptive AI you know the kind of AI that will change with every single decision it makes. Maybe you regulate one kind of technology and then it makes a million decisions on the first day. It will change a million times. So the regulating adaptive AI is still a a big challenge.

What do you do? Do you regulate the product and you check every single you know programming language shift in the system or you regulate the company developing the technology or the product and then you want you want evaluate that kind of regulation week by week or year by year. How about medical associations that should provide the right guidelines for the physicians about how to use AI, how to report AI based studies, how to talk to patients about AI. So when a patient comes in and I'm using an AI scribe to save time and not turn to a keyboard to input data, how should I describe it to my patient without losing trust, without not showing enough empathy or compassion?

So we have to solve so many things to to make it happen. My fear is always that um what's what's different now compared to 10 15 years ago is that when 10 years ago a technological breakthrough came out then regulators policy makers all of us had time to learn to use it because there was a a a time gap between when the technology becomes accessible and when it would be allowed to come to the market otherwise how could I access it but now I can access it. uh chit came up with ch health earlier this year. Clo came up with entropy came up with clo health. All these large language models have now health related versions.

Billions started using them on the same day. That's not comparable to how much time regulators used to have to deal with such a technological breakthrough. So what it mean what it means is that patients and physicians will keep on using tools which are not regulated, validated, safe and you know simply proved yet but they seem to be very efficient. So of course they will reach out and that's going to create like an ethical gap that I don't see how we can fill in uh in time. What was that moment like for you personally when you glimpsed the future? Everybody's got one in whatever industry they're in where they are they first started using it was for most of us chat GPT but it could have been any of the models and and you suddenly your eyes are just open to what's possible down the line.

I had two actually and I call these moments the you know the technological sublime when just witnessing one technology makes you hopeful for the future. One was hardwarebased and the second was obviously software based. The hardware ways was when a few years ago I used a mixed reality headset and when I turned it on then um 3D human body appeared in front of me but not you know in the way of an image or an animation being projected like in a cinema but I could dissect that body just the bones or the muscles it had a beating heart I could walk around it so that was that was like a technological sublime event in my life saying that wow that's really like something jumping out of a science fiction movie and a second was was software based and it was not when I started using the first iterations of CHP but when about half a year ago sometimes you know at the end of 2025 I started having really deep scientific discussions with that model about the hypothesis I was working on and I didn't expect anything from it.

I was just curious if I start sharing such things with you dear large language model about the reasoning I have in mind and why I focus on this on doing this study right now and what kind of database I'm building out of this or that started doing that and the feedback and the reasoning all these things turned out to be so extremely useful that it's I started changing my hypothesis plus at some point I felt like I cannot allow myself to think that I shouldn't make a decision without consulting it first. And you know that's awful. That's that means that I'm losing you know my human skill. I might think that I'm losing my human skills.

But it turns out I think after a few months I realized that it's all about how much I want how much re how many resources I want to put into that that conversation that I would never allow AI to make a scientific hypothesis instead of me. But I'm using it in the ideation process like never before like nothing before. and like nobody before and I don't want to to hurt those colleagues I've been collaborating worldwide for more than a decade but the the speed at which I can do ideation and the best part is and I think that's my really technological sublime moment was that I don't want the AI to find out something for me I don't want to it to come up with a solution for me I will take care of that but what I'm curious about is what am I missing and you cannot not know what you don't know.

Um um I think um vice president of the US said that a decade ago or so. You cannot know what you don't know. So what if there is something big I'm missing because I'm not aware of that. So I started asking these questions and reprompting myself back throughout the audi ideation process. Look, this is my hypothesis. I describe the reasons, the details and all the ideas I have in mind. This is my goal. I want to achieve this or that. This is going to be my method. Am I missing something? And it always nine out of 10 times it comes up with something I literally missed. And by including that and I worked out the details myself, but by not missing that, I think I gained a lot as a scientist.

So that's that's quite a you know sublime moment that what if I not it's not that I cannot work without AI anymore. I don't want to work without AI anymore because what if I miss something huge that I was not aware of? That's I think the biggest lesson for me. >> The technologically sublime. It's a great turn of phrase. >> I took part in a um Harvard Extension Extension School course about science fiction, the history of science fiction. And that's where I I learned that expression that many of these works, movies and books have a technological sublime moment when everything becomes so bright and you know the the sun is rising on the horizon.

everything is going to be better and promising and hopeful because of this exact technology that will solve all our problems. It's never the case, but that's the feeling it radiates uh that that one technological solution. >> If you find yourself guessing at your kids college odds, know that it doesn't have to be that way. There's a website. You pick a college, any four-year college in the country, type in your student's GPA and their SAT or ACT score, and you get to see their odds built from that colleg's own published admissions data. Change the score, watch the results change. You can look up any high school in the country as well to see their results at getting kids into the UC's.

It's free. It's called Reach, and you can get there at www. highchoolreach.com. How was the class? I'm I'm curious and the reason I ask is that I feel like it's a fork in the road for higher education as well in terms of staying relevant to teach people skills for the future. And you've got some colleges are really leaning into this in a big way. And some are kind of waiting and uh we'll see what happens. Maybe this is all hype. Maybe this is a fad. and they're just kind of going about business as usual. And I just know from families I work in education myself, they're very excited about schools that are leaning into this.

But I'm curious from your point of view, since you're so deep in the weeds with all of this, how strong was the class itself and how helpful was it to further your understanding of these tools and what's possible? I I had a very specific idea for joining that specific course and it was that I felt like I've read and watched every major science fiction movie and book out there. I've been a science fiction fan since like don't know six or seven. There are books I reread every two years because of I learned so much about them even though through the same stories and the reason why I chose this course was that I think my weak point as the medical futurist is always the the cultural social perspectives of technological change.

I feel like I'm very efficient at the other one tech not I mean not missing technological blind spots and and pinpointing the progress and evolution of technologies. I feel very comfortable in that, but not so much about the same evolution or progress with social or cultural change. And what I always find fascinating about science fiction works is that they they might present you a word or environment full of technologies, but there is always something like a social or cultural issue at the center of that. 2001 of Space Odyssey is not about the rise of AI. It's about trust between technology and human beings.

Mhm. >> And I I wanted to see how the biggest science fiction authors of our time and of past times have dealt with that. Like there is a science fiction author these days called Becky Chambers. I love all her books and she's I think she's the master today at focusing on the social cultural issues, challenges, interactions while creating a technological science vision like environment. So from that perspective, it has been extremely useful to do that course. But man, I do so much more. I I started going to a chess trainer to to become a like a semi-professional chess player because I feel like I needed a a language that's common with AI.

So what I mean by that I I wouldn't be able to learn to enough programming language skills to learn to understand how AI is being written because not even the sometimes the developers of AI understand how they work. So I wanted to have a language through which I can really understand to the absolute details the progress of AI and chess I think is the perfect platform for that because now by understanding you know higher and higher level chess games and trying to play myself a better and better game every day or so I've been doing that for almost eight years now I still have the chess trainer I go to him every uh second week I see where AI is progressing in chess and where it has a plateau phase or sort of that and I feel like I have a common language with it.

So it's not about learning skills or just you know learning about something specific from textbooks. For me it's more about constantly challenging myself when the world is challenging me with all these advanced technologies. Final example last year I had I think the world's most comprehensive longevity package and as a physician and scientist I've been pretty I've been pretty much into longevity for the last 15 years. I've had more more than 100 digital health sensors and eight genetic tests and all these. But last year a company offered me a package and I had 50 plus physical examinations, 300 plus blood markers, genome sequencing, microbiome, everything you can imagine, full body MRI, all the things that healthcare is possibly able to do on a human being.

So in those three months last year, I think I was one of the most examined human beings in uh in the world, including Brian Johnson himself. And and it it it makes you learn a new language and makes you learn new like emotional tricks about your health. You know, all these data of course creates health anxiety, but how to deal with that? I for example I came to a conclusion that I would rather lie in this very loud very closed MRI machine for an hour or so without having any symptom to you know seek the the the solution for just lying there it's very uncomfortable but I have nothing to search for if they find something I want to deal with that but I have no reason to lie into that machine besides just you know this for the sake of prevention so all these new like mindsets and knowledge sets and and skill sets.

I think these are just helping me helping us better rise ride ride the waves of change because there are so many of those too. >> I want to come back to that experience but you mentioned before there that the learning chess at a high level is helping you to speak a common language with AI. I just want to dig into that a little bit deeper because that is really interesting. >> Absolutely. Um I hope we have a few more hours on that. I'm just joking. [laughter] But but chess is not just something I play as a game but something I learn um through many books and through my trainer into understanding the the real deep concepts of it.

Just to give you a real life example for centuries it it seemed that chess is about materials. If I have more materials on the board than you then eventually I will beat you or you have to be so much better than me. So then maybe we get to a draw, but material advantage will simply decide who will win the game. And then AI came to the picture that doesn't even understand the philosophical background of what chess is. It knows the rules of course, but through reinforcement learning by playing millions of games against itself and only keeping the strategy with after the games that it could win, it came to a conclusion that no, no, no, not materials, but space advantage is what g what wins you games.

And you might think that it's it's ridiculous. Every single chess grandmaster today, including Magnus Carson, Gary Kasparov, all of them learned that materials matter and that's the only thing that matters. Of course, positions and strategies must be on the board. But if I have a rook more than, you know, I have a one rook more than you, I will beat you for sure unless you are much much better than me. And then AI comes to the picture. It starts playing games against against other AI based games uh players and it might have a material disadvantage, but by having a space advantage, it kept on beating them.

And now, and I think that's I hope that's the worst thing I can say about AI today. There is a I'm a I'm a very like casual player. Magnus Carson word number one and the the distance knowledge and skillwise between Magnus Carson and myself is smaller than between Magnus Carson and the best AI chess player in the world and it says nothing about me. I would need five lifetimes and a different brain to try to catch Magnus Carson. I have zero chance to do that. And just imagine I am closer to him than the next AI based player which he will never beat. And he knows that exactly. There might be many games ending with a draw because there are positions like that.

But if there is a more than zero 0.00001 advantage, the AI will find a way to exploit that. And that's even Magnus Carson is very fierce on the table. He's known for that. But that human skill cannot even be comparable to what AI can bring to the picture. So by for example seeing that mindset shift from material advantage to space advantage that's something the AI came up with I saw a a jump in AI progress which then materialized in how a large language model started started to provide simply better services for us and even take care of images. If you think back 2023, I had to go to Discord and to use the midjourney channel to create one image and it took a lot of time.

It was a very uncomfortable platform. Now I can upload an image and tell the chip what exactly do with that like translate the the text on the image. I don't have to label the data. I just describe what I want to do. It will translate it. It will change it. It will form it. Unbelievable. And imagine you know what's going to happen with the next iteration. So that's I I like to believe that that these are the kinds of progress jumps that I can predict in AI because I can see how Stockfish and Alpha Zela the best chess players out there the AI chess players keep on making like different moves or different strategies than what we human beings thought would be the best one in chess.

>> Yeah. Yeah. And as that comes back to medicine, something like a strategy shift is where AI could be really helpful in breakthroughs for say cancer treatment where we've been working at it for a long time, but some sort of a shift like no, it's not the material advantage, it's the space advantage in chess. just a slight reframing on some of those medical challenges that have plagued humanity for a long time. I think the hope is there from most people that that little that little shift could speed up our progress in some of these areas in these very persistent difficult medical problems and that we could have more breakthroughs.

That's an interesting way that you're approaching understanding AI uh more thoroughly and that's exactly what's happening in medicine. What you are mentioning now that you might be an extremely up-to-date physician reading a study every week or so that that's amazing. But there are 42 million studies and who knows how many clinical trials and technology announcements and all these. So of course it's physically impossible for you to be up to date. So, but it's not about then it means AI will replace you and that's it. No, no, no. AI will be a tool that you can use to read tens of thousands of papers, not one or two.

It will summarize what you need, but it will it will require your very specific prompt and and context engineering skills to get the knowledge out to exploit the the potential of the technology for your benefit. But for that it needs your proactivity to learn how to use it to learn these skills and to shape your mindset. But if you do that then you will become better. You will have more time. You will spend more attention dedicate more attention to your patients or the technologies you want to keep on developing. It's more it's all about your proactivity. And I understand that not everyone likes that kind of challenge because it requires them to change.

And that's why I mentioned the term, you know, riding the waves of change because I I I I like comfort, too. But I try to force myself to get used to riding the waves of change because I expect that's what I'm going to do throughout my whole life. As a futurist, I expect AI to replace me in about 10 years. I look at 200 news and stories every day. It will look at 20,000 or 2 million or as many as exist out there. I cannot compete with that. I might have a better reasoning that other people will understand about my forecasts and predictions, but AI will do a very good job at describing textbased, you know, descriptions how it comes to a conclusion about a prediction or highlighting a trend.

So, of course, I'm going to get replaced in that sense. I will have to keep on either look for ways of still contributing to that profession or find another profession. But you know to summarize all these I'm 41 I came through LA at the age of 80 I will uh apply for university for becoming an astronomer because I always wanted to be to go into cosmology from the age of six and I have a different professional I'm very happy about it but what if at some point I still go back to uh studying physics and cosmology and astronomy maybe at the age of 80 when I have more time not focusing on making you know predictions and and highlighting future trends.

I'm happy to ride the waves of change because I think that's what awaits us all. >> Yeah, I think that's great advice for everyone no matter what industry you work in right now with how much is changing across the board with these emerging technologies. And then as we wind this down, what did you learn from that testing experience where you were one of the most tested uh human beings on planet earth for a little while? What knowledge emerged from that experience? >> I learned like minor conclusions about health anxiety that I will become better at it. when you have your I don't know fifth colonoscopy over 25 years or your sixth whole body MRI or the 100th blood marker test you will get used to it it's like a learning curve I learned many sort of minor things like that but the most important thing I learned was that even as a physician and someone really having an experience with using technologies to measure my health and provide feedback about that maybe for some reason I thought before this that big experience that longevity such a longevity package will give me like definitive answers.

Here is a status of my health and this is what I can expect and this is what is going to happen to me. But instead of that I receive like um a map of different routes or pathways and I it's my decision how far I want to walk on each of these. In diet I received about 80 if 100 recommendations. I had a a continuous glucose monitoring service on. I I had diet uh diary. I had a nutritionist. I had metaboloms measured all these things, microbiome. I received almost 100 recommendations and and actually that's the route where I was happy to dive into it. And I think I follow about 60 of them. Not every day, but you know on average because I was very happy to keep on fine-tuning my diet based on these good recommendations.

Regarding exercise, I think I was already on the right path. regarding medical screenings, I got a clear package about what I have to look forward to and and I think in many cases when still health anxiety kicked in, it kicks in. No matter how experienced you are, when um your medical records look like an Instagram feed, you know, all the new things coming in, oh, these are, you know, verdicts about your health and the future of your health and it's it's statistically impossible for them not to find something. It took the company uh four sittings, four sessions, two hours each, eight hours in total to tell me what's wrong or could be wrong in the future in my health.

And I'm in quite good shape. They told me if they told me everything that was right, it would it would have taken me two weeks to listen to all the nice things about my health. So they can only focus on what might be wrong or what uh might go wrong later. So it's you would they will find something. The question is h how much how how far along those roots or roads I want to how far I can walk on those roads and I'm happy to devote a lot of attention to many of these because I I will have diseases statistically no matter how healthy life I'm trying to pursue here I try to live to 100 but statistically that's that shouldn't be my goal my population is like 74 75ish especially men Eastern Europe so my let's say population based and maybe genetic markers are not are not pointing to that direction of living to 100 but I'm rooting for that and I'm doing everything I can at the end of the day I I don't actually care whether I will live to 100 but this is going to be my goal throughout the way because I see myself boarding a spaceship as a tourist and even come back after the age of 100 and I think this sort of futurist perspectives gives you um um a calm way of approaching these results.

For example, when I had my colonoscopy that was part of the package, I had my first at 35. I wanted to uh started as early as possible, my second at 40 last year. So I said it's very you know uncomfortable process preparing for that in the 3 five days before that. But if I leave to 100 then I will I will have like more than 10 colonoscopies. So how about just getting to use you know getting used to the whole process. It's not something that's happening now. It's going to happen over and over every five years if everything is right only every five years and not more more often. So I think these perspectives just help me find out that longevity is not a road.

It's not a singular timeline. It's a map and I have so many you know elements and signpost on that map. I have data. I have a status update about my health. I know where this is the benchmark that this is where we are starting from and I have data to keep on making better decisions. Not all my decisions will be good. I I guarantee that. But maybe before the package 30% of my decisions were were positive and I was still living a healthy life. Now maybe it's 70. And I keep on trying to fine-tune it to reach I don't know 75. But that's a map. And I love to have a map to have a chance to reach that desired goal of reaching 100.

>> I love that. Well, Dr. Meshko, I could seriously talk to you for hours. Unfortunately, uh you have many things to do today cuz you're a busy guy like most of us. If anybody wants to find more information about what you're doing and the the work you've got going on, where's the best place to find you online? >> Absolutely. It's called medical futurist.com. I publish analysis there once or twice a week and I share daily news updates and the context, the background of news and studies every single day uh on um on X, LinkedIn, uh Facebook, Instagram and YouTube. I publish videos on my YouTube channel too, but all these links can be found on medical futurist.com.

>> Fantastic. Well, thank you so much for the time and the insights and the information. Um, yeah, really appreciate it and a and a little uh shout out to LinkedIn because of its excellent algorithm that brought us together and led me to your work and I was fascinated by all of it and you did not disappoint. Thank you so much Nick for saying that and I'm really thank you for having me on the show and let's keep on having discussions at least on LinkedIn if not in person but but thank you for this opportunity. >> Absolutely. >> Okay everybody until next time ask questions don't accept the status quo and be curious.

The next daily show.