This episode of Real Talk from AI and I Care is brought to you by Bardi Software. Welcome to Real Talk on AI and I Care with your host Dr. Scott Morris. And I'm co-host Dr.
Rayhan Ahmed. In each episode we discuss current news adding a little AI education, and debate innovative AI tools and topics that will change the industry. Welcome to the show to everybody. So Rayhan's on vacation this week, so I have a special guest Dr.
Ronja Habash out of Bascom Palmer and she is one of my favorite people to talk to. We have conversations often and it goes in all kinds of different places. I am so much. I literally wrote this one on my calendar no offense, random reason, I'm vacation going.
I can't wait to talk to her. A little bit of background about Ronja. So she's also on her innovator series. So if you haven't got a chance to listen to her podcast on that, please do fascinating discussion that she had with Rayhan.
She is currently a faculty instructor in clinical informatics management program at Stanford University. She also chairs the Artificial Intelligence Committee for the American European Congress of Ophthalmic Surgery. She takes up residence at Bascom Palmer where she's an assistant professor. She's also the co-founder of Metamad and I'm sure she's going to talk a little bit about that.
And I'm looking forward to hearing and learning more about that. And then she recently kind of as a congratulations, Ronja. She recently was put as part of the power list 2025, which is the top 10 innovators in the I care world. This is a global award and Ronja is.
I'm proud to say she made the list and she's one of our rock stars of the future. So Ronja, I want to say thank you so much to come and coming on the show. My pleasure Scott. Thank you for having me.
This is a discussion I've been looking forward to as well. Awesome. Well, Ronja, so we always do kind of news of the day. So I want to talk a little bit.
I don't know if you saw Amazon come out this last week. They kind of stepped into the world of AI agents kind of into that arena with what they call Nova Act and OVA Act. It's a general purpose AI that can actually do things like book your dinner reservations or order your food. It's our first real step into online agents and really it's a kind of a competitor to open a eyes.
They have one an agent called operations. Now this is still kind of in preview mode. So we haven't really got to play with it though. I'm super excited to play with it.
But this is kind of the next evolution of Alexa plus. In fact, it's going to get incorporated into Alexa. And so this is kind of think of it as actionable AI that can basically interact with websites and complete tasks and follow workflows. And like any, it just said we had a personal system.
There's going to be a little bit of training that goes into it. But imagine going that, you know, hey, Nova, you know, make my dinner reservation for Friday for two at Garden Grace, which is one of the really nice restaurants here in Denver. And it'll just make those reservations for you and say to me going online and even more cool than that is they're actually releasing. It's going to I don't say it's going to be quite open source, but almost it's a tool kit called Nova Act SDK.
And that's going to let developers build their own AI powered agents to kind of automate a whole variety of activities. So I'm really looking forward to this. I mean, this is kind of that first big step into what we all think of agents and how agents will assist us and talk to us and work with us. And I'm super excited about it.
I can't wait to see where it goes, but that's kind of some of the new news of the week. I just got to figure out how it can do my laundry. That's what I'm really looking forward to. I think that's robots though.
I don't know. Ron, you have you played with any of the agents that are out there yet like operations or any of those kinds of things? And you know, they like you said, anytime you hire a new assistant, for instance, you're, you're still going to have to train them. And so there is definitely a learning curve involved, but I, I see it as a way to really put AI into everyday work, just like we needed in our everyday clinical workflows as well, which I'm sure we're going to talk about.
So this is the first step to, you know, to bring it into the mainstream even further. It's super looking forward to it. Alright, so now what we do is I'm going to also cover easy right on and I kind of take turns on this, but I'm going to cover the educational segment. And this one's maybe a little deeper than the last couple ones.
We're going to talk a little bit about neural networks in the context of AI. And so let me kind of explain. So, you know, as well, no artificial intelligence is really kind of transformer entire world. And this is what this entire podcast or innovator podcast or magazine is all about.
You know, but it's going to include like how do we shop online as we just mentioned to how medical diagnoses are made. But the question is this is like we think about how do people learn we also got to wonder how does AI learn. And one of the most powerful techniques behind modern AI is inspired by none other than the human brain. An artificial neural network or sometimes referred to as an a n n.
Think of it as the workhorse that does the heavy lifting about how AI learns to do tasks. Now, so imagine your brains made a billions of neurons connected in this vast crazy network that is ever changing. It's very fluid. And when you learn something new, connections between these neural strength and weekend and these an ends mimic this basic idea that consists of a whole series of interconnected processing units called neurons or nodes.
They're organized in layers. And remember last week we talked a little bit about how layers, how we will learn how deep learning works in terms of layers. This is kind of the next step a little more detailed. There's an input layers we discussed last week.
One or more of these hidden layers were much of the processing happens. I think last week we've been talking about as a 3D multiple layer, you know, spider web of all these networks and then an output layer, which is this is the output. Well, each connection between these neurons actually has a weight kind of a number representing that connection or that node strength or importance. When the network receives input data like pixels or images or words, whether it be verbal or syntax and a sentence signals transfer through this trap signals travel through the network and then these neuron processes these signals and based on how much weight each one of these nodes have.
The network then produces output maybe identifying object or predicting the next word. It doesn't work that much different than our brain is we, for example, we all have the most important word in our English vocabulary is our name. If I say Ron, you're her entire cerebral cortex activates and she's like, oh, it's about me. You know, and I do the same thing.
Well, that gives that particular neural network in our brain or biological network gives more weight to whatever said next because it's about us. And so the synthetic neural network so the electrical neural networks as we do much the same way and initially these weights are really random. So, you know, when we're training these networks, many times the outputs wrong because we don't have the right weighting system on them. But here's kind of the magic.
I mean, as we train this network with fast amounts of data. And if the network makes some mistakes, the algorithms adjust the weight slightly to make it perform better next time. And as we learn the weights get better and better no different than like we talked the last week about, you know, hey, this is a block. This is a ball.
Well, when we first went to your babies and we said block and it was ball appearance or somebody else said no block. And we give more weight to what a square is versus what around is as association with the block. So this process repeats itself millions of millions of times until the neural network becomes more accurate. Now, the simplest type of these neural networks is called the feed forward neural network or you'll have many times here at FNN.
This is where information flows only in one direction from the input layer through the hidden layers and then to the output layers. No loops, no cycles, no feedback system. It's really just a very linear type of algorithm and many of the early algorithms are really in in AI are really just linear feed forward neural networks saying, you know, this image is a, I think last week we used this is a dog. Right, the dogs got fur.
It's got maybe it's got, you know, ears, maybe it's got a tail. It's got four legs and different. You move between a cat and a dog. This output just really doesn't affect the classification of the next.
It's just very simple linear type of algorithm. Now, what happens when we make multiple FNs and we make a deep in other words, we make it complex. This adds we add many different layers, sometimes hundreds and this consists what's this this builds what's called a deep neural network. And this is what we spent a little time talking about last week.
This depth allows the network to learn incredibly complex patterns and hierarchies of features from data now takes a lot more training because there's a lot more weighted spots that it has to take take. That we have to give weight to different nodes, but nonetheless, it starts becoming much more complex, much more similar to our own human biological neural network. But this is where it gets kind of interesting because what about where day and day to where the order of how we process that matters and we start thinking about medical care and medical decision making or clinical decision support. We all know that the order of this kind of matters.
No, our biological systems. We say the order is ABC. And maybe it is or maybe that's just the way our brains work and maybe a C B would be better. But we simply are going to know that until we have millions of data points to test.
But where order matters like language, you know, I mean, if I give you a whole bunch of words that don't make any sense when you put them together, you're like, hey, but if I jump on around kind of a word, I'm like, oh, now it makes sense. Well, think about, as we said, you know, if F and N's or DNN's that are very linear or even mildly complex struggle with these going because they don't always have a time sequence. And this is where these recurrent neural networks, which is kind of really where we're at in the way our brain works. These have loops that say what's the time sequence with these and they allow information from previous steps to persist and kind of influence the next step much like our emotions and our physical state do it our own body.
This internal memory makes them kind of ideal for things like language translation or speech recognition or stock prices or maybe analyzing OCT data. So as a recap, kind of neural networks are these really interesting brain inspired systems that learn from data by adjusting connections weights much like our own system does and I think it'll be interesting over the next few decades. Years decades that maybe there's a better way to learn than the way our human brain does it, but we had to start somewhere and the current network systems are built on the way we currently learn hopefully every learn to something new today. I don't know any family comments on that otherwise we'll just jump into a real talk.
You did such a good job, so you explain the concept called back propagation and and you did that so beautifully, but it's a very interesting and complex topic, but you know you handled it beautifully. The issue with back propagation, which is sort of the older way of training machines or doing neural networks is that you already know the answer before you train the machine and so this is almost like learning backwards, you know we we don't learn that way we learn forwards, meaning that we are exploring the world and learning and making new neural connections as we go out in the world, think of like a kid practicing on the piano for instance, you know they practice and that's what makes things perfect right. Well, the new generative theory behind AI is to use a different type of model rather than back propagation where we already know the answer and we go back and reprogram the weights if to come out with the right answer again. It it it's forward learning instead and that uses something called hebs rule, which is.
An alleges to the brain pathways. You know if you use a pathway more often than you can rewire it to be more used and that's the new theory behind training machines is to let them figure out things moving forward and then we go back and make corrections afterwards with that new information so it's a little bit of mix of both. Of back propagation and using hebs rule for computers to learn forward as well and again you know just back to your point this is exactly. You know what we're trying to achieve by training machines is to get a machine to think like a human and that's kind of the beauty and the irony behind all of this and maybe that's good maybe that's not good right now not always sure that.
You're right maybe best model for that exactly like you said you know maybe we can learn new new ways to learn ourselves. Maybe maybe we can teach an old dog new tricks I don't know what to see. Well from a neuro for a neuro plasticity mindset yes we actually can you know that used to be the old adage but now we know that we can teach an old brain new tricks as well. There's hope for me yet running is hope for me yet.
Alright so today this is the part this is the main focus today's talk we're going to talk a little about the future of education when AI becomes involved is such a huge topic. I'm sure Ronnie we could probably talk for hours on this one but they're going to limit to us the next amount of time but so I want to kind of start 30,000 food view and I'll just make this general statement and we'll kind of go from there. I think as I said before I think AI will be the biggest change to medicine humanity for sure and then medicine that we've probably seen in my lifetime for sure and I actually think the most likely place it's going to have the biggest impact is education. Of how we train consumers what consumers can learn what they can access how we train our student docs our residents our fellows how we trained ourselves in terms of can we as an old dog learn new tricks right as well as how marketing occurs and how we educate and industry about new products and new services and success and value based medicine and I think there's no end to the way AI is going to impact.
Education in its various forms. So what I want to do is Ryan this is kind of tell me so if the big picture view what are some of two three four the areas you go here's where I see AI having the largest impact in the way we educate humanity. Specifically around medicine. Okay well great well I totally agree with your statements actually and you know that's going to be part of our discussion right now you want some pushback about some things so hopefully.
You know that that's going to be a little bit easier to find but we usually agree on a lot of things so we'll see how this goes. As far as a AI driven education you know we're already seeing a big impact in schools so let's start with that first and then we'll go into medicine because that's a kind of an endless topic but in schools you know let me just start up by saying we have a couple of examples of schools which have implemented AI teachers. And these schools there's one in Texas I think it's called alpha. The school school yeah yeah you're aware of it okay good so it's integrated these AI tutors into its curriculum and it's it actually just teaches two hours a day and during those two hours a day the students learn these core academic subjects and then the rest of their day is devoted to like learning life skills like workshops and projects and things like that.
You do every day exactly right and so you might think the two hours a day just isn't nearly enough but it's shown now that their average SAT scores are I believe it was like 1550 out of 1600 and they're in the top yeah they're in the top 2% this school is now in the top 2% nationally. It's students are in the top 2% nationally and they're only getting two hours of core academic subject training by an AI tutor it's it's unreal you know and so there's actually another school I think in Arizona that does a similar thing and has very similar outcomes so there is something I think wrong about the way we're teaching our students. The same way we were taught 200 year 150 180 years ago right at the brink of the industrial revolution right everybody sits in the class everybody learns the same thing. But that's not the way our human brains work you know I have Salman Khan from Khan Academy and you know they have now come ego which is their AI tutor that instead of going to a classroom now it's virtual he's coming on to do one of innovator series here pretty soon and.
Fascinated guy I mean this thought process about where all this goes but it's kind of along the lines we said Ronnie is maybe I know that I can't sit still for more than a few minutes without having something challenge me intellectually and I think a lot of humans are like that and maybe two hours of really focused hard work is all we really need and if you think about from the biological perspective we only have enough blood glucose in our brain cells. To work for about 40 minutes before we need to break right before glucose levels drop and our brain function starts to decrease so this six seven hours of monotonous education it's not the way we're best built biologically to learn either. That's a very good point and then maybe the the way that the materials presented is different as well like if you're a visual learner for instance you know you you may need a different set of examples or if you learn by storytelling you know maybe it's a history lesson but told as a story and then. Not just the history part itself but also the politics of the region and the geography of the region and the scientific advancements of that time period and textual learning right exactly exactly exactly and.
Like my oldest son who's an engineering school right I mean he learned best if he could do an experiment make it work or fail he died okay I get it but read a textbook was worthless to him yeah yeah more hands on training exactly so and that's exactly the crux of what AI or what. This paradigm of AI teaching is getting at it's precision learning it's precise to how you you like to learn and if we parallel that with medicine the newest you know paradigm and medicine is precision medicine for the same exact reasons because we've been treating patients exactly the same way with the same set of. Of of of of efficacy or of treatments but they have different efficacies depending on the patient and so now with different you know different you know different genotype different phenotype and yet we say. All humans are the same that that's just that's exactly right.
Yeah so they describe as kind of I I call this an adaptive curriculum right based on every person is going to be able to learn the way they learn best in the type of. The best in the time sequence they learn best at the time the day that they learn best you know my brain is an overdrive from about 5 a.m. Till about 6 30 in the morning that's when I get my most productive work is when the rest of the world is asleep yeah power power hour the power hour right that's when I get that's where I really get creative and get my stuff done you know but my two o'clock in the afternoon I should have been born in a Spanish speaking country. I think it's theistic is my brains and off mode at that point time but this adaptive curricular where we can learn you know and this is a debate I've had with other people of dr.
Bersel and I had this debate on his innovator series where he's the dean at one of the colleges. And I said that you know the challenge is is that we don't really know what people don't know we just teach him what we think they need to know whether they actually already know it or whether they don't know it we teach in the same thing. And if we could figure out what knowledge gaps are and treat each individual knowledge gap to fill that gap that's everything you talked about ranya plus kind of that's what I think it was an adaptive curriculum is let's fill in the weak spots let's make the strong spots better but let's fill in the weak spots but your weak spot in my weak spot maybe completely different. Based on our experiences and what we retained and what our who our professor was right did we have professor really got and explained really well and you're like oh I get it I understand it right but maybe you went to a different school and had a professor maybe wasn't as good talking about that particular subject and you didn't learn it as well.
It's adaptive curriculum I think it's it's one type of let's level the playing field and make everybody have the capacity to learn and understand at the same level someday. This episode is brought to you by Barty the AI powered all in one each are built for I care with AI scribe voice over internet phones websites payment processing in our cm all bundled into one system. Barty is redefining modern I care spend less time clicking and more time with patience at Barty dot com. Yeah that's right I actually wouldn't say level the playing field I would say augment people's strengths to help everyone achieve more agreed agreed agreed and that kind of that's like you know it's kind of like morphs into that whole personalized learning kind of thing you know where everybody's learning styles different but now you have that as we said that's what we're doing.
So I think that's kind of 24 7 365 is not a 7 30 to 3 o'clock kind of deal with an hour for lunch in an open campus format like some of our schools are owned here. You know now it's hey I was think the AI and this is what Solomon says in his book you know he basically says that now AI will be your tutor right you'll learn from the AI and your teacher will be your coach. You will have a you learn what AI can teach you but I think the role of educators will be very different just as the role doctors is going to be different. Yeah I was just going to say that you know your doctor will become more like your health coach and your your AI will be more like the doctor a lot of times and honestly that that role could reverse as well you know your AI could be your daily coach to you know using virtual and digital health tools and then your doctor you know your doctor.
But um it's kind of like you talk about it in use of the day right we talked about agents and maybe in the near future your agent what know of our whatever we're going to call our agents in the future maybe your agents really your tutor that helps you learn so so learning can be continuous right I mean through our entire life time not just something we learn for. 468 12 years in school and then we move on and we hope that see somewhere along the line fills in the gaps. Yeah exactly and then think about the myopia epidemic you know these kids are in school for what eight hours a day they get very little outside playing time and then when they go home they have homework right so or they're on screens all day just imagine now if we cut back that curriculum to two hours a day like these other schools are starting to do and have the kids go out and move them. So if you play or do workshop tasks like you said then the myopia pendant or the myopia epidemic will will should tremendously improve so that's just another theory behind you know myopia management now.
Well that brings me to a thing you know and as you said that I kind of started thinking well you know we often get stuck in AI about this the phrase of transparency and bias and equity and equality and all that kind of stuff but it makes me wonder. Will AI really help and you correct me some earlier and it made me go yeah I said that wrong and will AI really make learning equitable across the global population the answer is i'm not sure i'm concerned that the differences will get greater than they ever have been for about the educated versus the non educated instead of the other way because at least in today's world. Not everybody has access to AI tools like you said the alpha school right there's there's a great example well how many kids got to go to the alpha school alpha go school. A couple hundred out of millions in the us alone and we look across Southeast Asia and you look at Europe and you look at African you look at South American go will the ability to access knowledge through educational portals are supposed to get make the world more equitable.
Or will it get less equitable would you take on that run you. Well that I disagree with you a hundred percent it's just like telehealth you know this is opening access to people who would never normally have that kind of. Of ability to access. Education through an AI system or teachers even so you know all these regions that you mentioned you know a lot of these are our war torn poverty stricken there are no resources.
The cost of an AI is far less than a human being able to teach and and have all all of the resources at play so i think that it's going to make it way more accessible and equitable the thing that i had mentioned before pointed out before is that the goal here is not a aggression to the mean the goal is to elevate everyone to a new standard of education. Blue ocean strategy yeah well in that you know that I actually agree with you I just set that up so we can have a little debate on that you know I think that when you think of and I can't remember what's the most recent style I think the most recent one was 2022 that is like 96 point. I don't know one six seven 96 points somewhere around 97% of people in the world have access to a smartphone not a cell phone but a smartphone and. And you know now all of a sudden if that many people have access to a smartphone and granted we're saying that everybody there's no cloaking AI to teach people about stuff they don't want to learn in certain countries and blah blah blah blah say but truly was open access.
I think our phone and maybe someday you know our. Our heads up display in our glasses and that's a whole nother issue we can go into someday you know but I think that I do agree with you I think this is going to make learning accessible for everybody and not just for everybody as a smartphone but people were as we mentioned earlier people are very kinesthetic learners maybe we're going to have. Virtual patients you know or virtual situations or a our simulations or right you know truly interactive learning modules I kind of say you know if we think about learning is somewhat of that you know FN and we talked about it earlier that it's a little bit linear and you make a decision what if you could go back virtually and back up that decision to make the other decision. Decision instead of yes make a no and go further down the no tree and go where does that bring me right so we're making decisions on it bringing it back to medicine we're bringing making decisions on patient care where we're not going to harm anybody by making this decision because it's all virtual.
How does that change care and more importantly experience right I think about how long does it take I've been doing this for almost 30 years now. I have a little experience because I made a lot of mistakes and I've tried to learn from my mistakes but if you're fresh out of school you might have education but you don't have experience maybe you can gain experience in a virtual way much faster than putting in your eight hours a day in a clinic. Yeah well you touched on a couple subjects there I'm going to go back to the telehealth example which parallels really the AI example you know we've had this debate over and over again people think telehealth might might expand the digital divide and further you know inequity but I always thought the different the difference and if you look at it as. Patients who never get care versus the ones who might get care because of telehealth you know for instance at Baskham we have a world famous specialist who some of our patients would never have an opportunity to see but because of the virtual care methods that we've employed we now have patients from you know China Russia anywhere in Europe India you know these patients with really severe injuries are now able to consult with one of our specialists our tertiary level care.
So I think that's a very serious area level care subspecialists anywhere in the world and that is opened up a whole new paradigm of treatment and so I would I wanted to just put that plug in for telehealth and virtual care because that that really does look exactly like what we were talking about when it comes to AI education too and the access to education. So far as virtual care goes for for patient care and practicing on these patients who are not human where we can actually make mistakes I mean that's exactly what we're doing now with surgical education and training using different XR technologies to do so so. These XR technologies are allowing our residents to practice cases in a pretty you know hyper realistic environment and an immersive environment and it's coupled with AI analytics to talk about to show them their stats basically how fast or how well how accurately they perform that caps are excess or what they could do to improve their accuracy or efficiency next time in another case or have them compare side by side with an expert surgeon to do so. So I think that's a very important thing to see how they can improve and so these are the newest you know the newest series behind extended reality and AI education which I know you wanted to touch on.
Yeah and I think that you know that to me and obviously I'm a tech geek I'm going to kind of qualify you in that group too right I mean we're kind of tech eeks which is why we're doing this kind of stuff. You know I think for the tech eeks is learning in a virtual environment where you can make a mistake back up learn to do it a different way and go oh I see now right and it's instantaneous feedback versus oh gosh I have to wait a week a month a year before I see another situation that's similar to that maybe I remember what I learned and maybe I don't remember what I learned. You know a year later or six months later whatever but now you can actually practice so many different variations of surgical techniques or clinical care decisions or how this drugs can interact that we just took and I use this phrase a lot what used to take decades will take days. Yes absolutely and now we're also going to be armed with data analytics to so the percentage of patients who might respond to you know a mix versus a laser versus a valve or two or something like that for glaucoma treatment you know that's the kind of precision medicine that we're talking about here where we're not just making these you know algorithmic well I said algorithmic but human algorithmic or tried and true you know pathways for care it's actually more.
For more precise for that patient situation and gives us analytics to back it up so now instead of you know talking to a patient and just saying this is what I think you should do it's this is what I think you should do and by the way the data backs me up as well there is a 65% chance your glaucoma might improve if we do this this in this. And so there gets into the foundation of evidence based medicine right is now we truly want especially once we start capturing this data crossed millions of people with different phenotypes and genotypes across the world now we have a basis for evidence based medicine that says maybe what we're doing maybe what we've been taught to do isn't really the best for every genotype phenotype combination come as we mentioned earlier. And that's the basis for value based medicine going are you still going to prescribe something that's not the best for that genotype phenotype combo that's right yeah I was going to say that it's not just genotype phenotype there was a study of Microsoft a few years ago a couple of years ago that said you know they looked at what were the biggest social or what were the biggest determinants of a patient's health outcome. And only 30% of the clinical picture emerged from the genotype phenotype 10% was from the clinical data and 60% was from the patient's social determinants of health.
Oh yeah I know that study yeah they're invited there are lifestyle and environment exactly and so imagine this you know whenever we make these decisions about our patients we're doing that with 10% of the full picture 10% of that information and that's what we're telling the patient right exactly. It is scary when you think of it think of it that way but now add genetics to it now we're up to 40% but we still have to add the rest of the patients you know living situation you know where they were they were they live how they have clean water they have every matter in the water I mean is the closest grocery state close to grocery store 7 11 or is it a whole foods. Yeah exactly it will kind of job do they work are they in a high stress environment you know and then we can actually change certain factors like their stress levels their sleep patterns their nutrition their exercise patterns things like that to help tweak their eventual outcome and as we know from things like diabetes and hypertensive retinopathy glaucoma all these things play you know all of the lifestyle and environment roles play play. Play a huge role in the patients health outcome with those eye diseases we really have to stop thinking of the eyes you know the only part of the body that we concentrate on it should really be a more systemic holistic approach for the patient.
I would love to do this visually but I always think my mentor used to say remember there's a body attached to those two things right and that's right. His name is Larry Alexander and he was so right I mean he's like when you stop thinking about the eyes and start thinking that's part of a huge century system that's in the much larger organism. Now you start thinking about the whole big picture right and maybe that's when we come back to you know where we started with the news of the day is you know because the problem with this provider you only have limited access to each one of our patients and you can only learn or know so much about them and got granted a eyes can build a little bit more than that but these agents someday I personally believe and I think it might be seven 10 years away. But I think we're going to have personal health agents that are friends that remind us in the nicest possible way.
Did you remember to take your vitamin D this morning Scott did you work out and did you really work out or were you kind of half assing it today right because they're going to know. Yeah that's right. How do they coaches maybe not do that right and maybe they say well you know hey if you do this I won't remind you later today that you can't have the talk right. Yeah exactly and then you can you can program them if you want you know a very a polite English speak you know English accent or British accent woman politely telling you what to you know only to put one cube of sugar in your tea versus two or if you want like an.
Army coach. Yeah it's your GPS right. You can't be asking. Exactly.
You made the wrong turn versus hey stupid go the other way. Yeah exactly and we've actually made a really nice segue into autonomous diagnostics from the eye to the rest of the body to which is probably a topic we should cover another time since we're almost at a time now but that really deserves that that's a beautiful segue into everything from everything that we've just talked about. Well I want to take two minutes because I want to cover another thing that in education that's maybe not quite about our side of medicine. But about from a consumer side of medicine and education is I think the biggest avenue of opportunity is the fact that now consumers can be educated at potentially about their prediagnosis or you know when we basically comes in and sees me and I go here's your diagnosis here's your treatment.
And you know you got 30 60 90 seconds to do that and then you think that they're going to remember that for the rest of the next six weeks of treatment. Right. Yeah that doesn't happen most people don't remember it when they leave the office right how many times did you say. Oh I had my glaucoma removed or yeah you know I mean okay really.
They're not so dilated they're hungry they're stressed they're overwhelmed by information and oftentimes they're alone and so you're absolutely right like the amount of stuff that they retain is actually like. Incomprehensibly good compared to what I would think that they they should retain and so having an AI assistant help it help with that translation or that education contextually so it can be in their own language at their own you know reading level for instance and really. Whatever time of day that they have the question right what's if it's nine o'clock at night and they're like am I supposed to take the steroid first or the glaucoma medication first what did he say. Yeah you have a reminder and potentially even a little 30 second segment about why that occurs right you know so.
Yeah I think that the biggest my personal belief is though education about training professionals and training students is all wonderful the biggest impact we're going to have is the ability to train. Train and educate humanity. About how to best take care of themselves. That's exactly right because you know there is a role for patients to take care of their own health here you know and I think we're seeing patients becoming more like consumers and more interested in longevity and health and wellness so I definitely think that's you know that that should be top of mind.
So we got to kind of close up here because I yelled at when I go over too much on running any closing comments about the future of education with AI is our helper. Yeah the future is bright and we're just now getting to see it in play and it's very exciting. Thank you everybody thanks audience for listening I hope you got something out of this I wrote tons of notes I do every time I talk to Ronnie she's just a wealth of information which is. You've been listening to real talk an AI and I care you're weekly podcast to keep you informed about AI technologies revolutionizing I care.
We would like to thank birdie software for supporting this episode of real talk from AI and I care.