Full Transcript 0:03 Could AI be your doctor one day? 0:06 I know it's a crazy question, but we have to talk about it. 0:10 It's something that we're going to about today on Everyday AI. 0:14 This is your daily live stream, podcast, and free newsletter where we help everyday people like you and me not just understand AI, 0:25 Before we get into the role of AI in healthcare, and we have a great one, we actually have the author of The Doctor Who Wrote the Book on ChatGBT in Healthcare, 0:35 literally. 0:35 So before we go there though, I want to run down some of the biggest things that are going on in the news world today. 0:41 We do this every single day, weekdays, Monday through Friday. 0:45 If you are joining us live, please go ahead and drop a comment like this LinkedIn user that said, So let's get to it. 0:52 A couple of big news pieces today. 0:54 So first, a new CNBC story says that the AI boom that's happening right now is actually the fourth industrial revolution playing out. 1:06 At least that a quote from an executive from Wedbush Securities. 1:11 Make sure to read more about that in the newsletter today, A Couple of other cool healthcare related stories. 1:17 Gleamer is a healthcare tech company. 1:20 They just raised $30 million to bring more AI and radiology products and services to the market. 1:28 Super exciting. 1:30 Speaking of tools and software, you've probably seen or heard these AI detection tools. 1:35 So we're going to be sharing in our newsletter today a link to a story that talks about how easy is it to fool these detection. 1:42 That's from the New York Times, so make sure to check that out. 1:46 And last but not least, and kind of a good transition into our show today, big news story. 1:52 This is the first AI-generated drug has entered human clinical trials. 1:57 So I'm not going to give you the name of the drug because it's a bunch of letters and numbers, but it has interred phase two clinical trial targeting chronic 2:06 lung disease patients. 2:08 We're going to have all of that and a lot more in our daily newsletter. 2:11 So make sure you go to youreverydayai.com. 2:14 Sign up for that free newsletter, we're gonna have a lots more. 2:17 But right now I'm extremely excited to bring on someone that I've been following in the kind of AI and chat GPT space. 2:27 Let's bring in on our guest for today. 2:30 We have, welcome, we have Dr. 2:33 Harvey Castro, the author of Chat, GPT, and Health Care. 2:37 Dr Harvey, Castro. 2:38 How are we doing? 2:39 Awesome. 2:40 Thanks for having me. 2:40 This is exciting. 2:41 I've been wanting to do this for a while now. 2:44 Absolutely. 2:45 So I can't wait to talk about a whole variety of things. 2:48 But as a reminder, this is for everyone out there. 2:51 Everyday AI is, for everyday people. 2:54 So if you do have a question for Dr. 2:57 Harvey Caster, please leave it, and we'll get to it. 3:00 If you're listening later on the podcast, you missed out to get all your questions answered. 3:06 We do quite a few people already with questions. 3:09 I have question you first, Dr Harvey. 3:14 I started the show with maybe a little bit of a sensationalized statement, but is that too far off? 3:20 Might we have doctors or a system one day that is just primarily AI? 3:26 Is that a thing? 3:28 I pause for a second and hesitate to answer because My gut says no, but technology is changing so fast. 3:37 And with robots that I'm seeing and with things that are coming, it's starting to make me think twice that, well, maybe I won't have to concede here and say, 3:44 you know, I do see a robot in our future where it is going to be a resident equivalent where there'll be my assistant. 3:51 I don't want to replace myself yet, so I want say not yet. 3:55 Yeah, that's the, I think you come from obviously a very unique position. 3:59 So you obviously have a background as a practicing physician, as more of an executive role in healthcare systems, but also you are literally pushing the 4:12 education around this technology with AI and healthcare, chat, GBT and health care. 4:17 I guess, where do you lean on it though? 4:20 Yeah, let me kind of explain my rationale. 4:22 So OpenAI, and I'm really bad with names, so forgive me. 4:26 They have a robot that's coming out by the end of the year and it will have what's called tactile function. 4:32 It will be able to sense and give information back. 4:35 Well, just extrapolate here a couple of things. 4:38 If you have the equivalent of, I call it a chat GPT-5, where it can see and take video and pictures where four has it, but we just haven't been able to 4:46 use it due to GPUs out there in the world. 4:48 But just play with me for a sec. 4:49 So we have this fake robot. 4:52 Let's put a white coat on him. 4:53 And say it's able to see, and now with the tactile, it can feel the patient's belly. 4:59 And for example, they will say, oh, wait, this is right upper quadrant. 5:02 It's by the gallbladder. 5:03 it sends that feedback, looks at the differential and says, OK, is probably this. 5:07 I'm going to go a little extra here with eyes. 5:10 In the eyes, there's already technology that the video camera can actually give my age, my hemoglobin A1C, which is the last three months of my sugar, 5:17 can give me blood pressure. 5:20 And it could give about five or six different parameters, just with the camera. 5:23 So I'm thinking, man, you put that camera on this robot, dude, like a lot of information. 5:29 You have your triage nurse, your resident, and it's all being packaged. 5:33 Due to the hallucination effect, I don't think it'll be chat GPT. 5:37 It'll another bio-GPT that's healthcare related. 5:42 And that why I hesitate it so much because I am in my mind thinking well, if you'd put all these pieces together, then you would have something very close. 5:50 Yeah, that's a good point. 5:51 So we're obviously already starting at the end, which I love, but let's hit rewind and let remind people just of your background, 6:00 because it's extensive. 6:02 I think people might hear us talking and say, oh, this GPT or AI thing is brand new. 6:10 Who are these guys to be talking about it? 6:12 But you have a huge, in-depth background in technology and in the healthcare field. 6:17 So just give everyone a brief overview of your background. 6:21 Yeah, I'm blessed in that I wear two hats. 6:23 I am an ER doctor by training, so I've done it for over 20 years. 6:30 But simultaneously, I'm an entrepreneur, have had over 30 iPhone apps into the technology world. 6:36 I've actually wrote my first app, which was IV meds. 6:39 And I spent a lot of time looking at technology and merging it with health care. 6:43 As an entreprenuer, a joke, i feel like I live in the clouds. 6:46 i live and dreaming and I always think of things and i love merging the two languages, let's say the language of medicine and the Language of Artificial 6:54 Intelligence for today's conversation. 6:56 And so that's kind of my background. 6:57 And just really quickly, back in November, I was playing with chat GPT, and I Was blown away. 7:05 I had that iPhone moment. 7:06 This is going to be the next iPhone, but better. 7:09 Let me write a book real quick and explain how it works and tell our teachers and doctors and patients. 7:19 Yeah, yeah. 7:20 And if you're not watching live, I do have Dr. 7:24 Harvey's book on the screen here. 7:26 So he literally wrote the book On Chat, GPT, and Health Care, Unlocking the Potential of Patient Empowerment. 7:32 What was that process like? 7:33 And what was the reception as well in the medical community? 7:38 Well, it's kind of funny. 7:38 Obviously, medicine is really conservative, and I'm here pushing the envelope and telling patients, hey, use Dr. 7:45 Google, but let's use ChatGPT and tell your doctor we're going to use it together. 7:48 And I know doctors were upset with me. 7:51 It's really tough to write this book. 7:52 You know, when I wrote the first edition, this was, mind you, January 5th. 7:57 Here we are like six months later. 7:59 I've already talked about several books out. 8:00 This technology changes so fast. 8:02 But the skinniest I see life as a bill curve. 8:04 We have people in one extreme that know their stuff really well, and when I wrote it, they're like, haha, this is just scratching the surface. 8:11 And in my mind, I'm like it's actually not meant totally for you. 8:14 It's more meant for people that don't understand this technology, haven't thought of all the possibilities. 8:19 So I am just kind of presenting it to them. 8:21 At the first two people, including my wife and close friends, were like, ChatGPT, who? 8:26 What are you talking about? 8:27 I have no idea. 8:28 And I was talking to my doctor friends and I'm like hey, the last thing you want is a patient to show up with some papers saying this is Chat GPT and you 8:36 don't even know what this and what the good, bad and the ugly is. 8:39 So when I made that statement, I said, you know, maybe I should start teaching people the bad, good and ugly as far as doctors and patients. 8:46 Yeah, absolutely. 8:47 And you bring up something that I wanted to talk about. 8:50 Did you say Dr. 8:51 Google, right? 8:52 Is that is that a saying in the medical community? 8:53 Right. 8:54 So I, like, I think when people talk About AI and healthcare, they think, Oh, no, Absolutely not. 9:01 You need to trust the doctor's knowledge. 9:03 But at least when I go to see the Doctor, so many times they're just Googling something, Right? 9:08 So I say, what's the difference? 9:11 So do you think that people in the medical community, in health care profession, are people for some reason biased against AI or are they scared of it, 9:23 even though it is kind of just a 10 times more powerful Google? 9:28 So I'm gonna hesitate one more time. 9:30 I think again, it's a bill curve. 9:32 We have our newer generation doctors that are like, let's go with technology, I got it, lets see it. 9:38 But then we have conservative older doctors, that honestly some of them can't type and I am not disrespecting them in any way. 9:44 They just didn't grow up with the typewriter and typing and so typing they literally have two fingers going or they ask for an assistant to type for them. 9:51 And so we two different populations. 9:54 The key that I'm trying to teach my fellow doctors and patients is there's a balance. 10:00 I don't think AI is at the point where it could just take over my job because honestly, it's called the art of medicine because it is a true art. 10:07 And AI can only look at facts, but it also hallucinates and makes some mistakes. 10:12 And in this world of healthcare, it's not like you bought a bad stock because AI told you to, hey, maybe someone died or maybe some bad effect happened. 10:19 So healthcare is a little different in that lives are huge and we can't make those decisions without a human as of today. 10:26 And so that's kind of my soapbox on that one. 10:28 Yeah. 10:29 So you ended it, and I'm going to poke and prod a little here because I am interesting. 10:33 You ended with a caveat of as of today. 10:38 Is that going change? 10:41 Do you see as an example in the future how the transportation industry, as example, is switching over to everything more autonomous? 10:53 Will it ever get like that in health care where the majority of decisions are maybe made by AI and then it's just more of a doctor or physician checking 11:03 everything on the back end? 11:04 You know, honestly, I think we're headed that way. 11:06 It's, again, a third hesitation here is. 11:11 The art of medicine is a true art. 11:13 And there's many times in residency that my fellow resident or my doctor colleagues, mentors would tell me, hey, Harvey, 11:21 sometimes you have to be a doctor. 11:23 Sometimes the facts in front of you tell you one thing, but you need to step back and look at all the fact and become a Doctor in the sense that you're 11:31 making that decision. 11:32 And so using that principle, AI may collect all the data, may look at everything, they point me and may bias me down a wrong road. 11:40 But my experience, my gestalt, looking at it from 10,000 feet high, putting all of the factors together, because there's also things like small cues. 11:49 Like if I look It the way you talk, your mannerisms, you're tone. 11:53 Yes, AI eventually can catch all those things, but you put it all together. 11:56 That's so many factors that is hard to train a model. 11:59 So my point is, I think it's going to take more than just having the doctor having a robot gather all that information. 12:05 I Think it can truly take a doctor to do it. 12:07 But here's the beauty of all this craziness. 12:10 I really think there's gonna be some AI bias in that I've worried that my younger, I'm gonna call myself an old man here, 12:17 my young doctor friends, they're not gonna have that gestalt because they are so used to using AI and artificial intelligence and other tools that they 12:26 gonna over rely on automation, over relying on technology. 12:30 And I worry about those doctors because if they don't do both, there gonna do many mistakes because it just won't have the experience. 12:37 Yeah, no, that makes perfect sense. 12:40 Do you have a comment or two I wanted to bring up here? 12:43 So Ryan Martin says, great book, read it twice. 12:45 Ryan, thanks for listening. 12:48 You know, Professor Dr. 12:49 Mohammed is saying, chat GPT is now binding professionals, too. 12:52 Absolutely. 12:54 I do want to follow up something that you said. 12:58 But also look, again, look forward, because I'm always trying to look for it. 13:02 But what areas of health care, you know, if we just talk about GPT, right? 13:08 Because AI has been around in medicine and in healthcare for decades, but if you're just talking about chat GP and what can be built around this technology, 13:19 What areas of medicine or healthcare do you see the GPT technology being most disruptive, whether it's for a bad thing in terms of jobs and careers and professions, 13:31 or for good thing, in creating more possibilities and more capabilities for the healthcare system? 13:39 I love your question because I could have answered it differently had you worded it different. 13:43 So let me go with GPT since you stressed on GPt technology. 13:48 I think one of the biggest problems is actually communication. 13:52 You can apply that to marriage. 13:56 One of the biggest problems with doctors and pain points is not being able to speak their language. 14:01 Meaning, as a doctor, you're taking all this medical jumbo stuff, but then to be a really good person to educate, You need to break it down to a five-year-old. 14:09 And if you can break down that level, then I think you know your stuff really well, But better you could converse with that person. 14:15 So bringing it to ChatGPT, Why not? 14:18 Just let's change discharge instructions. 14:20 Let's just take that one part. 14:22 Just go into the discharge instruction and say, okay, I have a 50-year-old. 14:26 It's way different how I talk to a male versus a female versus 5- year- old. 14:30 Not them being sexist. 14:31 There are certain things in their healthcare that is a factor that you would talk with them differently. 14:36 Taking it to another level, there's different cultures. 14:38 I grew up in New York City, so I'm very sensitive to different culture around the world, and I grow up with them. 14:43 And I went to my friend's parents' house, then we would eat and do customs. 14:46 Then I remember doing mistakes in their house. 14:48 They'd be like, no, we don't do that in our culture. 14:50 Oh, wow, I didn't know. 14:51 And so my point is, why not use the power of a GPT to help me create those examples? 14:57 So I'm talking to the patient, I've been respectful of their culture, and being mindful of how, but let me give it to you real quick, 15:04 another quick example of that would apply. 15:07 I usually talk about this example, but the skinny is this. 15:10 Have a five-year-old, pediatric asthma, first time, kids scared. 15:13 I help them out, give the parents discharge instructions. 15:16 That's how it's done today. 15:17 This one hospital did the following. 15:19 They took the discharge instruction, put it into GPT and say, hey, I'm talking to a 5- year- old, convert this into a coloring book. 15:25 Then they took a dolly equivalent, they made the pictures, They put all together and said, here son, Here's the discharges instruction but it is a color book, 15:32 Now that child can go through with the parents, with a doctor, the nurse, whoever, but it can actually look at it and be like, 15:38 oh, this is why I need to take my meds. 15:39 Oh, This is what I'd need a check my blood, blah, I mean, it really takes it to another level. 15:44 And then you can over and over. 15:45 I'm like that is the future of health care. 15:48 That's the way it's going to go, at least for that one example. 15:51 Wow, that example really hits home for me because I had my first asthma attack as a five-year-old. 15:58 So I don't remember a lot, but I remember being scared and confused. 16:01 And I can only imagine how much better I would have felt or how more comforted if my discharge instructions were a coloring book, 16:11 teaching me about how to treat this. 16:13 Even with that, a great follow-up. 16:14 We have some great questions. 16:15 I'm going to get to them. 16:17 Thank you guys for your patience. 16:18 But on that I think one of the knocks is on GPT or AI is that it's not empathetic. 16:27 And you just gave a great example of how you can use this technology to create empathy and everyone thinks it just like, 16:33 no, it is robotic. 16:34 It's non-empathetic and then obviously there's the very famous study now about chat GPTs versus doctors in terms of bedside manner and chat was actually 16:45 rated higher in how do you see the GPT technology improving empathy or is it or might it not? 16:53 No, I definitely think it will. 16:55 Let me give you an example. 16:56 I already think It'd be great use for medical students and doctors in hospitals that, you know, at one point I had over 50 doctors that were under me that 17:05 I have to manage and I realized at Not going to be disrespectful, but I felt like I was a high school principal telling doctors, 17:12 I'm like, are you kidding? 17:13 Why'd you do this? 17:14 Or X or Y. 17:15 And so my point is this, if I could use the power of a GPT to teach empathy, to create modules, and have them interact and make up fake cases where the 17:26 GP or the chat GP will give them corrections on, yeah, this is how you things. 17:29 This is what you did. 17:30 Because honestly, Again, I know Naka and some of my fellow colleagues, they're so smart and they are so into their one part of medicine that they forget the... 17:39 Some of them have really bad bedside manners. 17:41 And I think AI can put it together with the human and teach. 17:46 I see a future where we have a laptop or our iPhone, and our doctors are gonna hate me again for saying this, but we sit down with our patients and we 17:54 that laptop, we the patient and have the topic. 17:56 We're like, let's do this together. 17:57 Let's put into a GPT equivalent. 18:01 Give you my medical expertise and let's put it together. 18:04 I don't think it's one or the other. 18:09 Actually, now's a great time for this question from Yaddy. 18:12 Yatty, thank you for the question. 18:13 Because this perfectly plays into what we were just talking about. 18:17 So she's asking, how is HIPAA impacted if and when users input their medical data into the shared chat GPT servers? 18:25 That's great question, because I think that's one of the biggest hesitations, I Think, with GPTs in general, not just in health care, 18:32 is privacy and security concerns. 18:36 How do you handle that one? 18:37 Yeah. 18:38 So great questions, probably one of the top 10 questions. 18:41 And here in the United States, it's hip and outside. 18:43 I'm really bad with Europe's privacy laws, but it is equivalent. 18:46 They have it just everywhere is what I am getting at. 18:49 How to do this is as follows in my opinion. 18:52 Number one, I would never put any personal information into chat GPT. 18:57 Keep it very generic. 18:58 Make sure you're not using your address, things like that. 19:00 Obviously, ChatGPT has your IP address, so in theory, they already know who you are. 19:06 But for privacy area, the added in their settings a way that you're not training the model. 19:12 So go into your settings, especially if you paying for Chat GPT-4, and go in there and just take off that, you not train the models. 19:18 That would be one way. 19:20 Second, again, I'm going to reinforce, do not put your information in there. 19:24 Just put other stuff. 19:25 And how doctors are getting around it is they're putting the nuts and bolts of the information of what you're just coming in with. 19:33 So diabetic, or in your case, pediatric, asthma. 19:35 But they are not saying you are a five-year-old. 19:37 If they do say you a are five years old, they aren't saying your name and your social and other identifiers. 19:41 That would be one way. 19:42 Yeah. 19:44 Great question, Yadi. 19:45 Fantastic response, obviously, Dr. 19:46 Harvey. 19:47 But I have a question because we mentioned the security piece or the Security Concerns data, all of that, with Chetchi BT. 19:57 How long do you think until a lot of big health care companies figure it out? 20:03 Or maybe they have, I don't know, about just having their own large language models so they can tap into the power of GPT, 20:11 but keep everything private and then also train it on their own region or their policies. 20:17 I guess, why haven't we seen that yet? 20:19 Or is that just the next big innovation in terms of healthcare? 20:22 So you can tab into it, Perfect question. 20:27 So number one, there are technologies out there that are looking at scrubbing the data so that before it goes to an open AI, 20:34 for example, it would be scrubbed and so, that would never happen. 20:38 Two, they are look at models of what you just said. 20:41 Create your own LLM. 20:44 I'm from New York City. 20:47 They're looking at a database and creating their own. 20:50 I think that's what people are going to do. 20:52 There's two things, I know a lot of people don't know this, but with ChatGPT, what makes it powerful is the human reinforced learning. 20:59 My angle is, instead of saying it human, let's say it's doctor-reinforced learning, but then not just stop there. 21:06 Let's talk about doctor specific, meaning I'm an ER doctor. 21:09 I should only do human- reinforced learning or clinician- re-enforce learning in ER medicine. 21:19 I may look at it and say, oh yeah, that looks good. 21:20 But in reality, it's totally wrong. 21:22 That's not what's done. 21:23 And so I think that is the future. 21:25 and that's what was going to go forward. 21:27 Not promoting, I don't have any stock, but there is this company called Glass Health AI. 21:31 that what they're doing, what I just mentioned. 21:34 but it is a private company. 21:35 Another is The Hippocratic AI, they are working on this as well. 21:39 So these are just two that I know for sure. 21:41 I now there's more than that, because I write about it in one of my books. 21:48 Yeah, I can't wait to see how that unfolds and even as a patient, right? 21:52 Like I'm looking at this from that perspective. 21:55 I am excited for this technology because it's going to improve care, at least. 22:00 So a couple other questions and comments. 22:02 Sarah, thank you for your comment saying the human touch and interaction is so important and AI can be a great tool to enable. 22:08 us with communication with patient and families. 22:11 Thank you for that. 22:11 100% agree. 22:13 Ryan has a question here. 22:15 So following up on the mid-journey coloring books, never thought of it for pediatric cases, saying it's brilliant. 22:22 I agree, but a questions for you. 22:24 Do you see a drastic change in medical education with AI? 22:28 Because I know you've talked about this before, how hard it is for physicians and how much they have to relearn on an ongoing basis. 22:35 How will ChachiBT or AI in general help with education? 22:40 Love that question. 22:42 I've spoken to several companies and I have seen some on the news that basically are sharing that they are changing the way we educate. 22:49 And not only the ways we education, but the testing. 22:51 So how will it change? 22:53 In medicine, we use something they call up to date. 22:57 And obviously the goal is to stay up-to-date because technically textbooks take forever to print. 23:02 So it's online and they can change it. 23:04 But what's funny is that up today is technically not upto date and so what they're doing is they are using the power of a GPT equivalent. 23:12 Let's just call it auto GP T where it is going out to the internet or going on to medical references and is telling it go into PubMed or whatnot and then 23:20 plug that into up to date, and that is what the company I've heard is starting to do. 23:26 And so what's that going to in the future? 23:28 Well, I'm already seeing different companies talk about creating a GPT equivalent into education. 23:33 For example, i'm going blank on the name, but I think it's Wharton School of Business out there. 23:38 They have a branch where they're teaching using AI. 23:42 I remember hearing a talk where the professor said what used to take him Two hours to explain and to create the content is literally taking him minutes 23:53 and so he was kind of sad that it took him years to Create all this content. 23:56 He said if he just sat down with his team He could probably not get out in a month. 23:59 And so that's just the power of education So my point is a medical school will change as long as as well as all the other schools law school and all The 24:07 other undergrads Yeah Change I think change is the name of the game here you know even if you look at when chat GPT first came out in November of this 24:19 year to where we are now things have changed drastically because of GP T technology and other companies being able to tap into it and kind of build their 24:27 own models off of it. 24:28 So as a reminder, if you are tuning in live, please leave a question for Dr. 24:32 Harvey. 24:33 But I do have a questions because there's a whole area of your background that we haven't even tapped into. 24:38 And I would be remiss if I didn't ask some of these questions. 24:41 So we even opened up the show with one of the news pieces talking about just how much money is going into AI startups or established health tech companies. 24:53 You have a background as an investor and an entrepreneur. 24:57 How do you see just outside money combining with AI in the medical field? 25:04 Could it be too much all at once? 25:06 Or do think it's a good thing to see that innovation happening with this new AI and GPT technology? 25:12 Yeah, that's a tough one. 25:14 And let me explain why. 25:15 You know, and I'm going to go to my soapbox here for a sec. 25:18 So excuse me for doing it. 25:20 Elon Musk invested into OpenAI because he wanted to keep it open. 25:24 He wanted it to this information away from someone called Microsoft or just a big corporation. 25:32 The goal is if you think about it, if have Open AI or you have some kind of GPT equivalent, non-touch with private money or public money now, 25:41 Then the goal is you could do whatever you want because you're focused on the gold. 25:46 Once you have an external factor, and I love Microsoft, for them to come in and put money now, they have to answer to shareholders. 25:53 And I know OpenAI is not owned by Microsoft but 49% of it is. 25:58 So it changes it. 26:00 How does that affect my statement in medicine? 26:01 Well, it makes it a cash 22 because Now there's going to be that pressure, external pressure saying, hey, we need to produce money. 26:08 Now we got to do X, Y, and Z. 26:09 Whereas if you're just doing it of the kindness of your heart, you like, Hey, let's create a better health care. 26:14 There is not that external factor. 26:15 You can do more. 26:16 So that's one issue. 26:18 With that said, it's unnecessary evil. 26:20 We need the money to create. 26:21 Here's the opposite side of that. 26:23 If you fund it so well that the company doesn't do the judiciary duty by using that money well, and they waste so much money that when it goes out to the 26:35 consumer now, it costs so money for the hospital and everybody else, you kind of defeated the purpose. 26:38 Now it's actually increasing healthcare costs to use this tool. 26:42 Yeah. 26:43 Speaking of costs, I think one thing that can be offset, obviously, is using AI and GPT in the medical field. 26:52 So Monica's question here, a great one. 26:56 How will AI impact the medicine industry in areas such as procedures and surgeries, lab work, and robotics? 27:02 Oh my gosh, this is awesome. 27:04 So one of the favorite sayings in med school was, see one, do one teach one. 27:08 And so applying that to a procedure, I've already seen this amazing surgical AI. 27:15 Again, don't ask me for names because I'm horrible at names. 27:17 But the skinny is, either put on some VR glasses or a monitor. 27:22 The skinny, is I am in the surgical field doing the procedure. 27:24 but the AI is identifying the human body part. 27:27 It's telling me, cut here this angle, giving me little dots. 27:30 I can cut and it's walking me through. 27:33 Other things with AI, it's just that virtual AI. 27:36 You know, another soap box. 27:38 Basketball, they say if you can just sit in bed and just pretend you're throwing the ball, you'll actually do better at free throws than just throwing it. 27:44 I mean, that's another way of practicing. 27:46 Taking it to the medical field, if can see and visualize your patient. 27:50 Say you have a digital twin and you are able to do the procedure and adding AI and now you able navigate and see what you were able do and then you actually 27:58 cutting and practicing the procedures. 28:00 As a resident, I hate it when they were like, OK, Harvard, you see how we did that? 28:04 Here you go. 28:04 And I'm like crap, i'm having to do this procedure. 28:08 How nice would it be to have AI walk me through? 28:11 I am not plugging in this company i am consulting for, but it's called, not to give you the name, it is basically AI and Ultrasound. 28:18 And it's navigating you how to do ultrasound. 28:20 It's basically letting the common person at home be able to ultrasound themselves. 28:24 And I thought, man, this is the way to to procedures and to things. 28:27 So how will that apply? 28:28 I talked about procedure surgery, lab work, and labwork, I see it being more automated where we can talk to our data. 28:36 You know, one of the most frustrating things, including for me, is when I get my physical exam and I got my stuff in labs, 28:42 and I know there's graphs, but how nice would it be able to upload it to a GPT equivalent and say, okay, what's my trending glucose? 28:48 What's trending whatever? 28:50 And it would give me the numbers without me having to physically look for it. 28:53 Yeah. 28:54 That's how it's going to change. 28:57 So that makes perfect sense from the physician side. 28:59 But I'm also wondering from, the patient side, so an example I am thinking of, how you said, oh, you can ultrasound yourself. 29:07 that's amazing. 29:08 I think one thing, at least here in the US, that patients are maybe frustrated with in health care, is just the time it takes to get in to see a specialist. 29:17 Accessibility. 29:18 It can be weeks or months, depending on where you live, just to see a specialist. 29:24 Do you see AI technology maybe changing this? 29:28 As an example, could GPT come into the everyday health care system and automate so much of the manual processes, paperwork, 29:36 all of that to open up capacity? 29:39 Or how do you that playing out? 29:40 Yeah, love that question. 29:42 I saw this amazing model and I've seen two actually but the skinniest this just as you and i are having a conversation. 29:48 If we were if I was being your doctor I'd have to look down and figure out my keyboarding and type and then I hate when you meet with the doctor and they're 29:54 just like turned away and looking at their computer or their. 29:58 They call it cow, computer on wheels. 30:00 And so instead, what I see happening in the future is, and I already see the product, they have AI listening to the conversation, 30:07 transcribing the conversations, And I'll take it to a next level. 30:10 This camera that I have here, being able to take my physical exam. 30:13 So literally, I don't have to document anything. 30:15 If anything, just look at it and make sure that everything's correct. 30:18 And so that's going to free up my time. 30:19 It's gonna make it more efficient. 30:21 Now, talking about from a patient's point of view, how to see a doctor, I talked about this at AI Med. 30:27 I see it feature where I wake up in the middle of the night, um, having symptoms and I'm having a conversation with the GPT equivalent backed by a Doctor 30:37 and just kind of going through it. 30:38 And then all of a sudden, telemedicine pops up and says, OK, i'm your doctor. 30:41 i am here at the ER that is close to your house. 30:44 What's going on? 30:45 And then I'm like, you know what? 30:46 I need to have I needed to see me. 30:47 So come but then play with me here for a sec. 30:51 I have my ambulance, my virtual ambulance come get you. 30:53 It's like a Tesla equivalent. 30:55 Self drives, it comes as you're in the ambulance. 30:58 That camera I talked about early in this show, is giving my vitals, giving me information. 31:02 Me in real time, I am seeing it on the other end. 31:04 Oh wait, this person needs an aspirin now. 31:06 So then it's printing an aspirin, and then you take the aspirine. 31:09 And then by the time you come in, I already know you. 31:11 I've already seen you, haven't seen the feed, have all this data. 31:14 Like I mentioned, the hemoglobin A1C, already have your sugar. 31:17 Already know a lot of things that are going on. 31:19 So I think this is the future and I hate saying future because I see a lots of these products today and that's going to come sooner than we think. 31:28 So speaking of the future, and if you do have a question for Dr. 31:33 Harvey, ask it now, because otherwise I'm going to end on this one. 31:36 So Speaking of The Future, sometimes I put people in the hot seat. 31:40 I want to try it again here. 31:42 The same way. 31:43 Let's take an easy example. 31:46 Media, the way that we consume media has changed so much. 31:49 It used to be just, you'd go rent a movie from Blockbuster and read a newspaper, right? 31:55 Now it's streaming directly to the phone. 31:57 and push notifications for what you want, what need. 32:02 If you were to take a look at where healthcare was maybe five years ago versus where it's going to be in five-years, What do you see? 32:11 What is the equivalent of the streaming media, the personalized push notification that we get information? 32:18 what do see happening in that regards for healthcare in the medical industry? 32:25 Unfortunately, I see those advances outside the U.S. 32:29 before we see them here, but I think those outside countries will produce things because of legislation and because the need, 32:35 risk versus benefit. 32:36 Let me explain real quick. 32:38 In Africa, we have areas in Africa that don't have internet, don t have access to healthcare. 32:44 So it will be a combination of using satellite internet with a GPT equivalent and a physician assistant, but that will go out to the masses and be able 32:52 to take care of patients. 32:53 And then take it to a next level, there'll be places where the government will subsidize a gpt equivalent on your phone and it'll your healthcare and they'll 33:02 allow it because they will say, you know what, risk versus benefits, either this or nothing and we need our population to be to catch things. 33:08 And so the future will have these things. 33:10 Now, will that happen here in the States? 33:12 There's legislation, there's a lot of issues. 33:14 I know doctors are about to, I'm sure my phone's about go crazy for saying that, but I think it's gonna take a while here, 33:21 in U.S. 33:22 We'll get something like that but it'll be a lesser degree. 33:25 But I the pressures from other countries and depending who's the president, we'll change the legislation. 33:30 And to answer it, We will see those changes here. 33:33 It's going to be exciting. 33:34 But regardless, you have, I think, a great, like if you're listening to this, watching this you, have a, great resource because not to plug Dr. 33:44 Harvey too much, but he has literally written the book on chat, GPT and healthcare. 33:48 He's putting out great information. 33:50 So whether you are concerned, what this is going to look like from a patient point of view, or if you are a physician listening in, 33:57 make sure to check out Dr. 34:00 Harvey Castro. 34:01 In our newsletter, we're going be linking to all of his different resources. 34:04 We didn't even get into it. 34:04 You have a full podcast that is produced by GPT and AI Voices. 34:10 It's amazing. 34:11 But I don't want to take up your entire day, because we could talk about this for forever. 34:15 So Dr Harvey Castor, thank you so much for joining the Everyday AI Show. 34:19 Thanks for having me, guys. 34:21 All right. 34:21 And just as a quick reminder, go to youreverydayai.com if you want more information. 34:26 If you weren't able to keep up with everything we talked about, don't worry, we got you. 34:29 We're going to have all that in our free daily newsletter. 34:31 Also, make sure to subscribe to the podcast on Apple, Spotify, whatever people listen to podcasts on. 34:37 So we hope that you'll join us tomorrow on Everyday AI. 34:43 Thank you
Comments