Full Transcript

Dr James: 0:48

So, hey, everyone, welcome back to the Dentistry and Invest podcast. We’re going to depart slightly from finance today, although it is definitely related to the business aspect of dentistry, and we’re going to talk about something that you may or may not know about. If you do know about it, you will have been wowed by it because it’s flipping incredible, and if you don’t yet know about it, then you’re in for a treat on today’s podcast and the theme of today, the subject of today, what we’re going to talk about is chat GPT. Now, chat GPT. I could give my attempted description of what it is, but instead of me doing that and doing a terrible job, I’m joined today by my good friends oh, I almost said Dr Adrian Dray. Dr Adrian Dray.

Adrian: 1:34

Dr Dray thing, I’ll take it, but I’m not a licensed physician, that’s definitely what it was.

Dr James: 1:44

I’m joined today by my good friend, adrian Dray, and we’re going to cover everything in anything. Chat GPT and also describe its relevance to the dental industry.

Adrian: 1:53

Adrian how are you? I’m very good, I’m very excited about this and thank you for inviting me along Enjoyed our chat the other day and I think you could tell my enthusiasm and thought, yeah, you know what it might be good for the bensis to invest podcast listeners. So, thank you, thanks for the bringing me on.

Dr James: 2:11

It’s cool, man. You know I think that they slightly off the cuff and off beat. Stuff definitely deserves a platform, because it’s a bit like when we talk about NFTs and the dentistry invest podcast. You know there is no way that that information really else wise gets circulated. And chat GPT is so flipping cool and imagine at the bit, if anybody hasn’t seen it right chat GPTcom. Basically it’s a, it’s an AI bot, it’s like a conversational list, it’s a chat bot, but on a whole other level it’s like talking to another really eloquent, really intelligent human being. You basically cannot tell if I’ve done a somewhat decent job of describing it.

Adrian: 2:58

I think that’s, that’s pretty, that’s pretty good. I, like you, I sometimes struggle explaining what it is because it’s so different and new to us as as the common man. Now, for people who have been working in artificial intelligence, they might think it’s not that you know. All this other stuff is better, but this has been released to us. This is. This is incredible. So maybe, james, I can give you the chat GPT response. When I asked it introduce yourself, what do you do, what do you know and who is your creator?

Dr James: 3:37

This is how self-sentient this flipping thing is. You can actually ask it what it is, and it’s a long description.

Adrian: 3:45

The self-awareness is off the charts, anyway yeah, this is what it says is I’m a chat bot, a language model developed by open AI. My purpose is to assist users in generating human-like texts for a variety of natural language processing tasks. I’m trained on a diverse data set of internet text and have knowledge on a wide range of topics. My creators are the team of researchers and engineers at open AI. Now, that’s a lot of information, right? And I thought, okay, well, how would I ask chat GPT? I said how would you introduce yourself to my mum, who has no idea what AI is? Right? And this is what it said Hello, my name is chat GPT. I have a computer program that can understand and generate human-like text. My creators at Open AI have taught me to assist people with tasks such as answering questions and writing messages. Think of me as an advanced form of a word processor or search engine and I thought that was quite sweet and my mum would get that. I don’t know if you’ll like that. Your listeners have too.

Dr James: 5:03

Here’s the thing, right, here’s the thing, right. That description sounds so genero and unassuming, but this thing is so powerful that you can talk to it, have a conversation. Here’s an example, right? So the day after it came, one of my close friends called me, right, and he’s a bit of a tech geek. He’s really he’s a tech geek. He’s into all that stuff, right, and sometimes you’ll tell me about stuff and I can’t relate because that’s not my forte. And he’ll say, oh, this program does X, y and Z, and I’m happy for him, right, but I don’t fully get it, right, and that’s the level that we communicate on. That’s the dynamic, effectively, right. So then he called me up and he told me about chat, gpt and I, almost because of past experience, I didn’t get too excited till I looked at it myself and he was like Jim, just check out how sentient this thing is. He was like tell me a joke. He started typing into the chat bot. He was like, tell me a joke about a sausage dog, a kiwi fruit and a penguin in the voice of Dr Dre, right, and I was like no way, okay, no way. And he typed it up and then, right there, it started processing for a while and then the screen, the text started appearing on the screen. It was actually Ad Hock making this joke there, and then in the voice of Dr Dre. I can’t remember exactly how the joke went it actually did make sense, but I remember how it started and it was like hey, it’s your boy, dr Dre. I’m just like I want to talk about my illustrious rap career. Let me tell you about my pet penguin. It countered a kiwi fruit and then it just went off and it just told this whole story and it was this whole joke. I was like this is you made this up? I actually thought it was him. He created this thing, you know, and it was like a joke or something. Then I started to use it more and it could add Hock, just make these things up. Anything you want you could get it to tell, write your emails and tell your stories and all sorts of things. Anyway, I don’t mean to steal all of the limelight, but you can tell I’m blown away by it.

Adrian: 7:09

For anybody who hasn’t and James, exactly the same situation. Now I’ve been advising on AI from a privacy perspective for a few years now and in Dentistry because we’ve got AI clinical tools processing patient data and there are legal things around their legal mechanisms and considerations that are quite complex. So that’s been mine I would say narrow view, still quite different view to most other people, just because of the nature of the work. But when I saw this thing, the first thing I did was write a rap about my cats, and my cats called Classe and Ned, and this is what they do. This is the kind of breeze they are at this kind of stuff. And then now make it into a poem, now make a fantasy story, now write it in the style of Lord of the Rings. And just was playing with it and just thinking my word, it’s so versatile. It can move to all these different in seconds. It’s creating the stuff in seconds. I hope this is really interesting. And then I was out in Sweden just when chat GPC became publicly available. That’s what we see now. It’s been around for a while, but it’s publicly available and it’s beta mode. And then I started like researching and reading papers about it and speaking to AI engineers and seeing what else it can do, and I realized it’s like, okay, I’ve been like stuck on beginner mode because I just been like using it for raps and poems and stuff like this, but it can do so much more. And the reality is that I was actually a little bit skeptical of this being a game changer for business until I started using it properly. Now I don’t even know what level I’m at now. I know I’m a lot better than I was 150 hours ago with chat, gpc experimentation, because what it comes down to is knowing its limitations, knowing probably the kind of the ethics around it and then from that, being able to craft the right questions and instructions and follow-ups so you can really milk as much of the good information and rich outputs as you possibly can. I think that’s so important because when I see people online and I’m not hammering anyone in dentistry on this I see it’s across different industries operating. But when they discredit no-transcript chat, gpt and these other AI tools, I just think there’s two reasons for that Either you don’t know how to use it properly, or you’re threatened by it because it’s gonna put you out of business, or it’s gonna take attention away from you, or whatever it might be. The thing is, as us me and you like do we wanna be on that side or do we actually wanna experiment with it? And if we’re gonna experiment with it, should we get you know? You can pass go kind of card so we can just get to the good stuff straight away. And that’s what I’ve been really looking into, seeing the versatility of it, but how I’m really locking the right prompts to get the nice out of it.

Dr James: 10:42

And I’m really excited to ask you what you’ve learned on that voyage into the functionality of chat GPT, because I was much like you. I mean, I very much treated it as a novelty whenever I saw it the first time. I thought it was funny. I enjoyed that, dr, dr, dr Jenis Penguin. I didn’t explore it much after that. I thought, oh, that’s cool. And then I thought about it a little bit and I was like goodness, this is really powerful. I know that there’s some way that this can be harnessed. And then went back to whatever it is that I was doing before. I started seeing everybody talking about it and I was like, whoa, this is really getting so much. I wanted to say one thing you know, like I can even feel in myself and I’m like, wow, it almost feels like once upon a time that we build people out there. You feel hard done by it, like authors and people who very well read and learned. Okay, you spend a lot of time learning about a particular subject, refining their ability to craft a text on it and author something right. Who know, all of a sudden they don’t quite have that advantage, given that AI has made that ability more accessible to everybody else, and I get that that’s something that might make someone incredibly reticent and even part of me is like, well, that’s not fair. However, at the same time, I recognize that as human beings, we have this innate sentiment that makes us repulsed by things we don’t know and what I’m trying to figure out. How much of that is genuine concern and how much of that is that like caveman sentiment that. Have you ever heard of the Luddites? You ever heard of the Luddites? They were like in the 1800s and they went around and smashed up industrial machines because they were like everything should have come to my hand. We look back on that now and we think it’s not. So is that sentiment in me, the Neo Luddite speaking? Is that something that’s rational? Is that something that is actually gonna inhibit me and hold me back and let everybody else have this advantage? Or do we just embrace it? And I’m not saying I know the answer, I’m just saying that that’s the quandary. I’m not saying that you should think one way or the other, I’m saying that’s the quandary in my head.

Adrian: 12:50

Yeah, I see what you mean. I think there are a lot of unfortunately. There’s gonna be a lot of jobs lost out of this. There’s a lot of industries that it’s gonna be, it’s gonna impact them massively. There’s a lot of people and companies that are gonna get hit because they don’t embrace it and they’re compact to do. But I think that there is a mindset to using this thing and for me, I look at it not for cheating. I’m not doing it to try and cut corners or to be surreptitious Surreptitious, yeah, and fake who I am or anything like that. I’m looking at it to inspire me To really pick on something at speed so that it can help me in whatever journey I’m in on that particular moment, or to assist someone or whatever it is to consult. And I think, when you look at it from that perspective and you’re thinking, look, I’ve got so many tasks in the day. There are so many days where I lose at the productivity game. I’ve got tasks that are still on my list for the next day and the ones that have done are half baked. I think a lot of us feel like that. If you’ve got this thing, that can A help you generate ideas, because that’s a lot of hard time. The reason why we’re running on fumes is because our brain’s just trying to think of stuff to do and how to do it better, and also for it to do the heavy lifting for a lot of tasks. Then you’re gonna start winning your work day. So at the end of it you’re thinking I feel pretty good. You know, I’m at the point now Jay I can’t believe I’m saying this that I have time and I look forward to going through my emails. I’ve never done that, I’ve never had the time to properly do it, but it’s because I’ve managed to just find ways to let chatGPT pick up other tasks, things that would usually take a long time because I’d have to really brainstorm and think about it and it’s just me in my little consultancy. So seeing it from that perspective is really good and I think when you, that’s on your own personal use, but as a world looking at this, it’s not gonna replace everyone, you know, in the same way that smartphone cameras didn’t replace all the photographers in the world. There are some times where we do need to have the human-generated content, the human touch, absolutely, but we’d be fools not to see where we can deal with Use the AI-generated touch to make our lives easier and actually to be able to do that human stuff better.

Dr James: 15:48

Well, I just think I mean the industrial revolution, and every single technological revolution, is effectively a tapestry of increasing productivity. Okay, that is the game, right, and that’s effectively what this does, right? So that’s one thing to consider. And then the other thing to consider is it’s not going away. It’s already been released, right, it’s getting out there. Whether we choose to embrace it or not is another thing, but it is not going anywhere. And you know what? I’m not gonna take a side at all either here today. I’m actually still wrestling with it a little bit myself. But what I am saying is this was swirling around in my head. Let’s circle back to what you were saying earlier. You said that you were delving into chat, gpt, and you discovered there’s layers to their functionality, to its functionality, rather sorry. What did you find? And then you know what would be interesting how does that impact upon dentistry as well?

Adrian: 16:46

Good question. So I realized that, fundamentally, the power is in the way that you craft and curate your prompts. Now, prompts is kind of the broad term for your questions, your instructions, and these can be initial prompts or they can be follow-up prompts. You can even have a sequence of prompts during this conversation. Now, the thing for me is that before I came to this realization, through experimentation and also talking and listening, reading and watching content about it, I would just think what can I use chat GPT for? Okay, let’s try this. I wasn’t intentional with my use, so I’ve had some. I’ll give you a personal example, maybe a little bit later. But I was thinking okay, I’ve got this situation, I need help with it, I need support with it. If I was going to hire someone to do it, what type of person would I go to? Who would I Google? Who would I ask in a Facebook group? Who would I use in the yellow pages Jokes? I don’t even know where yellow pages is. So what type of person would it be? And that’s probably one of the most important things to think about when you’re using chat GPT Ask it to simulate a persona. So, for example, I needed help writing an article. It’s quite a complex article and if I had just said to Jack KBT, write an article about this, focus on this, it would write it in its default wording style and it would try and guess what I want based on all of the text and previous prompts that it’s got, and maybe some of the stuff would be helpful. But it would probably go off course. Now when I say to it act as my copywriter, you specialize in this particular field, you have decades of experience in it and you write in this particular style, then its output is far more fine-tuned and far more helpful to me. And then I was able to use that as inspiration. I didn’t copy and paste it and just put it on a website or whatever it might be. I thought, okay, you’ve got it, you understand how I speak and that’s great. And then I can use these follow-up prompts to tease out more information. So let’s say I’ve got this article. That’s 600 words, 700 words. I could say make it shorter. I could say put a compelling first sentence or paragraph, or give me 10 examples of scroll-stopping, eye-catching, compelling, gripping, whatever adjectives you want to use, titles, all that kind of stuff. Now I’m building on because it’s starting to learn through this conversation. Through this conversation, it’s thinking okay, adrian wants me to act like this, he wants me to know about this, so it’s basically just focused on a specific area of its knowledge base. If I don’t do that, the output is going to be very generic. I might as well just Google, search it and just look through the first five things and pick something from there. So being able to be detailed in your prompts, thinking about what type of persona that chat GPT needs to simulate in that particular conversation, gives you a massive leg up, and that’s just a tip of the iceberg with it. Now to give you a if you’ve got any questions or if you want to hear an example you know what that was awesome that’s so.

Dr James: 21:01

Basically, it’s how we articulate our questions effectively, is how we get the most out of the machine, and it’s almost like a skill in itself. Are you with me? Yeah? So maybe you have a quick example just to make that really tangible. And then the next, obviously the evolution of that, would be how does this become relevant for the dental industry, which it of course will, because it’s relevant for every industry.

Adrian: 21:24

Yeah, okay. So I’ll give you a personal example and then I’ll give you an example that I’ve written up in my book, which I loved doing. This particular example the books, as we were saying earlier is taking me longer than expected. I’m a bit off of all. I can chew they always do, it will be good.

Dr James: 21:43

That’s the flippin’ problem with books, isn’t it?

Adrian: 21:45

Yeah, you keep wanting to reword it and not getting it out. Anyway, here’s this particular situation. Actually, I put this in your Facebook group, james, so I’ve had a lot of calls and messages. People say how on earth did you manage to do this? So here’s a personal example. This happened before chat GPT. So earlier last year 2022, I saw this advert that said did you take out a finance agreement on a car between these dates 2010 and 2020, something like that? And I happened to have done that and it says you may have been missold your car finance agreement and you may be entitled to compensation. And I thought I’ll do that. I didn’t get much out of PPI to see if there’s anything out of this. Anyway, fast forward. I go through the process. They get their fee no win, no fee. They go to a law firm. They look at my agreement. They tell me Mr Dre, you’re entitled to X amount, you were missold your agreement. But we only take stuff that’s over 10,000 pounds in compensation. And I was like, ok, but mine won’t give you the exact figure, but it was just under that. So I said, ok, well, can I have all the calculations and all of your findings, because you’re not going to take it on. I might as well just have my file. And it gave me that. So what I did is I went into chat GPT and I said act as my expert FCA Financial Conduct Authority, because this is who decided that they were missold agreements. Out there, be my expert on FCA infringement and regulation and help me in this particular situation. So what I did then is then I detail what the situation is. A really good thing you can do on chat GPT is you give it the instruction, you mention situation, and then you just give it a little bit of white space between the prompts and your next word I put situation equals, and then I basically just write down what’s happened to make the situation with the car, give it a little bit of context, as it needs that. And then it says oh, it sounds like you’ve been missold and all that kind of stuff. But yes, I’m going to feed you a report, I’m going to give you a report and then I want some help with that. So I take out all the personal information and I feed it bit by bit. I think I have to split it up into about six pieces because you can’t just copy and paste a whole 25 page report into it. But I put this report into chat GPT. It analyzed the whole thing. It’s read this whole thing now, james. It’s got the calculations, the whole shebang. Then I said to it how do I pursue a complaint for compensation to the lender which is the ANROVA, the ANROVA violence? So it gave me all of the different steps. I said write me a complaints letter about this, the row of complaints letter. Then I said rewrite it to greatly increase my chances of getting compensation. So it just tweaked it again. So now I’ve got a full letter that is specific to my exact situation, that is referencing different types of FCA regulation, case law, stuff that wasn’t even in the report which I later verified, it checked, and then it’s saying that if you don’t respond within eight weeks, I’m going to take it to the ombudsman and all this kind of stuff and that’s gone up to Land Rover and I’m waiting, you know, hopefully get my check soon. But what was interesting with all of that is that it took about five minutes and it’s probably cost. It’s probably hopefully was saving a 30% fee. Now the thing is is that I did have to, you know, get a consultant to actually check the agreement. But you know what? I think chat ETT probably could have done it as well If you trained it well enough, gave it the right thing, so it was being able to using that as an example. If you’re giving it context, you’re giving it a particular situation, you’re giving it a persona to rely on, and then you’re getting an instruction and you’ll find tuning the output of that and it’s taking very little time. But you have to be intentional. You’ve got to think what do I want to get out of this? Who does it mean to be? And hopefully that will be a nice bit of money for me, right, and saving on legal fees. What do you think on that? And then I can tell you the, I can tell you the, the other one.

Dr James: 27:10

Damn bro. Did machines just become sentient Right on?

Adrian: 27:15

They’re not going to like Land Rover. Hopefully they’re not listening to Land Rover Finance. If you are listening, you’ve got my address. So here’s an exact, exact example prompts that I used for the chat EBT for a dental situation, and I’ve got lots of these. I’ve created a list of prompt templates that can be used for dental practices which really harnesses its skills and it explores how it can help you in different situations in the business of dentistry. So I thinking, okay, what, what are dentists really struggling with at the moment in the business of dentistry? Recruitment, right, it’s really difficult to recruit associates, nurses, receptionists, practice managers. So I stepped into the shoes, right of a practice, of a practice owner. I had a whole Dr Peter, he’s called. He’s got a practice in Ditsbury, you know he’s got three associates, hygienists, four, four nurses, few receptionists. Doesn’t have a practice manager. So this was the avatar, right? So this is this is only for the purposes of testing it for dental health, dentistry. So I’ve got because you know this thing, I’m not a dentist. I work with hundreds of dentists but I’m not a dentist. I understand your pain, right, understand enough to pretend to be a dentist and amusing chat to each other. So, dr Peter, in this particular situation, who I’m pretending to be. He needs a new practice manager, right? He’s lost his practice well, she’s retired, whatever, didn’t go into that much of the story of it but needs a practice manager and he’s struggling to find one, right? So I’m really struggling to find one. Doesn’t want to ask his mates because they don’t want to lose their practice manager, right? So let’s see how we can use chat GPT for this. So what I do is I provide a little bit of context. So that little story that I’ve given you about the practice and it’s got a good reputation, private only all this kind of stuff in about three or four sentences. That’s our little context, that’s the background information. And then I say to chat, chat, chat, gpt, act as my expert dental business coach. You have decades of experience helping hundreds of dentists grow their teams, all this kind of like who I would want to go to, who, who? I would love someone to recommend me being a doctor, peter, to help in this situation. I am writing it down with the best of the best and then I say to it I need a new practice manager, ask me questions to find out more about my situation, ask me one by one. So now I’m having a conversation with an expert, ask a question and I respond. I give a little bit of background so it could be. You know, what are you looking for in a practice manager? What are you struggling with at the moment? What would be the first tasks, what would their priority tasks in this first day? What, those sorts of things? And now what it’s doing. Even though it’s asking me the questions, just like any consultant, it’s building a picture of identifying the needs so it can better advise, right. So from that it gives me a little bit of advice. And then I said to that well then, give me expert advice, prioritizing uncommon advice first, and tell me how I can find a practice manager for my practice. And then it gives me a list of some of them are quite obvious, but some of them are quite like okay, that’s an inch. I’ve not heard of anyone actually use that as an idea to try and recruit. That’s quite interesting. And then I said okay, now act as my best in class world expert, whatever it might. Job ads writer. Write me a job ad for the practice manager role. To already understand about the practice, understands what I want, all this kind of thing, and it creates this job ad and it looks pretty good. It’s James, it looked like, and it’s doing this in seconds, right? The only thing that’s slow about it is my 42 words per minute typing speed. Okay, I’m the slow one of this conversation. So it’s got this job ad. Then I asked it how can you give me 10 ways in which I can improve this job ad to make it stand out as an ad in the on the job boards? It gives me 10 ways it can be improved. Then I said rewrite the job ad, incorporate these improvements. Right, so it’s already good, it’s already good. Then it just like puts it on and it just like boom. And I was like this is really good, right. And of course, you know in real life, maybe a principle, the actual principle, we want to change it and add a few things and that kind of stuff, but the way it was set out formatted, beautiful. So we’ve got that. So I’ve had a consultation with it, right, which has taken me about three minutes. It’s understood what it is. I’ve got two job ads to use. Then I say to it act as my expert CV analyzer and assessor and give me a guide on how I can assess the right CVs. Look at the right part, look at the strong hand that fall based on that CV, and it gives me some gunk about that. And then I’ve said, write it in a table. So now it’s written a table and it’s got it. So it’s really easy to look at specific elements right Of the CV. So now this particular practice owner, when he gets a CV gone, he’s got a Cripsheet to check. I’m like, okay, yeah, this one looks, this one’s ticking most of the things that are on the assessment guide that chatGPT’s created for me, right, it’s a little bit more to this and I’ll talk about it as I ask.

Dr James: 33:49

It just. You know what, the more I learn, the more it astounds me, I mean even just-.

Adrian: 33:55

You can see the level in that right.

Dr James: 33:57

Yeah, yeah, yeah, yeah yeah, you’ve got more.

Adrian: 33:59

You said there’s more on that. So now we’re on about six minutes. Then, okay, so we’ve got this. The next thing that I ask it to do off the back of that is yeah, that’s it. I said write a response to. I think I get it to act as someone else, act as a recruitment consultant yeah, especially as a recruitment consultant. And I said write me a creative acknowledgement email or an invite in the person. So for an initial chat, informal chat, write that. Got that email. Then I said write it as a DM, okay. Then it puts it uses more informal language, right. Then I said simulate a phone interview with me and a strong candidate. And now it’s playing the different parts of how that conversation could go. That’s me just playing around with it. To be honest, it’s probably people thinking I don’t need that. But you are all of the different steps that you’ve got. If you’re thinking about how do I recruit someone and what would be great to have really quickly, like those acknowledgement letters, like those letters where you have to let people down, right, you’ve got it. But the one that I really liked is I got it to become my expert interviewing coach and I got it to write a load of questions, follow up questions and then also identify us for a strong candidate, in the style of a checklist for when I would be interviewing this practice manager, because I think a lot of people when they’re recruiting, they just have a friendly chat and say what goes on in the practice but they don’t actually dive down deep. I mean, I’m used to corporate interviews. It’s very structured and it’s a uniform approach. But because ChatGPT already understands exactly what the practice manager role does, it’s created this list of questions, follow up questions, associated and related and relevant identifiers on what is a strong candidate, and all of this information and more was done in less than 10 minutes. So, being this made up practice practice manager, who’s gone from scratching their head thinking I put a job on Indeed, I’ve gone and put it on the Facebook, you know places and all this kind of stuff, and I’ve interviewed a couple of people with the interviews and go around. It’s now got a leg up right, probably motivated, inspired and got some assets and guides that you can use as part of that process.

Dr James: 36:54

Nuts. I mean yeah, like I say, every time, I learn more about it, even just as this conversation has evolved. The succession of how I’m blown away I am just increases and increases and increases. Right, and yeah, it’s incredibly powerful. It really is. And for anybody who hasn’t checked it out, chatgbtcom right, am I right, neil?

Adrian: 37:21

I would just Google chatgbt to openAI because it’s I don’t think it’s chatgbtcom it’s. I should know that. I just put chatgbt to the person that comes up, click on it. Actually, I’ve got a bookmark now to use it, so pretty much.

Dr James: 37:36

I was going to say something. Another area, and I think you kind of touched upon this right Copy for your emails for your patients, the practice flipping newsletter right, you don’t need a copywriter, you don’t need someone to do that for you, right? You just need chatgbt with the correct parameters and questions, or, at the very least, even if it throws back something at you maybe it’s like 95% how you want it, maybe it’s not 100% how you want it, maybe it’s like 95% and you tweak a few things. It’s an amazing place to start. That saves a huge amount of time, I think yeah, writing.

Adrian: 38:12

there are some potential pitfalls and hazards to be aware of when it comes to the writing side of things. Good to know. So the first thing I would say, and there’s probably I can’t see how this would have really affect non-clinical tasks, business identity type of use cases, but it’s good to know that chatgbt is built, it’s learnt all of this internet text up to I think it’s the end of 2021. So it doesn’t know who won the World Cup last year, it doesn’t know how many prime ministers we had last year. It doesn’t have the information, no, so it’s not plugged into the internet. So that means that there is potential inaccuracies in the information that it provides. So you have to keep a careful eye on that. Also, it is subject to bias. Now that can come from two things. Basically, the information is learnt and also based on your prompts. It’s got restrictions in there, some things it won’t do with an open AI. It can’t be used. It can’t be used. But generally speaking, chatgbt is acting as an assistant. So if you tell it something’s wrong, it’s like oh, okay, sorry, I didn’t realise I got that wrong. So you can actually kind of manipulate it if you want to. In that particular conversation. Like I said, I think only if you try to prove a point that it doesn’t work is that really going to happen. But generally speaking, if you’re not being an idiot and you’re just using it for general good tasks and business-adventure tasks, then it’s going to be okay. But it’s good to just be aware of those things. The next thing that’s just gone mad, which I think is really worthwhile knowing, is that AI generated text has been around for a while not to this level, but writing blogs for your website and this kind of stuff has already been on the radar of Google. Now Google’s policy is that I think it sort of deranks or it doesn’t play well with AI generated content. And you’re thinking well, how does it know? Yeah, there are AI detectors that are able to try and spot and try and guess and figure out whether content is AI generated or not. So, I would be careful using it for blogs. There are way around there, but you don’t want to affect your SEO and I always say you know, use a respected and well-known and well-used SEO person for your content on your dental websites. But when you’re talking about newsletters and social media content, it’s brilliant. You can write policies of it. James, I wouldn’t ask it to write data protection policy. It could do it. There’s some gaps in it, but you could write policies on lots of different things non-compliance things and it will write it. It’ll do it in seconds for you. You just set out what you want it to be based on, what it has to include, who’s the responsible person, and it’ll write it for you. So that’s really good. And then that was the newsletter, the email newsletter. I think when I speak to a lot of practices because they talked to me about consent and can I email to people and I’ve got loads of consent I just don’t have any ideas on what to email them about. They don’t have a copywriter on hand, can charge a lot of money. So now you’ve got all these people that have signed up and said that they wanna hear from the practice, you could just give it the practice news we’ve got new hygienists. Now we’re starting to do facial aesthetics, all this kind of stuff, and it’ll write it Done, got content.

Dr James: 42:18

You know, I remember in the 90s, when I was growing up, this sort of stuff was pure satisfaction, right. So that’s my frame of reference, right. I remember, okay, I was watching iRobot. By the way, we’re on like 40, 50 minute marks. We’ll wrap this up pretty soon, so I won’t walk around with this story. But basically, yeah, I was watching iRobot, which somebody in the audience will know when that came out, but my guess is like mid 2000s. You know, irobot was in the audience, right, and there’s a part where he walks into this room and he talks to his speaker and the speaker he says speaker off, like that, right, cause it’s playing music, and it’s like. And then the speaker talks back to him and it’s like speaker is now turning off, or something like that. And in the way it’s portrayed in this movie, it’s like a real wow moment and from that era, that was something seriously impressive, like that was something that was just not, that didn’t exist at that time. That was something for the future. Right Now we have Alexa and all of the things, and you can talk to lots of appliances in your house because Alexa can connect to them, which is crazy, right. But this sort of stuff, right, this literal chat GPT. I did not think that this would be. This is ahead of its time. I did not think this level of responsiveness and communication from a robot would be real until like 2030. So anyway, I’ve picked it up enough. Enough talking about chat GPT from me from today. You mentioned all of the stuff that you mentioned today. Will this be covered in your book that you talked about?

Adrian: 43:56

Yeah, so I’ve used some specific examples and it’s an ebook, so there’s some videos as well, cause it’s quite this is quite a difficult thing to explain verbally, you know, and also it’s quite difficult to even write about. It’s good to actually see it, but what I’ve done is I haven’t helped chat GPT has not helped me write my portion of the book. That’s all you can tell, cause that’s all spelling errors. But what’s important is that I’ve set out the things that you need to know. If you want to be the 10% of people who are actually using this properly and not the 90% who struggle with it or just use it for raps and all this kind of stuff, you need to know its limitations. You need to know how to work around to some of the errors that come up. And then we get into the juicy stuff, and that’s about mastering your prompts right, being able to get all of that stuff from just one conversation and getting so much stuff. I think you’re only needed to do this one thing, this one task, but actually I come across a situation, quite a lot, or there’s other things that come up, so I’m gonna get it all now. So, because it’s already learned, it’s got the context, it understands me, it knows what I’m talking about. So that’s why you can get all of these templates and policies and responses and newsletters and social medias and everything, all these ideas just from that one conversation. So I go into that and I actually show specific examples and then give you a prompt template so you can really get the most out of these different use cases. And all I ask people is don’t be shackled by these prompt templates. Use them as inspiration and completely get to go, get your 200 pop bounds. That’s what my book is is to save you the 150 hour learning curve that I had to get to this point where I’m using it effectively and optimizing my business. And short dentists we have today too.

Dr James: 46:09

Good to know, and what’s the title of that book?

Adrian: 46:12

Working title, Working title, but it’s gonna have saving money and saving time in chat TPC and dentists in one kind of all. I’m gonna ask chat TPC to actually help with the title of the book. I promise you that.

Dr James: 46:26

Well, that’s cool If it does all of those things. I’m flipping saw, because that sounds awesome, adrian. Thank you for your time today. My friend, that was a lightning podcast, really bringing this information to the dental industry. It doesn’t really have a platform elsewhere, so it’s really cool that we can talk about this and disseminate this sort of stuff out there and be a little bit avant garde in that sense. And you know what? Anybody who wants to reach out to Adrian, you’ll be able to find him on the Facebook group Adrian Dre, not Dr Dre, adrian Dre. We will find him on the Facebook group Should have been a doctor.

Adrian: 46:59

could you imagine?

Dr James: 47:00

There you go All right, cool, Anyway, Adrian, thank you so much for your time this evening. We shall catch up very soon. My friend See you later. Cheers Directorines, supposed to be reading on this subject.