Practical AI for Dentistry – Save Time, Achieve More – PDP260

If you're not going to be using AI, you'll just be left behind because everyone else who has that advantage of basically having the expert in your pocket on anything, I'm never going to be the best communicator in the world. AI can close that gap 100 times 100% better than what you already are and that's how I see it is that an extension of yourself. Context is key and this is the biggest thing that people miss out on because now AI's get better and better with worse and worse prompts as the models improve. But context is everything. For example, let's say you have a patient with an external cervical root resorption and I need to explain that to them. If they're like a mechanical engineer, I can say like, okay, well imagine you've got a cardor and then there's a bit of rust coming from the bottom of the cardor and you have to sort of file it down and fill it before it gets worse. So it can come up with these amazing strategies that I'll never be able to think about. It'll make an experience much better for patients and for dentists because it'll be less effort for the dentist and there'll be much higher up quality for the patient because the patient will have something that they can really really understand. Welcome to the Prochrusive Dental Podcast, the forward-thinking podcast for dental professionals. Join us as we discuss hot topics in dentistry, clinical tips, continuing education and adding value to your life and career. With your host, Jazz Gulati. AI is currently everywhere. Some massive buzzword has been for a few years now, but can you believe that the time of recording this 86% of the world of the entire world had had no interactions with AI. And only about 0.3% of the population I actually paying for AI. So I have a monthly subscription with AI and that does our guests say we'll talk about that as well. And so I'm one of 0.3% of the population. So the reason I'm mentioning that is we barely scratch the surface. And so this episode is about making AI in dentistry tangible. Me and my guest, Dr. Das Kaspreck, assume zero knowledge that you have in AI. Like we cover what is a prompt, what is an AI model? Which one should you use? Is it safe to put patient details inside of AI? And more importantly, how can we actually start to get our time back be more efficient? Enjoy our lives more. Have an easier life through the power of AI. And so over the next weeks, we'll talk about other more advanced ways we can use AI. But this episode is very foundational because many colleagues, especially on protrusive guidance, message me saying, hey, I don't have time. I am a busy clinician. I come home to my family. I haven't had time to properly give AI a go. Most of the people listening are watching this right now, even though I've been banging on about AI notes for two years now, are still not using AI notes because they just don't know how to get started. Well, let this episode be your guiding light. Let this episode pave the way for you to get into AI because trust me, it is worth getting into AI is super powerful. It can do bad things. But if we use it ethically, it's going to save us time. It's going to reduce our stress. And we talk about all the different mechanisms in this podcast. So this is for every single dental professional who hasn't really given AI a shot yet. You will understand it a lot more after this episode. And for those of you who are already deep into AI, you'll hear some cool things that Daz is doing with AI right now. Now every PDP episode, we give you a preclusive dental pill. And today is a free AI tool that you can do something really cool. You can feed it an image, an image of a smile like a standard smile photo that we take for documentation and the AI will do its magic and give you the same image back, but with a smile makeover. So this is like one of those instant smile makeover things like I remember five years ago, I used to send cases like these. I send these images to a dental lab and they would use some software to give you a few smiles and then they would email it back. And even they are a bit sketchy, but at least by 50 pounds for that service. So if something you could show to a patient, hey, this is kind of like what your smile could look like. And so now we don't need to pay anyone. And certainly I would say you don't need to buy any expensive software that does this. Now there are some smile makeover simulation software is there out there. And I'm telling you, you don't need them. It's free. And I'm going to share it with you. And I just want to let you know that I did not invent this. And even Jason Lipscomb, who actually presented this or curated this, he didn't invent the technology, but he did the hard work of making it into a model into a project and allowing it to be embedded into chat GPT. So I don't want to take any credit for this. All credit to Dr Jason Lipscomb who's doing some really cool things sharing with AI and dentistry. I'm going to find his Facebook group and give it a shout out. So the Facebook group is dental disrupt AI and automation currently 550 members. He's posting daily uses of AI and dentistry and beyond. So massive shout out to you, Dr Jason Lipscomb. Thank you for creating that resource. And I'm just passing on to you guys wherever you're watching or listening to this, just scroll down. You'll see it. It's in my show notes. So if you're on a progressive guidance, I'll put a clear link to this software. If you're on Spotify or we in the description, if you're on Apple, scroll down to description, there'll be a link there. And let me explain how this link works. So this is a custom GPT, which means that if you go inside chat GPT and then you click on the link, I'm going to give you, you kind of install this custom GPT in your chat GPT. So whenever you're having a chat with chat GPT on the side panel, you're going to see this tool. It's called dental disrupt smile simulator. And did I mention it's free? Like it's freaking amazing. I've actually been really impressed. In fact, I've been doing an experiment. I've been giving it before photos of actual real smile makeovers that I've carried out. And the simulated after image it gives you is pretty on point when you compare it to the real after. So it's not too far fetched. It's actually quite accurate. I've been really impressed. So it harnesses the technology that chat GPT has. But Jason's just done on the hard work to actually prompt it and give it some instructions for smile design. So once again, the link is freely available. Just scroll down wherever you get your podcast. And I'll make sure the link is there. If you hope you enjoyed that tool, you're welcome. And now let's join the main episode and I'll catch you in the outro. Dr. Daz Casper, welcome to the professional podcast, my friend. A long-term producer, I've even seen from the Facebook days, right? Well, back when we had a Facebook group and stuff, you have made so many protrusion writing happy with your cross polarizing filters that you made for them. And it's so nice to actually see you in the well, not in the flesh, but see you're in front of me discussing something that's like changed my world. And I know it has yours. Well, Daz, tell us about yourself. Why you are now studying Endo at Liverpool. And like a little bit about your journey, because I always believe in the journey and how you got here and how now you're doing like research in AI in Endo, right? Exactly. Yeah. First of all, thank you so much. Thank you so much, Jazz, for having me on. I've been a long fan of your podcast. And I still remember back in the day when you started at the beginning, you know, when you did that first episode on Singapore, I think. And you heard that. Yeah. I remember you were doing like the intro of that American guy, you know, so cool. I just remember listening to you. But I graduated from Sheffield. So I graduated from Sheffield in 2018. And I think at some point you were DCT there. So I think I might have like senior round. And when I was a student. But yeah. And so right now I'm doing my doctor in the non-techs University of Liverpool and something. Can I ask what that actually means? Like, for example, I'm familiar with Masters. I'm familiar with an M-clin dent. How is the doctor different? Like, is it significantly different? So it's still a specialty training pathway, but we've got a lot more research. So a big part of our course is a research component. So. Like the other level? Yes, precisely. Yeah. Yeah. So I'm also an ACF, which stands for academic clinical fellows. So I'm doing my end of sort of specialty training with the doctorate concurrently. So I'm also a big part of my specialty training. So it's also like a big research project. So everyone who's doing the DDSC, you know, we have like a proper, you know, long, three year long research project that requires a full dissertation, thesis, defense, etc. It's a very big commitment, but I have, honestly, so happy at Liverpool. I have the best supervisors like my professor, Professor Jarrod, shout out to her. Oh, legend. Yeah. He taught my, to my wife back in the day as well when she was a Liverpool manager. He's very well known. Yeah. Yeah. And honestly, it's such a nice place to be and extremely supportive. So I'm really, really happy. And yeah, and a lot of my research is in AI and also augmented reality as well. So I'm looking at that as well. So I'm sort of looking at all the digital technologies and how we can integrate them into. And so. Yeah, really such an exciting space because literally every week there's new things that come out and you know, you can go on holiday for a week and I feel it's like once like a few hours to the G catch up for a missed on so it's incredibly exciting to be exciting that you're at the interface of end of the ontics and AI, right. That is so exciting. I can, I can hear the enthusiasm and your voice and that, you know, that makes me so happy. Look, AI has just changed our world completely, right. And it's an interesting thing. Why I actually remember something interesting from when I was 12 years old, this year, I was like year seven. And I remember reading something in IT class, right. There was like the first chapter in the book back in the day and using like these big old desktop, e-machines, Dell and that kind of stuff, right. And I had this chapter about the disadvantages of having the internet and one of the disadvantages of having the internet is that it creates a big divide in the world of the communities and a percentage of the population that is information rich and percentage of population that's now going to be information poor and how the information rich will really be massively like advantage over and it creates this like inequality. We are at the same space now where most of the world has now caught up. Okay, it's internet access, especially now things are Tesla and what they're doing the styling and stuff is it's not pretty much anywhere now. But it's now who's using AI and what they do and who's not. And you can see that in the next few years, the AI is like getting so crazy. And the time I'll take people who are behind to catch up is just like you said, I have been like, AI for me has allowed me to come home to my children, has saved me so much time in the podcasting world and what we do now, like we obviously launched Ask Jazz and made everything accessible. Like before, someone used to mess with me years ago before the AI thing took off and says, "Jazz, and this was when we were like 100 episodes in." And it's like, "Jazz, your podcasts are so rich information." You know, you should create a resource that can e-book or something with all the stuff in it, basically. But now with AI, it was amazing to be able to just bring it together so you can chat with every single piece of content ever. And this is where just barely scratching the service, right? I just want to share that perspective and excitement. So tell us where should we start? Where should we start? Where do you think we are now in terms of AI in Dentistry? I think that's a good overarching place to start. Yeah, so I think you're absolutely nailed it. I think we're still in this infancy days because I mean, AI, I mean, chat-gipoties have been out maybe for just over three years. You know, it's become mainstream. And every week, like I said, you know, have things that are improving. And because all these big companies like Google, Cultura, and Tropic, OpenAI, they all competing with each other constantly. So they're all pushing themselves to get better and better and better. And obviously as consumers, we benefit from that because it's a healthy competition because everyone wants them to use their AI rather than the competition. So this is, I feel like that's why we're pushing so hard with the models that are becoming smarter, they're becoming easier to work with, they're becoming more intuitive. And I see you with every iteration of every new model, a little new answers that, you know, you might have not picked up if you haven't used it enough that are being ironed out behind the scenes. And I think with more and more people adopting it is just going to be mainstream. I think, you know, it's going to be like the internet. Exactly like you said, it's just going to get better and better and it's just going to be more integrated. And like you said, if you're not going to be using AI, you'll just be left behind because everyone else who has that advantage of basically having the expert in your pocket on anything and is precisely tailored to you, I think that's the key thing. It's like having a tutor that's omnipresent and you know, they have expert knowledge. And that's what's so amazing about it. And I think it's only getting started. I think, you know, there are specific dental AI's in terms of like, care's diagnosis, there are research done, there's segmentation, CBCT. There's a lot of stuff done in medicine, but they help clinicians dying those patients. So that's, you know, there's a very exciting space. And these are old tailored models that are trained specifically on very specific data, very specific instructions. But one of the things. You need an example, Daz. I last year, around this year, this time last year, I had a pneumothorax, a tenious pneumothorax, right? My left lung decided I wanted to burst. And the doctor had no idea it was going on. And she's like, maybe we should send you for a CT scan. I don't think it's anything, but we'll send you, but we know it was a chest x-ray, sorry, send you for chest x-ray. And then the chest x-ray popped up. And within a second, it said pneumothorax. And it highlighted that the diagram came up of the area of the pneumothorax. And that was amazing to see. A bit of use in medicine real time on the very tailored, specific stuff. And that's how it's revolutionized diagnosis. And I think that's really where dentistry, where does dentistry start with AI? Well, it started with diagnosis, right? And that's really, because diagnosis is, once you've gathered the data, right, which we have to do as clinicians anyway, it's allowing us to eliminate our blind spots and actually tap into multiple, you know, language models and parts of our brain that we just don't exist with us yet. We're not intelligent enough, but we can now give it to AI to give us a differential diagnosis and we use a human judgment. So no wonder that diagnosis is really taking the lead. So tell us about what you've seen as good case studies at the moment and what is potentially not there yet, but is something to be excited for the future when it comes to diagnosis and AI and dentistry. So I think in terms of diagnosis, it's like exactly like you said, it's you have an expert for a second opinion, right? Because ultimately, you're always going to be responsible for your clinical judgment. You know, AI doesn't have a GDC number, but you do. It's always about, you know, having a different perspective because a lot of patients will give us a lot of information from the history, from the clinical findings, you know, signs and symptoms. And we get those patients that we're not quite sure sometimes what's going on. And you know, you order a different scan, you have a CBCT, you may get them to see another specialist and it's exactly for that purpose. So you can even feed information like patients, signs and symptoms anonymously, but I think that's the topic that we also need to touch upon as well. You know, you can describe the signs and symptoms, your clinical findings, and then you can get AI because you know, it has all the knowledge of the internet has been trained on the older knowledge, all the websites, books, et cetera, right? So as humans, we don't have really the capacity to have so much knowledge available to us. And it just gives you that ability to really hone in and maybe get something to this persona really to tell you, okay, maybe this could be. And then you're the final sort of boss that's connected altogether and ultimately you make that decision. But I think, you know, in terms of like diagnosis, I think that's really, really crucial because there may be something that the patient could have told you that you could have missed and completed this missed, but that might be very, very specific. That was mentioned somewhere else that you're not familiar with. So I think that's the key thing here right now. The danger for me with AI and let's say, let's use chat-y-p-t-is-a-case-use, right? I've done it before, is that, okay, obviously you don't want to feed any patient data in. We're going to talk about that, but let's say, if there is a scenario where a patient has this feature, a sinus tract, a blah, blah, blah, what's a percentage chance that we should save it, what's a prognosis, blah, blah, you feed it some more data, you even read a graph, it reads a read graph, for example, or an example read a graph, like, it gives you information, but my worry is when I've tried to discuss with chat-y-p-t about various other aspects and said, give me the evidence for this, give me the evidence for that, it literally made up papers that don't exist, right? And it hallucinates, right? It's called hallucination. And so that's one of the reasons why I created R's Jazz, because I said, if I don't know something, it was like, we don't know. And so only use of the information that's actually been verified from the mouth of me or an expert, that goes there, rather than scraping from Reddit and hallucinating. And that's one of the reasons why I was worried about it, because a lot of it depends on the prompt. Now, we mean you, we know what prompt means, but some dentists and colleagues I speak to, one of them who messaged me on a protruding guidance and says, look, he said, I'm 30 years qualified, I feel like a dinosaur, I feel like I don't even know where to start with AI, right? So someone like that who's listening into this conversation, we need to even say what is a prompt. A prompt is that what you feed the AI to help you unpack the problem. And depending on what prompt you give, you'll get wildly different answers, right? And so that itself is like biasing the AI in different ways. So for example, the way patients do this is, patients would type in say, hey, I think my dentist has messed up my diagnosis, I think I have this, this, and this. And the tendency of the AI, because that's the way you've presented the problem, it will look for reasons to support you. And so that is a big issue. So what advice do you have for a dentist when they are starting to use chat TFT or any other models? And I'd love to hear what models that you use daily to help them to make better judgment. Yeah, so I think one of the key issues that I found with when people come to ask me about AI or when come to me when sort of ask me about troubleshooting is the fact that they are not customizing their model for them. So when you sign up to chat GPT, when you sign up to Google Gemini, Cloud, et cetera, there is an option in the settings, which is the personalization. And there you can put information about yourself. So what you do, what you do for a living, what you're interested in, where, where, where, G-lib. So basically, in my personalization, I told that, this is my name, I'm doing my Dondick training here at this university, under this professor. I'm interested in AI, I'm interested in technology. And these are like my preferences. So I tell it a lot about myself and a lot about my preferences. So any time I'm having a conversation with AI, it already knows all this information about me. So because let's say you're asking a dental question about AI. AI doesn't know. Are you a patient? Are you a child? Are you a dentist? Are you a specialist? You could be anyone. So you get a very generic answer back. And that's, I think that's the key thing that people sort of struggle with, because they'll try it, I tried this rubbish, it doesn't really work. Oh, it gave me some garbage out. And exactly how you said, the first key thing is always like log in and use your own account. Don't share accounts with other people, because at the moment it doesn't really work, you only sort of, you have one account per person. And personalize it with your own personal instructions. And once you've done that, you dramatically see an improvement. Now, second thing I would say is after you've done this, is, so from my prompting, like the basic thing, I would say is, I have this little acronym, RCT. So I've broken out treatment. So obviously I love those. It has to be underrelated. So R is the role. What do you want the role of the AI to be? Are they a fellow clinician? Are they fellow dentists, a specialist? Are they an admin? Are they an emo writer? And the C is the context. Context is key. And this is the biggest thing that people miss out on, because now, AI's get better and better with worse and worse prompts as the models improve. And I feel people sort of, context to me is like a lab doc that they send to your laboratory when you're trying to make an indirect restoration of a crown, right? So if you just sent them, say to them, make me a crown. They don't know what to do for this, what material, what shades, what characteristics, what's the occlusion? is you're going to get a generic crown back and then you get a generic crown back and you're like, well, what is this? This is not what I want because any time that's a person, I love that. Thank you. When you think of, you know, when you enter the chat box, you always have a model answer that you always want back. And then when you have that expectation to get that model answer and you get something else back, you just want, well, this is rubbish, you know, but the context is key. So always give AI a lot more information than you think it needs. And the last thing is the task. So exactly what does it need to do? Does it need to give me some brainstorming to the diagnoses? Does it need me to write an email? Do I want it to help me brainstorm some ideas because I'm traveling to a conference somewhere? So that's our CT. That's, you know, that's the main thing. But with C being the most important thing because the context is everything, especially now with the models being more and more smarter, but nothing can replace context, you know, I think in the future will AI will just be integrated in our lives, whether it will be through like wearable devices or glasses or anything like that. So you automatically get the context fed into it with Leslie. So it'll know from your day what you've done, who you've met, who you've spoken to, what you've bought on your phone. So when you say it's like, oh, right, an email to Jamie already know who Jamie is and who, who, where they're working, you know, and so at the moment, still, I think you really should get the most out of AI. You have to spend a little bit of time customizing it and making it tailored to you. But once you've done that, the results, you know, the skyrocket in terms of the outputs they actually get. In terms of, I mean, you made a wonderful summary. So role context task. I love that so much that that's a real gem already. Even the role like prompt engineering is like act like a specialist act like Jackie Chan. Tell me this, whatever, right? It's like the role you're playing. I love that so much as well. That's one of the key ones I use. Like, you know, well, now I love speaking, by the way, to chat to you, Peter, I love having like a proper chat as I'm driving. It's like a phone call. I love that feature about it. Right? So that's awesome. Yeah. With Asja as I sometimes talk with myself is that in this episode, what was the gem that was shared by this dentist? I love just hearing the voice back and how you could do it in different languages, which is just crazy, right? The AI can speak or Chinese and I can't and I love that as well. But when we come to tasks and administrative tasks, I think that in density, there will and I, either whether we want to hear this or not, we are getting close to a point where we will need less receptionists. We will need less admin people because now we have AI agents that can do so much. For example, AI agents work for you will be able to expand as well, but they can work 24/7. They are literally the person, the AI agent that will read the email from the patient, right? Attribute them, help them with an appointment, book the appointment SOE already for them, send them the email for the welcome, check in with them a few days before, send them the treatment plan, all these things that a human is doing at the moment. I think we're two years away from AI agents just doing this. And so this is going to massively shape the workforce of the future and it starts with density admin staff as well. What do you think about that? Do you think I'm wrong? Do you think I'm, do you think it's scary? Thought for sure. But what do you think about my prediction? No, I agree with you. I think the agentic workflows and the way actually the latest chat GPT 5.2 models trained on is actually amazing at agentic tasks. In the chat GPT especially, you've got the agent mode, which is the agentic mode and then you can also use the thinking mode, the 5.2, which is the reasoning model. But the reasoning model, the recent reason model that was actually released by chat GPT is really, really good at agentic tasks. And I think that's what the industry is going because all these repetitive, boring administrative tasks that can take a lot of time, but also just very time consuming, it can be easy replaced by agents. So I'll give you an example of one of the things that I've done recently with GPT 5.2 thinking in the agent mode. So I had a lot of CPD certificates. And you've got, there's like a GDC sort of template, they have to fill in with all your certificates, where you went, how many hours. So basically I've uploaded all the certificates, I've uploaded my Excel spreadsheets and I said to it, read all the certificates and copy all the information from the certificates, from the hours, all the objectives, you know, and put it all to go in the spreadsheet. And it basically just did it and then I can just go and do something else. But previously you either have to hire someone to do this or you would essentially have to do it yourself and waste a lot of time. So I think now with the agentic workflows becoming more and more prevalent, I think it's, I think absolutely right, is getting more and more, you're going to get more and more personalized advice, right? So for every patient, every patient, I think that's why air is brilliant because you can make personalized consent forms, you can make personalized recommendations or explanations. So for example, let's say I have a patient with an external survival route resorption and I need to sort of explain that to them. Well, if they're like a mechanical engineer, I can say like, okay, well, imagine you've got a car door and then there's a bit of rust coming from the bottom of the car door and you have to sort of file it down and fill it because before it gets worse. So it can come up with these amazing strategies that I think I would never be able to think about, you know, and it can, it can really make personalized medicines very, very accessible. And I think it will make an experience much better for patients and for dentists because it will be less effort for the dentists and there will be much higher output quality for the patient because the patient will have something that they can really, really understand. And then they can really sort of, you know, I want to always want to get better communication, but I'm not, I'm never going to be the best communicator in the world. You know, AI can close that gap. AI can make you, it's 100 times 100% better than what you're already at. And that's how I see it is like an extension of yourself. So it just amplifies your skills and then sort of fills in any sort of things that you may not necessarily have the ability to do at the moment. Dan, I'm going to add to that. So one of the, this month is AI month on the podcast and one of the guests that we have this month from the States, this is exactly how he uses AI. Like he will, I don't know if he's actually putting the audio in or the transcript in. He's putting the transcript of the entire conversation with the patient in, right? And analyzing it. And then eventually he'll give the outcome, okay, well, patient went ahead with treatment or not, or the patient said this. And it kind of audits, okay, well, actually your sentiment score like, okay, well, actually, I'll give you seven out of 10 for this next time when the patient said this, you could have said this because the patient was trying to tell you this, right? And so it's amazing how you can use that data to improve yourself in communication. It's just when you think of it that way. But I'm just going to pause a moment because we can't gloss over this wonderful tip that you gave us whereby certificates are annoying thing, right? CPD certificates. Yes, we're getting less and less paper ones, which is good because the more PDF ones we have, the way to do it is you stick it all in a Google drive, right? You can allow your chat GPT to have access to a Google drive or Claude, whatever, right? Because Claude's agent stuff is pretty up and coming good as well. And then exactly it can translate all that information. And then he could say, okay, make me a PDP to summarize everything that I've learned so far this year and leave a space for me to decide what I want to do next year. So the whole thing about having to manually input, like one of my friends Alex, he was saying to me that he logs in his website, he types in the title of every single lecture he's went to, he types in whatever, like that's BS. Even the AI agent can just fill that page for you, right? So if you want to still use that website, whatever he's using and fill it for you can. But we've come to a stage now where AI, like the only limit is our imagination. So now guys, you don't have to imagine anymore. All those purchases, certificates, all the certificates you're getting, right? In fact, one thing I can even, it'd be so easy for me to do this, right? I can make on chat GPT a custom GPT that you just up, give it your Google Drive folder. It will make exactly for you. Because behind the scene, I say, when the dentist uploads their Google Drive folder, do this, give them this output and that's their PDP. And it's going to be, it can be free. Because it takes me minutes to make and it just saves time for everyone. So this is the reality that we're in now. Like you can, the only limit is your imagination. Absolutely. And I think, you know, a key thing is how you set this personalization with like custom GPTs, but also projects. So projects are big things. So chat GPT projects, Gemini, Gems, the Cloud Artifacts, they're all the same thing. They're basically a little subsets like imagine like you're building like a little folder of a very specific task. So anytime I'm doing a big task and I need to sort of talk to without having to give the same context and background every time I'm having a new conversation in this space. So for example, I've got one for emails, right? So I can, I know exactly, you know, every time I dictate my emails and it was exactly what style I do my emails. And so it sounds like me, British English, it's exactly the way that I sort of finish my emails. I want my emails to be professional and polite and it basically matches my style. If I was to do my email dictation every time for every email reply, I'll have to give it the same instructions again and again and again, but with a project, you save time because on top of your custom instructions, the personalization instructions that you have in a normal chat in the project you've got project specific instructions. So that's really, really good for repeatable tasks. For example, if I, if I'm doing my like, for example, like a consent form or if I can do like a, if I'm doing my end of treatment notes, they're always going to be the same. So, you know, it's basically I can just quickly dictate all my information there, put all my work units in. If there's any things that that's relevant and then from the same project, I can literally have a full treatment that I've completed. I can then generate a full summary to like the dentist. you know, that I'm referring to and I can make a social media poster if I wanted to. And it's all in the same project. So because it all has, it has all the context, it's minimal amount of effort that I have to put in then to get a really high quality output. The only caveat is. Right, better than what you did on manually and like a hundred times faster. Absolutely, absolutely. And because it's the only caveat to it is just like you have to spend a tiny little bit of time at the beginning to set it up. But once you've set it up, Chad GPC is really, really good at following all the instructions. So you're just recycling the same thing and it already knows what you're talking about. So, you know, it's then minimum amount of information, minimum amount of prompting that have to give it to have a very, very high quality output. It's because it knows all these instructors in the background and it will give you that high quality answer. And again, I think one of the key sort of things when it comes to interactions with AI, the majority of people, they type, you know, and I think that's a massive sort of mistake. I think the key is transcription and voice transcription. Chad GPC is out of sort of. So at the moment, Chad GPC and Gemini are the two sort of main ones that have the voice transcription enabled. But in my experience, Chad GPC's transcription is way better. It's just because you can talk for longer and the quality is just higher. And, you know, you can talk probably four or five times faster than you can type and you can put so much more information and it just doesn't take as much effort, you know, it can just be mumbling around. And this is where that you give the AI that context. And first of all, it saves you so much time because I tend to, you know, people say to me like, "By the time I've typed it to Chad, I would have written it myself." And I'm like, "Well, that's the key mistake because if you can voice transcription, if you can voice type it, it's just so much faster." And you get so much higher the quality output because you can put so much more information in that Chad to really sort of get the most out of the model, really. So it's a good, the best response. Haram Sennagreet, I also want to add an ergonomic reason as well. Like, why are we looking down our phones? We need to be just speaking, right? And so this is a kind of an ergonomic reason, right? There's no reason to be typing. I do so many voice notes anyway. And so with the AI, like, are you using, Dan, at the moment, I don't know if you're still working on practice part-time while doing your DDS at the moment? I am working practice as well, yeah. Okay. Are you using AI notes software? I know using ChatGbT in a way, but are you using a dedicated dental AI note software? So I use a customized version of ChatGbT. So I've made my own sort of AI voice sort of notes software, but yeah, I do my notes with AI. But I know this high D. There's Heidi, there's Digital TCO, there's a Density Dashboard, even Plaude, you know, the Plaude AI things, right? So it's got a dental note taking mode on there, right? So hopefully this month, if I got my ducks in a row, I'll be posting the video where I'm going to compare the my biggest review yet, right? So I'm going to feed the same consultation into all these softwares and see, okay, who makes the best notes, the fastest, the best letters, but just whatever software you're using, the amount of time it's saved us. So one of the orthodontist colleague Stuart Yuton, onark, a protrusion guidance community, he loves it because the ability to write back to your referring dentist and having these templates, it's called Molymail on Digital TCO. Like the other day, just yesterday, I was having a telephone consultation with my patient, right? So I'm following up with her, saving her a visit, I get patients coming from all over for TMD stuff. So she wanted a telephone consultation and I put her on loudspeaker, I said, and all my patients know that everything's been recorded, they all consent to it, no one's ever said no. And so I had her on loudspeaker and the whole note just wrote themselves and you just had to read and check it. It's just a little time saving us that that mean that we get so much more done. And if you, I just think that for those dentists were feeling stressed by the notes, not only are their notes lower in quality, right? But they're having to spend so much time as well. One objection I received from a young dentist about AI notes was, well, what if you forget to warn them about something, then it's not going to, so compared to templates. So your template, for example, for an extraction, has got everything. Whether you warn them of a sinus, a breach or not, it's in there, right? And so they said to me, well, what if you forget to say it, right? And I'm like, how is it fair? Right? How is it fair that we're allowed to be the template? You didn't say it, but the other way around isn't, right? And so we just have to, it's a reminder for us to be checklisted, right? And so if we go a checklist manifesto and we always want, it's making us better clinicians, better communicators and patients are better. Now in these AI software, it's your TCO specifically when there's something called a super dentist cheap mode, right? And Lichy is like, it's like what hospital consultants do. They say, I saw Mrs. Myths today, she had a mouth opening of 39 millimeters. And as someone there is typing in the back end, writing the letter within a few years, that's going to be gone. All those people unfortunately don't lose their jobs. It's sad reality, okay? So that's going to be replaced by AI if it hasn't been already. But the ability for me to literally for a few minutes be like, bust a rhymes and say things so fast, right? In like 30 seconds, I've written a whole letter and it has captured everything beautifully, right? Yeah. And it just, it's a massive time safe. Now I want to show you, and for those listening, and I want to show you something cool, because you mentioned custom projects and stuff, there's a dentist called Jason Lipscomb in the US. He is a podcast in US, awesome dentist. He's got like these free groups online that he gives away so much, so massive shout out to him. Let me show you this cool thing that he made. And then for those listening, let me describe what I'm doing. So I'm going to now open up my chat GPT. Let me show you. So let me go to my chat GPT. And my team will blur out my stuff so you don't get to see what I'm saying with chat GPT. But why it's loading? Okay. And then I'm going to go to dental distrups smile simulator. Okay. So dental distrups smile simulator. This is his free thing he made. And basically, his Jason was saying that listen, stop paying all these companies to do your before and after. Right? But before and after like visualization, and there's so many other software's out there, like, hey, you upload a video, we'll do it. They're basically using AI anyway, right? To do it for you. So what are you saying? Hey, listen, this is for free. And took me a minute to make. So you upload a smile. And then it will give you the data. So let's do it. Right. So guys, what I'm doing? I've got my chat GPT loaded. I'm going to load a photo of my sister before she had author. Right. I'm just going to load that up now. So I'm going to find it. Here's my sister before she had author. I'm going to add it in and a smile photo. And then let's see how long it takes. So clock is ticking is creating the image. And by the way, chat GPT's image creation is just phenomenal. Right. For someone who's been like you, you two being in a thumbnail thing, it's been a revelation for us to be able to just save time and effort. And it's so much fun. Right. So much fun to be able to work with Gemini and chat GPT image creation. So I put my sister's image in is still having a thinking is still creating it. But it's creating it based on what Jason has trained this project on, which is when someone upload an image, follow these parameters of smile design to create a realistic image that we can show to a patient. So now it's easy and ever. So it's now starting to load it up. How do you use these things before? I have. Yes. So, you know, with custom GPT's, you're uploading your old custom data sets of things that you want it to reference as its knowledge. And then you've got your own sort of master structures in the back. Yes. I've made like my own endo one as well. What does your endo one do? Like does it assess for complexity or what does it what does your endo one do? And so the endo one is just basically, you know, it's kind of like a reference guide. So if I want to ask it about a specific type of a very, very complex case, you know, since I've referenced all the sources that I've uploaded. So it's a small reference, like a second opinion, like another condition. Very special. Very cool. So here we are before and after image. I mean, how realistic does that look? Look, looks amazing. Right? The funny thing now is it. Obviously, those listening, they won't get to see this. But I'm going to now load up her actual before and after. Right? And then we can actually compare. Was it was it actually close or not? Right? So let's put this here. That's so not not bad, right? Not bad. Obviously flashing stuff is different. But what a wonderful, almost identical result that we managed to achieve. The AI is very clever, very good. Obviously, we have to use our judgment. We don't want to like over promise and stuff. But we knew that already from when we're using these simulation softwares. But by next year, the video stuff will be free as well. Right? It's so easy. I'm sure probably there's a mode in there that I can just upload a video someone speaking and on send a video back to me. Grok can probably do it to be honest with you. Grok AI. So which AI softwares should we be using personally and then professionally? Right? So give us some personal things that you have found good case studies for AI that that just because we're dentists, we can still benefit from because we're humans too. And so we can actually benefit from these AI case studies. Yeah. So I think what I would say, if if any of us are sort of starting with AI, or if they want to sort of get to do and get more use out of AI is to try the different models first. So you know, there's obviously the chat GPC, which is the most sort of famous, most popular one. You've got Gemini from Google, this Claude from Antropic as well, and then Grok from XAI. And there's also other ones like there's Metastuff and Chinese models like when and deep seek. But sort of the main four, like the main three or four ones will be like the open AI chats Gemini, Claude and XAI and Grok. Now try different one and then see which personalities and which type of responses you like. So again, get your custom instructions and get you know, introduce introduce yourself to the AI. So there's exactly who you are who's dealing with. So you can get a high quality responses and then see which personality you like and then just stick to that one. Because especially now with a charge GPC and Gemini, there's a lot on more emphasis on memory. So as you're using these models, they'll learn more and more about you. And now with Chad Jeperti, they've introduced not only just the memories that you tell it to remember. So you might have a chat, for example, in your dental practice, there may be an nurse who's got a certain name, a practice manager's got a certain name, an associate's got a certain name, and you can basically introduce all these people to chat or to Gemini, and then it will remember. So then every time you say like, oh, Jason, my associate will write an email to him. And there's exactly what it is. For example, I AI knows exactly what on things of all my supervisors, what are names of my professors. So if I just say like, oh, write an email to prof, and there's exactly what prof I'm talking about. And so I'd have to give it that extra context. So just again, it just makes it faster and say small time. And it will say it in your voice as well, because you've trained it on, this is how I write my emails, this is me, and then you get to check it. And it's not like taking a. It is a shortcut, but it's not like, I don't think it's disingenuous because you've literally trained it to sound like you, and you approve it, and then you copy and paste it over. And this is your words, basically. Yeah, yeah, exactly. And I think, you know, so once you sort of start learning about it, just pick the one and sort of stick to it, because then learn more and more about you. In terms of what things, what certain AI is a better, I would say Gemini is definitely better if you are really in Google ecosystem, so you've got integration with your Gmail, Drive, so any sort of. It's got the biggest context window out of all the models that the moment just got one million tokens. So Gemini, if you're working with really, really big documents, and there's the only AI at the moment from the West and side that can accept the video uploads as well. So you can, for example, upload your video, you do a menu practice marketing video, and you say to it, have a look at this video that we're making for practice marketing, and you know, score it, and tell us what we can improve, or is it good, is it bad, is there any sort of things that can be improved, and you know, Gemini is really, really good at that. And when it comes to image generation as well, Gemini has got the Nano Banana model, so the Nano Banana, Nano Banana Pro, is faster than chat at the moment, and I, I'll say, is better quality, in my opinion, compared to chat, but they are very close, but Gemini is just a little bit faster. So I think any sort of big context documents, photo generation, also video, you can use Vio, so that's their video, FlaxShift Generation model, which again, so I'll tell you a really interesting use case that I did the other day. So one of my, one of my friends who's a specialist, he is doing a course for, I think, FDs or GDPs on, on Endo, and he had a photo of a access cavity and a normal handpiece with a bar over the access cavity, compared to an autosonic. And obviously he wanted to showcase that when he used the normal handpiece, you don't really see much because the handpiece is in the way, when he's an autosonic, you can see a lot more. So he had these two photos next to each other and he said to me, I, oh, can you make a video out of it? So I went to Gemini and I used the Vio model and I basically got it to move the handpiece, got it to move the autosonic, so it just becomes, your photos become life, you know, and it just, you know, it's just so much nicer because it just makes it easy to communicate your idea. And, yes, I think Gemini again is also really good with maps because it's got all the Google maps stuff. So I was, I was buying a car and I had to go to like different dealerships and, you know, sometimes if you have to go to like different places, you sort of like go and Google maps and say, okay, if I go from this place to this place is like that, that much time, but if I go here, and then you just do all these calculations, maybe, should I go to this one first or this one first was the most efficient way. So I basically told this to Gemini and it just did all of that and then it did all this, all these calculations, which, you know, which dealerships to go first and then said, like, okay, if you live in this area, this time of the day, if you leave it this time, this is the best way. And any sort of Google staff map staff video, I'll say Gemini is, you know, is better, but I use chat most of the time. I think I would say is, you know, don't just use one model. I'll say use one concurrently, but you can also ask other models for their opinion as well. So if I'm doing something very big or if something that's that is very serious or I need a lot of opinions, I would literally go and I open chat, Gemini, called and Gork and I'll paste the exact same thing into all four and it's like a council, you know, I've got all the different opinions of all the different AI's and they all have different slightly different trainings, slightly different instructions because the exact same prompt, but then it gives me different opinions and you're like, oh, actually, you know, this is a good idea. So that's really, really useful in terms of, you know, if you're, if it's like a big task that we're using the reasoning models, the thinking models, so that's, you know, that's what I would definitely recommend. But, you know, I prefer chat, but, you know, some people might prefer, I know, for example, my parents prefer Gemini or some of my friends prefer Gemini, just because of the personal history that it has. So it just doesn't mean that you have to, you know, you have to use chat because everyone else uses chat, but I think that is just try them and then stick to one. It's a real rabbit hole you can go down. It's just, I think it's just one wonderful thing. I mean, there's so many dangers to it. We'll touch on in a second or talk about patient, confidentiality and stuff in a moment, but for me, I'm rubbish at DIY. I just never paid, my dad was never into DIY. I never go into DIY. And so anytime I have to do any sort of DIY, Ikea kind of stuff. I'll literally, but if I'm like confused, which happens very soon into a DIY task, which why don't do implants, if anyone ever wondered, I get at the voice mode with chat, GPT and suddenly just switch on my video. I was like, wait, which screwed up, does it need? And as you say, oh, the one over there that's labeled B. I was like, okay, cool, I got you. Like little things that will help you navigate life. Like it's now I don't need to pay someone or or or or recall someone. I can just get chat, GPT to give me that opinion. And it's just mind-blowingly amazing. So the case use for personal things is just your, you know, imaginations, your limits. It's a wonderful thing out there. But for dentistry, where are we at now in terms of, you know, in the States, things have to be hyper compliant, UK and your GDPR, that kind of stuff. Like, are there any rules of engagement in terms of sharing patient details? Because there might be someone the world right now who says, write me a letter to a patient. Here take this, here's a patient details, write it, and they would have given the date of birth, the name of the patient, address, etc. That kind of stuff. That's surely that's dangerous. Yeah. So exactly. So I think, you know, I would certainly say at this point, you want to any interaction that you'd have with a chatbot, you want to be, you know, completely anonymous. You don't want to be giving, and I mean, when is it really needed? How much more, you know, how much more high quality will you get out of the output just by putting the patient details in? I don't think it's going to make that much of a difference. But one thing you said earlier was the context, like, and the whole communication side. For example, if you're, if you want to really enhance your communication and the patient is a mechanic and you want to say, look, so you could say, use my template for my end-on-to-consent form, adapt it for a 15-year-old male who's a smoker and is a mechanic. I don't think that's, that's not patient-identifiable. It's not patient-identifiable. But it's good context to help you make a better letter. So I think that's okay, but as long as not patient identifiable, it should be fine. Yeah, absolutely. Yeah. So any patient, you know, date of birth, patient names, things like that, addresses, you know, I think at the moment it's a big now, you know, this, this, you can do it yourself and this really take that much time. You know, I think the quality of the AI comes from, you know, the background information rather than patient-specific information. So, you know, at any time I interact with any of the AI models, I never, ever would put any patient information in. And this is, you know, it's quite interesting that we're talking about this because like some time ago, there was, there was a, there was an issue where, you know, where you, when you've got your, your chat with chat GPT and there's like a share button. And there. So it could be. You're a rock, right? You're coming onto the rock issue, right? I think it was chat. I'm not sure. I'm sure I remember it was, it was so basically what ended up happening is the, it created like a public address of your chat. And then these chats started indexing in Google. So, so people were able to find people's private chats in Google. I mean, obviously they fixed it now, but you never know where an issue like that may happen again. A big data leak could happen against the deep search where it's called in China, for example, whatever. Like anyway, it'd rock AI could be affected. And then suddenly they got all this data. And so I think very easy debate here. Don't over share. Give what you need for the context, but don't give any patient identifiable information. Now, here's a curve ball does. Here's a curve ball. You just saw me and those everyone who's listing right now, you heard me, upload my sister smile to chat GPT. I didn't ask my sister. She's cool, but I didn't ask my sister. So Jay, Jay lip, Jason lip, so made this wonderful thing out there. And so at what point do we need to do? We need to say to our patients, look, I'm going to be to help you to visualize this. I need to share your image, your data to chat GPT. And so where, where do we draw the line in terms of your patient's image? Or is it okay if it's not the eyes? Like, do you know anything about this? Because I don't, I don't know. To be honest with you, I'm not sure. I think it's always better to err on the side of caution, you know, in terms of, you know, I feel like as long as we follow the confident shallots, the patient will confirm the shallots principles, we should be, I think that's the key. And, you know, I think in terms of. I do way does is that if you were to upload it to a model to give you an AI simulation, I think the best way to do it is ask a patient, look, I can give you a really cool simulation. It's going to, is mind blowing. And I'll tell you if it's realistic or not, but it's really cool thing. I just need your permission to share this with AI. AI keeps it to itself, but I need to be able to share this image to chat GPT or whatever, dismiss this model. And as long as a patient gives you consent, then that's fine. But I think where we can get trouble is if we don't get that level of consent. Absolutely. Yeah, and everything in dentistry is to do with consent. You have to get consent for treatment. And I'm sure patients understand the fact that we're doing this to help them. We're not doing this for our gain. We're only really doing this to give them a better understanding of the things. We obviously, to us, it's evident and very easy to visualize what the outcome could be. But for them, it may be a little bit more tricky. And if it's just another tool, right, is that can actually, is just another tool that we're using. So it's just another way of being able to convey that data in a way that's very patient specific. And then again, you're the judge at the end. You are the one that says, okay, this is actually because this is quite realistic. Or you can be like, oh, actually just ignore it. You know, this is never going to happen. I think at the end of the day, it's all, you're ultimately always responsible for the information that is given to the patients. And you know, you cannot blame the AI for sort of saying to you, like, okay, it told me to do this. And I did this or, you know, it just wrote this consent form when I never read it or in the root, my notes and I never read it because you always have to be like responsible for everything that goes with your name. So I think that's the key thing. And yeah, I think I think you're absolutely right. It's just that it's all about patient consent at the end of the day and ensuring that we're on the same, we have both the same understanding of our patients in terms of what's the best way of being, you know, communicating with them. Let's get our crystal balls out, right? What will AI replace in dentistry in the foreseeable future? Admin team will reduce massively. I'm not saying it'll be fully replaced. You still need someone to oversee it because the child you really can make, or AI can make mistakes, you need to check someone to quality control, right? Once we've, you know, in my, in my business and education and protrusive, we've massively used AI, I haven't let anyone go. Got team of eight or nine now and I haven't let anyone go. I just help them to use AI so we can achieve more and serve more dentists and stuff when the RCI is et cetera. We've all been part of that rather than me getting rid of someone. I wouldn't want to do that. But as clinics want to be profitable, now they can get rid of a, not I'm getting rid of such a harsh term, but they won't need so many written. They have a team of six receptionists. They might only need two now, right? Which is, which is, which is good and bad because a very sad thing about workplace reality. So I think admin tasks and admin team will reduce my, my prediction. AI notes, we're pretty much already there and I think the vast majority will be using AI for writing letters, the notes, communication, that kind of stuff. What else do you think it will replace and then we'll come on to what will AI not replace in dentistry? So I think, again, like, I think education is quite big, you know, and tools, just sort of, so I know you mentioned the context of education. I think as dentists, especially, you know, Patricia Arty, we want to get better. We want to improve. We want to learn. We go to courses, we want to revise, we want to study. And I think personalised education is a big thing as well. And as like you said, you know, with your custom AI, you've got your own custom knowledge and you are able to essentially, as almost give that, that, you know, like a copy, like it, of yourself to mentor that dentist. It's a digital mind. They call it your digital mind. Yeah. Absolutely. And with tools like notebook, where you can, you know, upload your PDFs, papers, they can turn it into a podcast or presentation or infographic. It makes it so much easier to revise and really, you know, to consume knowledge and to learn. So I think this is going to have a big effect on teaching and training, you know, teaching training, stuff training, because different people in different ways, you know, one person really likes to read both someone else likes to do a little sense of podcast or watch a video. And with, with AI being able to deliver all these things and then deliver personalised quizzes, you know, tools like notebook, I think just going to be incredibly useful. They think, think of us, imagine we were students again, like the students of today have no idea how much information they have access to. Like all these concepts that literally, so many times I've had a chat with charity and said, look, explain to me like I'm 12, what this paper's trying to say, right? Because there's some very complex and bruxism, this, they're really complex papers, all these neurotransmitters and stuff. And I've, and it's helped me out so many times. I will think of the bruxism like a computer. And when it goes on standby mode and like, okay, that makes sense now. And so for education, perspective, even for self education, it helps so much. And the notebook, LM and very no worthy mention, the students of today, you know, we used to do like flashcards and past question quizzes. And we now have access to, like you can literally tell a German IDK, here's a few past papers from a few years ago, create five, create me a bank of 500 questions in a separate document, create me die, your answers that I can reference after I've given it a go. Wow. Yeah, it's incredible. And I mean, that was actually what I was talking about. And so I just went to conference in Paris for the ESC European Society of Endosontics. And, you know, we were talking about notebook LM and, you know, we actually use it in Liverpool for our journal club. So every week we have to read papers. So we, you know, we've introduced the podcast as part of the journal club. So you just listen to the podcast. And then you're already, you're already primed. So by the time we actually go and read the paper, it's almost more information is already available to you. And you're not sort of falling asleep when we're trying to read it. So I think plus in our busy life, sometimes as a parent, I don't get time to read. I do audiobook stuff, right? So now if you're part of a journal club, right, and you're struggling to read because of kids or life or whatever, but you have a half an hour commute, wow, you can literally just group notebook LM will make your podcast summary of that paper. You can listen to it and be clued up and a lot of people prefer to learn that way anyway. Absolutely. And, you know, I think, you know, you mentioned the fact that some jobs may go, but I think it's just a transfer of skills. You know, I think there was the same, same worries when the internet came out saying like, I was going to take everyone's jobs, but look how many jobs were created because of the internet. And I think the exact same thing will happen over AI. The expertise will just shift and your roles will be created. So, you know, rather than doing boring admin reception work, you just be the prompts master. You know, you'll be, you'll be supervising the AI and making sure that the output is correct or troubleshooting, you know, when the different AI sucks to each other, different people sucks to each other. So, you know, especially with coding, AI is amazing at coding and it can teach you, you know, it can teach you how to code is the best teacher. So, I think over time, you know, I don't think jobs will necessarily lost, just the transfer of skills will happen and new roles will actually become available and will rise because of the higher demand of this type of information. So, and I think, again, patients always appreciate personalized care and personalized attention and, you know, you can have an AI child, well, that's personalized to the patient, but then, again, I think, so let's say if the patient can talk to it, then the AI can flag of this something wrong. You know, this patient was talking to me and then maybe like the soccer is in healing after the extraction. So then I could ping your practice and say like, okay, check on this patient because they're interacting with our child, but they're reporting some complications. If anything, I think it's just going to improve patient care further and just give a better overall experience to everyone really involved. Talking of experience and you also mentioned augmented reality earlier as well. One cool thing I'm doing at the moment is patients wearing these, what they call, extra glasses, extra pro-2s, where I've got right. They wear the extra glasses and I'm screen mirroring my loop camera to the glasses and I'm doing my checkup. So, I'm doing my checkup live, the patient's hearing and seeing, like, as I'm doing the mouth cancer, they're sticking their tongue out, they're seeing their tongue through the glasses, right? And this has been like, revelation in the last four or five months, me experimenting with this. I'm re-painering this and I'm just loving patients. Not a single patient has said that was terrible or I hated it. They've been blown away. They like, never want to have an exam ever again, whereby they don't get to see that level of insight, right? But the AI element is now, let's say the patient says, oh, you know what? Can you send me a video of that? That's totally cool. But previously, my team would spend hours editing my clinical videos, right? In the education. Now, within the software, there's AI. So every time you look away, it removes those bits and it keeps the bit where it's relevant and you can just package it up, send to a patient that hears your video of your checkup a bit like when you get your car serviced and they send you a video of your car and his story here was good. This is the future of dentistry, right? And it's got AI, it's got the glasses experience and I think this is only going to evolve. What I want to know from you now is what are the limits and what will AI not replace? Where do you think where we're actually, you know, AI can't replace an dentistry because AI can cannot do this and it won't do this. Yeah. So I think, I think absolutely right, it's just in terms of the amount of information that we can now transmit to the patient in an effortless way, really to educate and engage them with us as well, you know, because we can take the sort of dreaded boring task of going to the dentist to something that, you know, some patients will be very interested and then I think that's a really good way of building extra rapport with them. I recently went to a conference in Venice on digital dentistry and there was a company there from China which made a robot that can do implants and they have hundreds of these robots deployed across China and the robot can literally just place implants and they said that I think they've got like 30,000, 40,000 implants that have already placed with the robot and they have like robot implant clearance, which is crazy. Wow. Wow. So you know, and then they said like now the next thing that's coming is the robot will do the crown preps and everything. So you know, it's incredible how much technology is advancing in such a short period of time. But I think there's still a lot I think what AI will not replace is definitely the human connection that we have, the emotions behind, because AI is just an algorithm with probabilities in the back of it. It can say to you, "Okay, if it's a borderline, restoreable case, it will say like, "well, from the training data that I know, this should really be restoreable, "but then you, with your experience as a clinician, you may know of a way to actually predict "to be restored out to." And there's just having that empathy. AI can generate the most empathetic sounding letter, but there's always that additional human touch that I think people, when patients will always appreciate and staff, it's having that ability really to just have that additional layer of interaction with people that AI will never replace. And I think the artistry and dentistry, there's so much nuance in terms of how much how creative we can be. AI can sort of give you suggestions and recommendations, but I think we all know our skills, we know what we good at, what we're not so good at. And I think we can really sort of use AI to amplify the things that we're not very good at, and also give us additional edges to define all the gaps, but I think we always be experts in terms of patient communication and also the artistry behind it. So I think the creativity and the artistry, I think it's, I don't think it's going to be replaced, but I think with new relinqu coming out and I think over time, I think humans and AI will start to merge and then you have pure humans, half humans, half AI. It's like something from a sci-fi, you know, all the sci-fi stuff is coming to life really, you know, cyborg, Mac, how fast they think, does, tell us how we can reach out to learn more from you and also just tell us what how you, what are you actually studying with AI and Endo? What are you actually doing your thesis on? So one of the things that I'm looking at is I'm looking at off the shelf models. So so basically, tragedy, beauty, Gemini, so basically I'm learning how they work and I'm looking at EndoDontic Complex DSESPON. So I'm seeing how they work in terms of EndoDontic Complexity, you know, can they, because yes, we've got a lot of sort of custom trained models for analyzing carries, CBCTs, you know, all this stuff. But I'm trying to see whether these models with all this vast training data, whether we can actually pinpoint and and see whether we can tailor this to be used for everyone, because you know, if you have, if you need a separate subscription for everything else, obviously, you know, yes, the model is arguably better. But what I'm trying to look at is how can we use the psychology that is in your pocket and in my pocket to help us, you know, and how we can customize it to get the best outputs and we're having to have custom AIs for everything. So that's that's one of the main things down down down looking at. I imagine it's scarierly promising how good it will be at assessment based on image and a radiograph. And I'm sure it's scarierly good at giving you complexity. Yeah, so, you know, if there's still there's still some nuances and the things that I'm sort of working out is trying to sort of understand how their models think, how they perceive things. And then, you know, sort of tailoring these things in order to sort of get a better output. But, you know, that's that's that's sort of my AI. And then basically every every few weeks, I do like an AI update for everyone that's doing, you know, that's that I'm doing underway with all my professors and specialists. And I'm doing some like AI teaching sessions with the the union of Liverpool people. Yeah, it's super exciting and it's just it's it's really changing the world. What a time to be alive. Absolutely, you know, and yeah, I think it's just incredible how fast things are progressing and how much they're improving by. And yeah, I think I think it's is like you said, this is the future in it is it's only it's only going to it's only going to get better. Yeah. So, I mean, I'm on Instagram. How can we send some yes, how can we send you some love and thanks and follow what you get up to, man? And so on Instagram and those ads and so end up on the score, that's so if you want to reach out to me more than happy to help. And I think, you know, if people are interested in it, I think I'll be more than happy to sort of organize some things and and see if people are interested to learn more about AI. But yeah, that's sort of my main sort of hub. Yeah, so if anyone's interested, like, yeah, give me a shout out and that's thank you. Amazing. And I want to say thanks so much for being part of the community. Thanks for all those put your us, put your respect to you serving the past with the cross-paralyzing filters. You're a man of many talents, I have to say, now endo as well and AI and mate, thank you so much for coming on. Maybe in a couple of years time, we'll do an update and how everything we said today was totally irrelevant because it moved so fast that everything was like, laughably irrelevant, no time to come to that. Yeah, it's changing so fast, I won't even be surprised. But yeah, thank you so much, Jasper, having me on. It's amazing. It's a massive honor to be here and I love what you're doing and your team and everyone's amazing. So yeah, excited to see what's going to happen in the future. And yeah, thank you again. Yeah, all the best with your studies, my friend. Thank you, all the best. All right, take care. Well, there we have it guys. Thank you so much for listening all the way to the end. You got through this episode with Daz and wasn't he great sharing all his enthusiasm and ideas and his geekiness. I'm hoping you're feeling a little bit more clued up about AI and hopefully a little bit excited. So at the time of recording, it is AI month on the app right now. And hey, if you're not yet on Procuse of Guidance, what are you doing? We are the nicest and geekiest dentist in the world. It's not like Facebook, it's not toxic. We're all helping sharing, chatting, posting and of course as hundreds of hours, CPD on there as well for you. We are a pace approved education provider. And this episode has a quiz. If you scroll down on the quiz, you get your CE for this episode. Your CPD queen Marie will email you a certificate and every quarter she'll send you your folder, which has all your certificates inside every month. I do a live webinar, a new consent form for you guys and every so often we release a new mini course. We also now have our own protrusive AI, called Ask Jazz and it's an expert at clinical reasoning. To join that, you have to join the Inferrency plan and the best place to go to find out more information is protrusive.app. That's the website, protrusive.app. Once you make a login, you'll be able to download the iOS and Android app. And hopefully you can make an account in time for us to get involved with these discussions every day where discussing new things about AI or whatever the theme is that month. Last month, it was all about personal finances and next month is all about preventing stress and burnout. So check it out, protrusive.app and don't forget to try that smile simulator. The dental distraught smile simulator link is below. If you like what we're doing, give us a like and subscribe wherever you're listening. And thank you once again, I'll catch you same time, same place next week. Bye for now. (upbeat music)