Archive.fm

Bulletproof Dental Practice

The AI [Dental] Revolution - with Shreyas

Duration:
59m
Broadcast on:
14 Aug 2024
Audio Format:
mp3

The Bulletproof Dental Podcast Episode 361

 

HOSTS: Dr. Peter Boulden, Dr. Craig Spodak

GUEST: Shreyas Parab

 

DESCRIPTION

In this conversation, Peter and Craig interview Shreyas Parab, an AI specialist, about the applications of AI in dentistry. They discuss the potential for AI to increase efficiency in various areas, such as revenue cycle management, chart notes, and phone calls. They also explore the idea of using AI to interact with insurance companies. They explore the benefits of AI receptionists, automated insurance verification, and claims processing. They also discuss the concept of a decentralized DSO and the use of AI in improving operational efficiency. The conversation highlights the importance of leveraging technology to enhance the patient experience and alleviate the burden on dental teams. Overall, the conversation emphasizes the positive impact of AI in dentistry and the potential for growth and innovation in the field.

TAKEAWAYS

  • AI can be used to increase efficiency in various areas of dentistry, such as revenue cycle management and chart notes.

  • Using AI for phone calls, especially for interactions with insurance companies, can save time and reduce frustration for dental practices.

  • AI is not meant to replace humans, but rather to augment their abilities and allow them to focus on tasks that require a human touch. AI can be used to automate various tasks in dentistry, such as receptionist duties, insurance verification, and claims processing.

  • A decentralized DSO model, supported by AI technology, can level the playing field for solo dental practices.

  • The rising tide of AI in dentistry can benefit the entire industry by sharing knowledge and improving processes.

  • The potential of AI in dentistry is vast, and it offers opportunities for growth and innovation.

CHAPTERS

00:00 - Introduction and Background

03:11 - The Energy and Mindset of the Bulletproof Community

06:02 - Understanding Large Language Models (LLMs)

09:28 - Applications of AI in Dentistry: Revenue Cycle Management

13:43 - Automating Phone Calls and Interactions with Insurance Companies

20:45 - Addressing Bottlenecks in Dental Practices with AI

25:20 - The Future of AI in Dentistry: Outbound Calls and Insurance Interactions

29:19 - The Benefits of AI in Dental Operations and Patient Care

34:58 - Automating Receptionist Duties

37:45 - The Rise of Decentralized DSOs

42:23 - AI and the Future of Dentistry

48:49 - Selling to a DSO and Decentralization

56:37 - Conclusion and Looking Ahead

REFERENCES

Bulletproof Mastermind

Shreyas Parab

 

we're live. So today, Craig, we've we have a guest and this is going to be highly technical. Okay, or highly technological, maybe not technical. And I want to introduce him because he came to Summit. He and I connected over Twitter because everyone tagged him that I was looking for an AI specialist to come into my practice to look for inefficiencies and audit my processes because I know that that is the future of where we're going. So, so he and I connected, he came to Summit, he came to my practices, and now we are we are friends. He's made me shoes. He's made you shoes, Craig. I know. I know you've been kicking him around. And anyway, Strayesh, you made a big impact at the summit when you were there. You you attended kind of my quote unquote AI presentation, which I felt like I had imposter syndrome the whole time I'm in that lecture because I'm like thinking myself like he's listening to all these things I'm saying and he's forgotten more than I know about AI. But anyway, you made a big splashman and made a bunch of contacts with a lot of bullet proofers. And we're just glad to have you in this ecosystem. So I thought it'd be a great a great thing that to increase the awareness of AI to our listenership as well as kind of you're going to be speaking at Summit 2025 in Orlando. And yeah, I think it's just part of our part of our mission is to really encourage dentists to become bulletproof through systems, processes, technology, all the things, right? We always say, Strayesh, don't bring a knife to a gunfight if there's if there's one and dentistry is getting more and more competitive. So anyway, but well, sorry for the long intro. Welcome to the show and welcome to the pod. And we're glad to have you. No, glad to be here again. Like I like the summit was just it was a life-changing experience for me too. Again, like I think, you know, I spent all day talking to dentists working with dentists. And then when I went into that community into your family, I was like, Oh my God, these are like the dentists who have the right mindset. They're just like thinking about the future. They're thinking about how to grow their practice. I mean, I was like, I was blown away by the energy. I mean, not only was the summit like it did involve like non dental stuff, like just thinking about how to take care of dentists as humans. You know, my mom's a dentist. That's how this whole thing got started. And I was like, man, I wish she had this 10 years ago when she started, right? I wish someone was thinking about like, Hey, you know, how do you take care of yourself as a human? Because that helps you take care of yourself as a dentist, which helps you take care of your patients, which is all she's spent her whole life doing. So I was like, I love the bulletproof community. And so grateful you guys have welcomed me into it. And I'm glad you like the sneakers. I'm glad you like the sneakers for sure. Yeah, man, my kids were jealous when I got home because they're all about the kicks in my house, but I need a new closet for my shoes. But it's a yeah, they were they're a welcome addition and super cool and obviously custom. So thank you. Craig, let's say you. Oh, he says nothing because he's muted. You can never have too many Jordan ones is what I think you can never. Well, first off, I don't are they Jordan ones or Air Force ones? First off, they're Air Force ones. Well, whatever, you can never have enough like retro Nike's, leave me alone. I'm very happy. And thanks for getting my size right, Trias. Appreciate that. Um, so, Shreya, so as I was kind of talking about the, first of all, thank you for those compliments about the energy. I agree with you. We always say that like our most recent summit was our most favorite. But Craig and I offline have said to ourselves, like, it was probably our most fun because of that energy, because of we had Uchi and because we were kind of talking about, you know, a lot of health components and energy components. And you know, we've launched, I don't know if you listened to the podcast, but we launched kind of a BP health component because we just feel like dentists are abused. And you have, you have real data on in your family life that your mom kind of has done the same thing, right? We risk at all or risk it all. And, and our health becomes a consequence of trying to succeed in, in this quote unquote gunfight, if you will. Yeah. So anyway, thank you for that. All right. So here's where I want to go with this pod is, is basically assume that people understand things from a level of like, yeah, there's chat GPT and it's fun and it's cute. But like you're in it all day long with your company looking for how to increase efficiencies. That's why, you know, you were kind enough to fly out to from San Francisco to Atlanta, from Silicon Valley to Atlanta, and basically spend a couple days with me just ideating and looking at solutions. And where can I fix my efficiencies? Because Craig, maybe I quote you with this all the time, but you say that, you know, Tony Robbins says that, you know, you have to be continually innovating in marketing. And that's what the job is of an entrepreneur, right? And so, like, if we don't, if I said to myself, I don't keep looking at this stuff, right, and keep innovating with technology, then like, I need to hand it off to I need to hand off the reins of the company to someone who's going to, because it's it's negligent, in my opinion, on my part. And this is just speaking for me. So, Tres, let's go with the understanding that everyone understands chat GPT. Maybe they think it's cute, and they've used it. And it's a, it's a cool little thing, right, that helps this in that. You know, there's, there's another layer of that, which is called long, large language models, which I spoke about in my presentation. Maybe you could go into a little bit about that. But then also, like, where's the rubber going to meet the road for dentistry, right? What's the road map, things that are, that are maybe hyperbole, you know, in like, oh, it's going to replace this and that. And what's like real life actionable things that you're seeing? Yeah, no, that makes sense. And yeah, I mean, like, I think sometimes it's useful for people to like kind of understand what, you know, I call LLM's like prediction machines, right? They're basically trying to predict either the next token, which is a word, or it could be trying to predict the next pixel in an image, right? And they're really good at like predicting things. And the reason that they've gotten good at predicting things is because they've basically just seen a lot of information, they've seen a lot of data, they've read every book, they've scoured the internet, they've seen thousands and tens of millions of images, right? And they're basically predicting things. And when you think about like what I'm doing when I talk, I'm kind of like predicting my next word, right? Or let's say when I'm drawing, I'm kind of predicting my next move. And you end up seeing that as you get trained on a lot of stuff, you get really good at predicting the next thing. That's kind of what these language models do. And chat GPT is like generating a response. They're basically like, hey, I have seen a thousand sentences that look similar to the question you're asking. Let me try to like form an answer and predict what what kind of answer would be an appropriate one here. And that's kind of like what chat GPT is doing. It's like, you're having a conversation because it's predicting the next word to say or predicting the next sentence to say. But one of the things LLMs are really good at is reasoning. When you start to have like multiple predictions like next to each other, you start to be able to like figure out like really interesting things. So I'll give a quick example if that's okay. If I would say like, you know, I'm hungry, let's go get blank. You might be able to fill in that sentence. You might say, let's go get food, right? Let's say I before I told you, hey, I love Burger King. And then I followed that. Let's say I went like, you know, I love Burger King. I'm hungry. Let's go get blank. This time around, you might say something like burgers, right? You might say something related to Burger King. Even though you know nothing about like what I'm hungry for, you can kind of sense from what I'm talking about before. What's the next word? And now you might say, Oh, well, like now I know Trace likes Burger King. Anytime that Trace is talking about food, he might be thinking about things related to Burger King, right? He might like french fries or whatever. And they have this like internal representation of this information, like they've associated burgers french fries soda with Burger King. And now they can like tell you weird things about what you might do next. For example, if I was like, Hey, unrelated, I'm going to have a barbecue and I have to go to Costco to get some blank. You could easily substitute hot dogs or burgers, right? But because they know something about me and about they have data that says like, it's likely that he's going to want burgers, right? So you're giving a differentiation between just the agnostic GPT 4.0 kind of thing, right, which has no specificity towards the human versus then getting, you know, creating a custom LLM, which stands for a large language model. Is that right? Yeah, and basically what I'm I guess I'm getting to is like, when you give it more information about your specific thing about your specific desired behavior, like, let's say you wanted to just like predict the next, I don't know, token, right, which is like guessing the next word. But let's say I wanted to predict like, Hey, I have 10 appointments that are, you know, where you know, I have 10 appointments and three out of the 10 people never come. Well, what are the similarities between those people who don't come and what are the similarities between the people that do come? When I get that, if I were to get a new set of 10 people, I could kind of predict, Oh, hey, like, I noticed that this person, you know, lives in this zip code, which is not my zip code. There's a less likelihood of them coming, right? I'm able to predict certain things about the future based on the context that I have on my data. So Craig, yes, I'm going to pause there because I'm going to use Craig as the an avatar or non technological avatar. Okay, Craig, given that stuff, where would you see, where would you see the applications where the rubber could meet the road in your practice, right? Where would you see that maybe, okay, here's a problem that I didn't know could be potentially solved? Well, my brain naturally went to distinguishing quality of patient. So my brain went to understanding there's certain we have raving fans of our practice and they show up for their appointments, they do lots of work, they refer people, they are fans, they are, they're the core. And my brain instantly, when you're talking about this, took all the data from my PMS to say that these are your ambassadors for better word and then treat these people especially well or do something kind for them and strip down my marketing budget and focus all dollars on these people. That's where my brain went. I love it. Sorry. That's great. Very intuitive, actually, Craig. No, that's awesome. And you know, like, it is like exactly that your PMS has all of this data, like, and a lot of it is unstructured, right? Like, for example, I love the fact that when my mom sees patients, she like remembers a fun fact about them from the past eight visits, even though this might have been eight years ago. And right now she kind of like reads all of her notes, clinical notes from like the past seven years and she's doing that, you know, five minutes before they walk in. Well, like, how nice would it be if you threw it at an LLM and you gave it context about this specific patient. So when they come in, you're like, oh, hey, like, I remember you noted, you know, you had, you know, pain in your lower left mouth last time. How is that going? Like, I remember we kind of addressed this. It's such a different patient experience than what, you know, you already have the data. It's so tough for you to surface that. That's, and again, in your case of marketing, that's also, yeah, I never thought about it. Craig, in terms of marketing, you know, yeah, so I just want to add, sorry, just to cut you one second. And the reason why that my brain went there is because this riddle of who are these core ambassadors, and I've called them that in our practice, I know that the ROI on each dollar spent marketing to them is unbelievably, exponentially higher than any other form of marketing. I know that if I could readily distinguish who these people are, every dollar spent could yield 20x. And I promise you on every other medium of advertising, it's spray and pray at best. Straight, you're not working specifically, like, on that, right? Well, someone listening to it is going to start working. You want to delay this and what we're working on is like specifically revenue cycle management. You know, like, I think, you know, for example, one of the things we do right now is when you get verification information from all these different insurance companies, it's super unstructured and unstandardized, right? You might notice your receptionist or front desk is like also calling the insurance company trying to get additional information. Everyone says that they're doing verifications, but you usually get like 30 to 50% of the data you need. And what ends up happening is the receptions of navigating to the insurance portal, typing in all of this information, getting it. We can actually train an AI agent to do that, which is basically like you give it the login information, and then it will learn how to navigate the insurance portal and the way that it's still meaning meaning and and so it's not a it's not I don't know the word I'm looking for. For instance, if a human is doing this, it's one to one, right? Meaning one process has to finish before one process can be a new one can be started. And I think with your tech, it allows it to kind of like 25 things could have happened simultaneously. Yeah, in parallel. Thank you. That's what I'm looking for. And I think, like, where that's really cool is like, if you think about what I was describing, where you're predicting, you're basically teaching the computer how to predict the next step of how to use a computer, right? Oh, like, I log in here. Okay, now I see this screen. What is the next button I should click? I kind of predict the next move. And it's really, really robust that way. It's really robust. Craig, this is the first time you hearing this, but I have Shreis as we have integrated with Shreis as stuff and we are piloting a program here at ADS for us, a pilot for us. And it's been, it's been crazy. The exposure of redundancies, the efficiencies we're going to experience, just so much more streamlined. I'm really excited. And Shreis, you spent an enormous amount of time with my team and I really just want to acknowledge you publicly for that. So thank you. Very good. I think a lot of people are worried, like, Oh, hey, like, what does this mean for my team when I automate things? I know, again, I know we're going to talk about that at some point. But no, I mean, like, the beauty of having an AI agent go and navigate the insurance portal is your team can spend more time talking to the patient. That's exactly it. So let's double click on this because Craig and I are in throws of writing a new book and it's called The Patient Experience. And as so many people, I think Shreis, they put their head in the sand to AI is like, Oh, it's going to disrupt or it could, you know, team members could get fired or whatnot. And that's not the design of it. Imagine if you if you had abundant time in your procedures now or abundant time to look at those Craig, those disciples in your practice that you want to market to like it frees up time to enhance the patient experience versus doing these rote, like get on the phone with this or write up my chart notes thing, right, or whatever it may be. It does thing that can be automated and it allows the humans to do human stuff. Right. And that's the way I see one. Yeah. So, um, yeah, did I just go off? Yeah, I feel like I kind of went off script there a little bit there, but is that the way you see it as well, Shreis? Yeah, yeah, where I see this as an augmentation to your existing team. So they can focus on the things that like you can only do face to face, right? Like, I think a lot of people have already started like outsourcing, like, for example, revenue cycle management to these like billing agencies, like, you know, most if you're already outsourcing it, like, you can also be automating it, right? And do it in house and automating it. I think that's really like, so this is where we had a guest on on David Ensey. I don't know if you know him from Twitter. Yeah, yeah, the DSO guy. Hey, he's actually speaking on our summit too. But we were talking, he gave us some cool things about what he does for for his DSO about, you know, looking at global outsourcing and some, and some jobs like here. And I feel like, and that is great. I think that's a next level entrepreneur. When you look at your, when you have enough intel to look at your P and L and say, Hey, Craig, we just discussed this on mastermind. I mean, our mastermind call. When you have enough intel to look at your employee cost, as it relates to a percentage of your top line revenue or profit, right? And that's, and so a next level entrepreneur looks at that and says, hey, I'm too high or I'm too low, right? If I'm too low, that yay for me, I have some room to bonus and be discretionarily, you know, the good guy. If I'm too high, means I'm operating well. I'm not being a good operator with that. I'm not being a very good steward of the inbound revenue. And so where I think that, that said, David's taking it to the place, Shreya's bear with me of, all right, let's look at some global, you know, let's look at some global talent to do X, Y or Z, whether it's a CMO or Craig, what were some other examples he had? Maybe call center thing, I forget. But like, I feel like that's a great thing. But but AI is in the same thing, meaning you're just looking for cost efficiencies. Help me out, help me get me off this ledge here. I'm trying to say I'm not saying injectors not to get people. Yeah, you're not looking to look. I think it's, it's really important to look at this. Like when you look at any industrialized process, there used to be a person that used to plant the seeds and pick the corn and all that stuff. And food was super expensive. And the reason why we have had global famine is not because soil wasn't good. It's just we didn't have industrialized farming. So modern farming techniques have allowed us to have population growth and stable nutrition. So this is the same thing, you know, and my dad's practice or Shreya's mother's practice, it's your mother, correct? You know, there was a person for every single process. And now just like there's been industrialized farming, there's going to be industrialized processes within our organizations to create operational efficiency, which is actually a great thing for our team. Because that allows, if you're like most of us in the bulletproof network, we actually have agreed to spend a certain fixed amount of our operational profit as team expenses. So if we have a tool like this, that allows us to drop dollars to the bottom line, it means more income for the humans. You know, when the AI works well, you don't give it a bonus. It doesn't need a bonus. You know, but when the AI works well, your team gets a bonus. So I think, you know, what's his face from Proleid? Oh, fear, sorry, I'll fear. Oh, fear is a good buddy of mine. Hopefully, hopefully, he gives me for this. But he always says like, it's not that there will be, it's not that AI replaces radiology, but it replaces the radiologist who don't want to use AI. So autopilot just made workload management better for pilots. It didn't take away the pilot. So now you can focus on looking at the turbulence or whatever. You don't have to physically fly the plane. And that's what we're talking about, the practices. There's certain things that don't matter if your human doesn't do it. And it creates more opportunity for those that want to use that tool. So for whoever's thinking of this as an, as an obsolescence tool, they're, they're really going out about it the wrong way. Well said, Craig, by the way, go ahead, Trace. There was this office manager I spoke to this morning. She, you know, basically right now is posting these payments. So we have this thing that allows us to basically like read an EOB or like an invoice and is like, parse it and read it and understand it as a human would, and then write it back into your practice management software. So this is very, very helpful for a lot of practices, right? Because they're just inundated with payments, have to do this reconciliation and posting. This office manager, I think she's in Pittsburgh, she wakes up at 4 30 a.m. Because 4 30 to 6 a.m. is right before her kids wake up. And before she has to go into the office, where she can get focused time to just basically do this data entry work, right, which is posting these payments. And like, I think about that experience. And I'm like, that is not what she wants to do. That is not what the dentist wants for his team member. But unfortunately, that is the only uninterrupted time she gets in her whole day. Because as soon as she's patient facing, she's a great office manager. She wants to be up and active. She wants to be engaging with the patients, engaging with the staff, really making sure that in office experience is running well. And so the only time she's getting to do this, like payment posting is at 4 30 a.m. That is like, not how it should be for anyone, right? And I understand, because like, you know, I've been in the practice myself, I've like, I've worked there like, I understand patients are coming in and out, people are coming in with questions. You don't have that unfocused at that point. Yeah, exactly. And so like, if you can, if you know, Liz is fantastic, hopefully we can take that work off of her plate. So she can like, one, sleep in and like wake up for kids, like, have a good time in the morning. But when she's in the office, she's in the office, she's present, she's not worried about making all of these payment posting before payroll. Look, and Liz doing that, not, not, uh, not disrupting that process for her, Shrayos, Liz burns out in six years, right? Right. And hates her job and becomes, there's, could become some apathy towards that. Why did I have to do this person? And now like, you may have created longevity with this, with this human, you know, and happiness with her job. So that's, that's awesome. I look at kind of four areas as the practice owner of, I mean, there's four bottlenecks that I, that I am looking at personally in terms of what I see AI potentially solving. And I'd like to hear your thoughts on this, Shrayos. One, of course, your business is focused around it, right? The RCM or the revenue cycle management aspect of it, insurance verifications, insurance posts, payment posting, things like this. And there's more to what you do in your wheelhouse. And I want to come and touch on that. But another bottleneck I see, Craig, and I'd like to hear your thoughts on this and your ecosystem is, I feel like chart notes are something that stresses the clinical team out to no degree, right? Because you can never be, if we write a chart note up, and it's in a soap format, right? You know, it's soap format is it, Shrayos. So, and no matter how thorough you are, if you have another dentist or another hygienist come look at that, they're going to be like, Oh, you didn't include this. And you know, you didn't have that. It's incomplete. And if you didn't document it, it didn't happen. You know, like you hear this all throughout dentistry. And, and so we have this, where we have people who write these books and novels, which take 30 minutes, each patient kind of thing. And we have some people who are like, MOD number 12, right? And that was it. That was the entire chart note. So it's a problem of being way too anemic and be a problem of being a bottleneck, because now, Sally, the hygienist is overwhelmed because patient number two is coming in. And I haven't finished my notes. So I save all this stuff to the end of the day. And now I've lost some pertinent information to the things. So that's, I'm hoping that you tell me that there are fixes on the horizon. I know the care stack is kind of working on some of this to kind of monitor the appointment and then create automated AI chart notes. But like, is there something in the works that can can alleviate that? So again, we can go back to focusing on the patient and not worried about, Oh my gosh, am I gonna have time for chart notes? Yeah, yeah. So that that is definitely happening right now in medical, we actually, you know, one of our investors invest in a company that does this for medical, where again, they basically like voice to, sorry, you have speech to text for clinical notes, and then they formats it in the soap format at the end. So again, like dentists can basically talk out hygienists can talk. Sorry. Yeah, no, I'm already doing this. So I use chat GPT 4.0. And I say it, I guess it knows I'm a dentist because I have memory turned on. So I was like, please help me write the note. This patient walked in and she had pain under lower right. I topped on tooth number 30. It was positive precaution. I told her she needs a root canal. There's a little bit of bone loss, not much. I'm speaking like this. That's a great basic level to do. I think you should see the note. I know, but but imagine not having to prompt it. Imagine have it just to somewhere like being listening in the background and then having it output to the same structure that you would prefer in your practice. Sure. Yeah, you'd only have to have a microphone, like this is the way I like to do it, right? You say this is the way I like to structure it. I like to actually put my objective before my subject, whatever it is, right? It comes out every time in a perfect format. No, it's amazing. It's amazing. And I was saying is that with the 4.0 that I'm using right now, I know it's rudimentary. I have to go in and prompt it and it's not listening. It is absolutely amazing to me that from that just random casual conversation, how it put it in bullet points, I just literally copy and paste it. And I say I talk to a patient about a bridge, but she doesn't know about having her teeth connected and wants to be able to floss. So I talked about an implant. It's always better to have three individual teeth. Patient really doesn't know what she wants to do. So she's going to call back and then it just it's so beautiful. But yes, if I was miked and I said, okay, about to walk into patient number one, Sally Jones, and then and then wait. Well, let's pause on this for a second. Let's pause on this for a second. Trey, I'm sorry. I know you're wanting to. So Craig, you could do the 4.0, just open it up and do that. The second thing you could do is actually, if you're a paying member for OpenAI, is just go in and create your custom GPT and just say, hey, I'm going to give you some stuff. And here's the format. Here's the structure I'd like to have it in. So you're not having to go to the level of what Treyus is about to talk about where there's microphones and stuff. You're like, hey, every time I say this, I want you to structure it this way. And that's just a quick little custom GPT. I mean, I could have one set up in what, five minutes, Treyus? In Craig, then it would output every time the same way for you. So that's level one was you, level two was what I just said. Level three would be, Treyus, what the investor. So go ahead and speak on that. Yeah. And so basically, you know, instead of you like clipping a mic, you still use your phone and you basically just speak it out. You pre-configured templates. So kind of like you're doing some prompting in the back end. But a lot of the tough thing with AI is like, you can get 80% of like the job done very quickly. But it's like, last 20% of perfecting it to make it reliable and robust. So that way you could like trust it consistently. That's the hard part. But yeah, like, I mean, I would even pause it like, imagine not only this, like, you know, I've seen a couple of associates, like they're getting use of writing clinical notes the way that your practice does. Imagine the LLM asked you a follow question. They said, Hey, what did they think about this? Or, Oh, hey, like, that's not this. That's like, that's even like crazy. And then I'll take it even one level further. Level further is like, you mean to make sure that it's comprehensive, Treyus, comprehensive. So that way, if the insurance company were to ask a follow up or if there were, you know, a follow up, you know, maybe a specialist who would be looking over this, you would basically remind the dentist, had Jen is like, Hey, like, make sure you don't forget this, you know what I mean? Because I've noticed over time, like you just kind of forget to mention about this specific condition of their mouth after seeing it, right? Like all of these like specificities you can ask follow up questions. And I think that's like, that's amazing. One layer, one layer on top. But the area that I'm really excited about any and I'm biased because I do like revenue cycle management is like, a lot of times clinical notes like help can make or break a, you know, approval from the insurance company. A lot of times like you don't have enough documentation or you're, you know, they ask, they like ask you to spell like this magic. I call it like Harry Potter. The insurance company expects you to know like the magical spell to get the insurance claim paid. And I'm like, how the heck is the dentist supposed to know what the magic words are? Well, is it their job to not to delay, delay, delay, not pay kind of thing? I mean, isn't that the role? It's a war of attrition with every insurance company. Like whoever's gotten in a fender bender, they will make you fight for everything. The only reason why dental insurance actually works is because we are a proxy for the insurance company. You go on at nadental, there's nowhere to write a bad review. But on Jones Family Denno, the atna provider, they torture them. Right. So if you have a business where the customer reviews another person and not you, it's a winning formula. Every time you piss somebody off, they write a different review for somebody else. It's great. That's the only reason why they exist. It's a war of attrition. Who will stand the phone long enough for the predetermination? Who will fight? Who will this and not? It's but usually we back to the bandwidth of, you know, back to your example with the lady waking up at four in the morning, we have a limited amount. We have limited bandwidth of available time to do that. I'll get on the phone with them later. I can't handle this right now as I would have heard before my office, right, which means I will never get back on the phone with them because I don't have time because the next series of events is going to come in my door today. Right. And so the war of attrition is usually not won by the solo provider. And I my favorite episode you guys was that don't bring a knife to a gunfight, right? Because you kind of describe like if you're a solo practitioner, even if you're a group like you are fighting against Goliath here, right? Yes. And you know, there's this crazy stat, at least in medical, I think, I think it was like, let's say they were to deny 500 claims, they've noticed, signal notice, hey, only 2% actually appeal. And so we might as well just like be denying a bunch of random stuff because what's the likelihood that they end up appealing, right? This is like, this is like, this is like, this is the business that is the inherent business model they have. Let's just hope that people get pissed off. That's what Geico does with their insurance, you know, like you have a $500 deductible or a $1,000 deductible and the benefit is 1,800. Screw it. They're probably just going to quit. And that's where like, that's where again, the automation is really nice is because again, like, how many times can I be on hold with the insurance company? Yeah, you'll wear out. Yeah, you'll wear out. How many times can your AI fight it though? Unlimited? Unlimited any day, any time of the day? I mean, that is like, you know, one of the things we've mentioned, imagine if you had to play a person in chess 24 hours a day for 15 days straight. Hey, I was going to win. You're going to wear somebody out. You know, like, wow, this guy's great. I can't believe I'm playing digital chess with this amazing person. I'm it's been 36 hours. I need to go pee and eat. And this thing's just kicking your ass. So it's the same thing. All right, Shreya, can I go to example number two? Yes. A bottleneck just so I want to hear where the where the technology is. I love the notes one, by the way, love that. Love that. I know. I know. I think that's just that's a huge bottleneck of our time. So phones are another example I want to talk about. And then I'm going to bring up something that you said that was just brilliant at Summit. So phones are, I think, one of the things that we at different times in the week, we have different call volumes, right? Like Monday morning, pretty high, Wednesday afternoon, pretty high. All the other times, maybe anemic. So so the worst thing we want to hear as a practice owner is patience, eager to get started. And they call and no one answers and it goes to voicemail and no one ever calls them back kind of thing. Like that's a gut punch from a whole lot of operational things from a marketing perspective. It's just it literally makes us vomit when we when we hear this as as dental owners. As well as maybe maybe the correct script not being followed or a patient not converting. You know, I always tell people if a patient calls your office asking me a new patient, they've already chosen to use you. It's just going to be your spot. You know, the only thing that can happen is your team could potentially talk them out of it by whatever they've said, right? So this script might go rogue a little bit. We may not have created a script and Craig, Craig's not a fan of scripts, but like, I like structure because because we're all the time just trying to get into that place of saying, yeah, great. See you on Wednesday. See you button the chair kind of thing. So there's two things that I'm hoping the day I can address phones answering and being human like I know that's probably down the way. And then and then what was the other thing? Just being able to convert. And then lastly, I want to say that one thing that you brought up is you're like, look, the tech may not be there for an inbound call, meaning an AI to answer the call and saying, hey, welcome to Jones family dental and it'd be indistinguishable from a human. The tech might not be there yet, right? We're probably six months out from that, I would say. Maybe you can give Intel, but you say, yeah, patient may care if they get quote on an AI agent. You said, but you know who will not care? And I was like, who? You're like the insurance company. So what if I could have your AI human at scale call insurance companies, right? Yeah. So go ahead, Chris. No, I mean, holy shit. You're right. Yeah, like, I mean, there's so many different phases of it. But yeah, like outbound calls, much easier, because I think the challenge that I think I've heard a lot of dentists have is like, hey, I don't want my patient to think that I'm just like not that they're going to deal with a robot, right? But on the outbound call, insurance companies don't care, right? They've already outsourced that function to some somewhere else. And again, it's like pretty believable. In fact, we were joking like, actually, for your practice, Dr. Boudin, like for the two factor authentication code, we have an automatic phone call that can go out to your office. And I was just looking for it. I can play it out loud. It kind of sounds like me. It sounds pretty good. But it's basically like, hey, we just sent a two FA code. Can you like read out the code to us? Just because like, on insurance verifications, we have to log into the portal all the time. But yeah, like, your team doesn't mind that it's not that bad. Like, no, you just want to get the job. I don't care what insurance companies think about my call, my calls. Listen, I'm thanking chat GPT now when I talk to it. You are. Yeah, my kids tease me. They say, oh, you know, everyone, someone's like, you know, anybody asked me any question. I'm like, hold on. Like, oh, here you go. Here goes dad talking to his boyfriend, chat GPT. Like, I'm like, listen, it's a well, it is proven. Craig, new data came out. If you think it, you get better outputs. There's new data. Yeah, I mean, it's coming. It is becoming if you're not an asshole, you get better, better results. So mine, yeah, so it's becoming literally like an extension of my brain. So it's like, I was out in downtown Aspen last night and I'm sitting on Hyman Avenue or whatever wrote up a street, I'm on a pedestrian area and I'm looking at the water running through in my son's like, where do you think this water comes from? You know, it's water that runs in the sidewalk. And I'm like, chat GPT, you know, and it's my, it's my hot button on my phone. So I'm just like, when did the water feature in Hyman Avenue, like, well, in 1984, it comes from a natural spring and blah, blah, blah, I don't wonder about anything anymore. The minute there's a point of wonder, it's instant. Like my son, when he worked for my office for two weeks, right before I left for Colorado, he said the chat GPT prompted it like, Hey, I worked from eight 42 to 11 15. And then I took a break from, you know, 11 13 to one. And I got back to work and I make $12 an hour. And it's a, and gave him this, and he put it together in a sheet and gave it to Erica for his paycheck. It looked like he had spent hours on it. So I mean, I, you, you bring up this point that people don't, people aren't bothered by, you know, the insurance company is not bothered by a robot. I think we're just early days and we're all going to be thanking our robots. Oh my God. You know, so when it's like, Jones family done all nice to hear you, you don't give a F if it's a robot or not. Like, thank you so much. Wow, the receptionist was great. I'd like to talk to the AI, please. Yeah, you know, exactly. Sounds like you're having a bad day, Cindy. Can you ask me the AI? So you stink today. Can I talk to your robot, please? Yes, your robots. So no, I don't care. It's like the, it's like the matrix. I want to eat the steak. Give me the pill. You know, what's that blue pill, red pill, whatever. Remember when people, you know, when we started doing like patient communication via texting, I was shocked at how many people preferred that they were like, just text me. Like, I don't want to deal with more humans. I'm, you know, like some of the patients, like, I deal with humans all day. I just want, I just want my appointment scheduled. I like don't. Human beings are messy. I don't want nice service. I don't care if it's a damn robot. Well, you're saying like, you're saying messy, this like psychological, we bring our baggage or bad day or good day or whatever tone you can, you know, the human language, you know, interaction is so complex. It's we think it's just words. But if your receptionist is having a shitty day, even if they say all the right things, the pausing, the size, the breathing, it leads to a bad experience. You know, Tres, back to your text me. They've actually done studies on this now. And I think this was kind of exposed after COVID or because when we had to do, you know, lack of human contact, people just, we found the people preferred that. So like, the new trend in dentistry is like receptionless dental offices, right, where there's not a reception person you check in, you go check into a kiosk. And there's data that proves that people would actually prefer checking in on an iPad than they would human. Let's take a real world example. I have a lot of meetings at Starbucks. So I show up to Starbucks. I sit down with the person that I'm meeting. I pull out my phone, I'm saying, excuse me, I just want to order real quick. It knows what I like. It knows I liked exactly the 2% blah, blah, blah, just I just reorder my drink. I don't have to stand online. I'm not tipping because you always feel about when they turn the screen around. It's like 20, 30, 90% tip, you know, how it goes or whatever it is now, 50, 70, 140% tip, I guess it is now. But I don't have to have that. I just place the, I just reorder, and I'm still sitting with my guests, the person I'm having a meeting, and they call my name when it's ready. I just see everything. So net net, it's so much better. And I'm a social guy for an introvert who hates people like you, Peter, it's amazing. That is not true. I just had to just jab you. It's been a while, but you know, listen, I would actually enjoy like kind of BSing with the Starbucks cashier person, you know, but for even a person like me, it's just so much more efficient. So I think, you know, right now we're talking about one level of not minding when our insurance companies talk to our AI, I fully agree that we're going to expect and enjoy talking to AI receptionists within 36 months. Yeah, I don't think you're wrong. And I think it is, it is one of those things where the technology is just getting better. And like just at scales, like we're just not capable, like we're not as humans are not good at thinking of like exponential growth. Yeah, we're good at linear growth. We can think yes. Yeah, we're good at forecasting linearly. But when we look at these technologies, like, you know, 3.5, you know, GPT 3.5 versus GPT four, like I remember watching and being like, Oh, this is not just better. This is a thousand times better, right? Like that's, that's the exponential growth and quality. And again, like I remember when, you know, for our like AI voice for the two FA codes, people thought that was me. Like, and I was like, it was crazy to me that they probably wouldn't have been able to tell. And in fact, they kind of liked how direct I was. I was just like, Hey, did you get the two FA code? Yes, you know, no, and it, no, it is, it is a different social paradigm. But I think in your case, like the fact that the barista isn't like typing in numbers, looking down at the screen, like, ferociously, like thinking about a thousand things allows her to look up and ask like, Hey, how is your day going? And really mean it? Like that human connection is there. So when you want it, it's there, not like, Oh, yeah, like, you know, I was, I was a reception at my mom's office, like, I would be having a bad day just because people would be calling the phone, you know, which is not asking you something about the tech. Actually, it's just a 30,000 foot view. So I always talk about in medical, I always use a term called a half life, which I don't know if they use that. It basically is like, from the dawn of time to now, right, what is the amount of time, what is the half life that's going to have to exist for that corpus of data to double, right? Because things move fast. And so I think in medicine, it's like, I don't know, three and a half years or something, right, it doubles kind of that corpus of data doubles every three and a half years. But in a situation like this, where you have machine learning, and technology begets faster technology, begets everything builds on itself. Like, what is the half life of something like this? I know that's kind of a hard question, but like, when do we have, when do we have, when do we have Jarvis is where I'm going, like the AI Jarvis. Yeah, timelines are tough. I always say, like, victory is inevitable. The timing of it is uncertain. Okay. It is inevitable. The question of timing is harder. I think, you know, when I, when I first experienced, you know, I was playing around with like earlier versions of GPT before it became super popular. And I remember it wrote, you know, GPT three, this was like 2018, 2019, it wrote me a rap. And I really did think I was talking to Jarvis, right? But then my standards for what intelligence looked like increased because the models got better. And I remember like, as we've kind of think of like what our bar is for human intelligence, it just kind of keeps on getting higher, right? Like, I would say like, when I talk to chat GPT, I really do think it is like Jarvis, right? But that's like the, is there something called the Turing test, right? Yeah. So yeah, like the Turing test is basically like, hey, could you talk to a, you know, if you had two people, a human and a computer, could the human not realize they were talking to a computer, would they think they were talking to a human? I think like we've, we've kind of like met that. We've passed that. Yeah, we've passed that. And so we haven't met the second layer, which is like, as people talk about AGI, right, which is, can you kind of talk about that? Yeah. So, you know, AGI is kind of like artificial general intelligence. You know, in the case of LLMs, like the way that they're almost like the smartest toddler you'll ever have. Like, they know a lot about everything, but they don't know how to do specific things, or they don't know a lot about a very specific thing, because they're kind of just like, like, directionless experts, right? In order for them to be able to be good at a specific thing, you have to kind of teach them, you have to give them examples. AGI is one of those things where it's like, I could just have a general machine that could do anything. The same way a human would like open up a website and be like, I've never used this tool, but let me figure out how to use it. That's AGI, where, like, right now, I am teaching a computer specifically how to use an insurance portal. That is a very specific task. That is like, I don't know how many, you know, examples I would say like, let's call it 500 examples. I need to give it examples and teach it how to do this one thing. Eventually, AGI is kind of like, hey, I don't teach it how to use an insurance portal. I just teach it how to like, use a computer. Actually, I don't even need to teach it how to use a computer. I just give it a computer and say like, start, and it'll figure out how to use a browser, for example. And the early day, I remember it was maybe like 2020 when I saw this, their earliest examples of this where they would have computers learn how to play video games, and they would tell it nothing about the video game, and it would just start learning how to play it after a bunch of trial and error. And they got pretty good at playing video games. And I was like, that's when I was like, this is what I want to dedicate my life to. This is awesome. This is awesome. So where is the product suite for lack of a better word? Like, let me understand, because it sounds to me, and again, I'm just learning about what you and Peter are doing. Obviously, you have this RCM aspect to your business, but where does it go? What's the scope of services of what you envision? What are you trying to create long term? Long term, let me be, let me be focused within 18 months. There you go. Yeah. Yeah. I think on the technology front, like I want to basically from verifying the insurance, like we pull data automatically from your PMS, we verify that patient's insurance by going to these portals, navigating these portals, getting all the data, and then inputting it into your practice management software. Then we, you know, right now we have humans who will actually like go and like submit the claims and do the attachments. I want that to be automated as well, right? Where we can actually do analytics and say, Hey, you know, like for this specific payer, we want you should attach these things. Hey, you're missing a medical necessity letter from the primary care physician. Let's automatically send that. I don't know if that can happen in 18 months. I'm kind of ambitious there. But yeah, basically that from claim submission to claim appealing where we're, you know, for example, for Dr. Bekora right now, I can tell you the most common reasons he's getting denialsed, and we can kind of pinpoint why these claims are getting denied and maybe what we can do about it. And then finally on the payment posting, which is automatically writing back into the practice management software. This is like basically the end-to-end RCM process. I think where we like really want to go not only in the 18 months, but in like life is something you mentioned before that decentralized DSO, right? Yes. Can I give a little? Well, we mentioned that when you were in town, when you when you were in my office and we sat around in '98 and I said, look, whenever Bulletproof does a DSO, if we ever do one, that I want it to be kind of the first one that is like AI functioned first, right? If you really find those efficiencies with tech and so, yeah, you can expand now. Yeah, like, you know, one of the things that we continually hear is like most practices, even the best ones, have cash flow challenges, right? Where, you know, you might be doing a service today and getting paid 30, 45 days from now. Like the world that I imagine is like, if we know data about this claim and we have historical data, we can probably predict what the likelihood of this claim being paid out is. And we can lend you that money at a much, much better rate than anybody else in the market. Because again, we are doing, we have all of this data about this claim. We have all this data about your practice and essentially allow you to get paid instantly. That is like, if I get my card. Craig, I want to give you a little more context. Sorry, Shreya, let's cut you off. So we were talking, it evolved into crypto because, because, you know, Shreya and I think we're talking that way and all smart people are in crypto. So, so it evolved that way. And I said, you know, it's just funny how, like, we talk about, hey, you know, in this DSL world dentistry, hey, if you centralized all your services, and if you've done the things of like legacy banks and put everything in one location, I'm like, wow, but in crypto, it's everything is decentralized. All the nodes are around the world. And that's what gives it strength. And I was like, you know, and so Shreya's kind of brought like the future is kind of in like decentralized, you know, autonomous processes, if you will. And I was like, wow, another, sorry. No, that's it. Go. Well, I mean, taking another, another layer back from the decentralized DSO concept is look at the centralization of funding. So going all the way back, Peter, to 2017, our comments about, you know, the tokenized DSO, the fractional investor DSO. So like we have these events, Shreya's, where it's like, you sell 51 and 99% of your practice or 100%. It's a very draconian move for someone that just wants a little bit of investment. So a decentralized investment mechanism as well, where everybody could own a piece of everyone's practice. The whole theory behind Bitcoin is that, you know, you're, if you're in Uganda and you don't have faith in the currency there, you could have a central opportunity to invest something. Well, it's just interesting that there's convergence of like methodologies of technology, right? Of course, there always is. And it was just like, wow, the thing that Shreya's was talking about from an RCM and a decentralized kind of DSO group of dental practices was like the same as like the where the money is going, right? Kind of digital money. And I was like, wow, what a parallel. And so I was like, that's a good, that's a good tell, right? That's a good, that's a good tell. I like that all decentralized. All narratives converge. All narratives converge. I mean, all business systems, what happens in aviation makes its way to dentistry. That's where like, you know, again, I'm an outsider to dental compared to most, like again, like I'm not a dentist, right? Much to my mom and dad's disappointment, I'm not a dentist, right? But, you know, when you kind of like come in with that technology perspective and you can you look at it, it's kind of inevitable that, you know, you have these big DSOs, which, you know, some people have great experiences, some people have not so great experiences, whatever, who say, hey, I'm going to do all of this back office services for you, so you don't have to think about it. Well, like, actually, we can automate some of these back office tasks. You can continue to own your practice. Like, my mom never wants to sell to a DSO, and that's like a personal choice for her and her practice. But there's no like alternative, there's no like third or fourth door here, where it's like, hey, I would love if you could take like RCM off my plate. Hey, I would love the marketing, maybe the calls, maybe the phone calls like, yeah, take that off. Well, it's the whole narrative for selling to a DSO in the first place. Do you want to focus on clinical dentistry? Are you not enjoying the operations? Do you want to focus on why you got it to being a dentist? Yes, I fully agree. If they can just do reviews as well, you know, like personnel reviews. We are hiring, firing. Yeah. The like vision over the next like months, years is like, if we're all sharing this like knowledge, for example, like when I'm doing verifications in Georgia for your practice, I'm, I'm verifying that plan. And like, now I've verified that policy. Anybody who has a policy in Georgia, right? Or in that, in that specific policy? Yeah. So one plus one is, is four in your world, because you're cross ref cross pollinating your data and saying like, Hey, what we learned here, maybe good for Dwight's practice, maybe good for Craig's practice, because of X, Y, and Z. And so you're getting, you're getting not linearly smarter, you're getting logarithmically smarter, right? Speaking of practice, when can we implement in my practice? I knew that was coming. I was coming before we stopped to record, by the way. Yeah, because it's like, look at what we've done. I mean, we've increased operational efficiency. It's incredible. I'm sitting here like, what the what the hell is going on here? What about me? Well, we can talk about that. I mean, I think, sorry, we're full, Greg, sorry. I mean, I've noticed a lot of the companies like scale too quickly, because they're like, we like really want to nail these automations down as much as possible. Okay, so I think Dwight would be amenable to being dropped, right? Is that what I heard? You know, the minute this podcast drops, by the way, Dwight is on me for that. So, Treas, we need to drop Dwight, and we're going to go to pilot mid me. Okay, we pilot Spodeck. All right. I'm kidding. Dwight, love you, bud. He's going to call me. So sure. Yes. I didn't prepare you that this pod was going to be very unstructured. We were going to go all over the place because you want to hear a funny, you want to hear a funny story about it. I knew this story was going to come up, Peter. So I knew it. Let me tell it at least, if I'm an insult to everybody, I'm going to insult myself. So we just dropped our way. We just, this is this morning. So the first bulletproof health podcast, BPA dropped. And my wife was actually, she's a health nut. So she was actually interested. This is the first podcast she listened, start to finish of what 350 some mods in 60 episodes. So we finally hit something that she wanted to listen to. So she listened to it on a hike and reposted it. I'm like, what'd you think? It's like, it's really good. You have Peter. I'm like, why? So you and you two are all over the fucking place. I'm like, Peter, like, actually, I'm like, Oh my goodness. I'm like, I can't believe you're saying that. And I told Peter this morning, Peter's like, dude, if I would have married your wife, I'd be president of the United States by now. She's such a powerhouse. She's such a powerhouse. She's such a powerhouse. It's unbelievable. She was like, that's a good thing you and just a good thing you guys have him. And Craig, it pained him to call and tell me this thing. I had to tell you, just like, listen, you're entertaining. But Peter actually brought like, Hey, here's step one step two. Yes, just raised. That's the running joke is that like, I want the one through five. Here's the steps you want to go through. And Craig's like, yeah, but what about the vision? And how do you feel about this? And like, can we create something new? I'm like, just do the steps, you know? So anyway, you're probably as an engineer. I guess you are an engineer, technically. Yeah, yeah, I studied a biomedical computation, which is like computer science and biology. It was I basically asked two subjects rather than full-assing one is what I jokeingly say. But they were definitely very useful subjects. Little did I know I would be end up doing this work, but uh, and that was a Stanford, right? So just not a mid-level school. Pretty terrible. Yeah, small liberal arts, Belgian Palo Alto. Oh, I thought that was Stanford. Um, and now about Alabama, Stanford Tech, a senior parent. If it's any, like my co-founder, Anton, who is by far the smarter one of he's, he's, he's, he's, I'm just like kind of a pretty face that happens to talk about dental. He's the brains by the operation. He went to Berkeley for three weeks before he dropped out. So, and he's, he's definitely like, this is too easy. It's too basic. He was building a company and they ended up doing pretty well. Uh, and so he was just like, well, like what's the plan here? Like, I'm going to go to college for four years and then like just get a job as an engineer. I could just like keep on building the company. And he chose to keep on building the company. They, I think they cross like seven figures in sales and then they got acquired. Um, I think they got acquired like when he was like 20 or so. So that's like, that's like the analogs like Craig. This would be a sports analogy like Kobe just being like, what do I need college for? I'll just go straight to the pros. Like, you know, I'll just get, I'll just get drafted right out of high school. I mean, not to go to tangential, but college is definitely becoming a lot less valuable every year. Yeah. You know, for sure. Yeah. I mean, it's fun for fun. I, I guess it's good for fun and connections, but I mean, unless you want to be an architect, a dentist or something like that, it's hard to really start to defend the value of a liberal arts degree, no offense or like a degree in English literature. So you're going towards Central Schools, Craig? Is that what I'm hearing? Oh, yeah. Very bullish on trade schools for sure. And so funny because the elite class always puts them down like the AC repair guys, whatever. I have to come back to Sanford Tech, but Stanford Tech. Stanford Tech. No, I made that joke. He's like, I went to a small, I'm like, was that Stanford Tech? And I'm sure there's a Stanford Technical School. And that would be great to say that you went to Stanford. So I shout out to Stanford Tech wherever you are. There's not Craig's googling it as we speak. I know it. I am. Well, Strayis, I hope, I hope the ride was fun with the Pete and Craig roller coaster and the and the tangents we almost fall off on anything. Kind of you want to say, I know you're going to speak at Summit. I'm hoping to get you to talk to even lean into our mastermind. You know, we've got a new mastermind class coming up throughout the year. It's actually Craig. We I think we have some some availability there, even still, Craig and I don't promote things very well. But I think we have some availability there. I'd like to have you talk to them. So you got the summit, we got the mastermind. And then is there anything kind of that you're obviously bullish in dentistry? You're very, you're very optimistic about the future of dentistry, especially for the sole provider. But is there anything like we missed on here that you're like, Dang it. I wish I could, I wish we could have talked about this. I think it's, you know, AI allows us to do a lot of amazing stuff and automate a lot of this, like the tasks that you already feel comfortable outsourcing or you don't want to deal with or most people on your team. If your team is like, Oh, I have to go do this. And that is something that's automatable. That is like part of the job is taking that off their plate and allowing them to like really go in to work excited and what you mean part of your job is to take it off their plate. Yes. Yeah. Part of our jobs take it off their plate. And one of the things I've noticed is like it just well not you I wasn't meaning you specifically. I mean as like if you're a business owner listening to start interrupt, I want to give clarity that like if someone dreads a task on their in their in their job, it could be incumbent upon you to figure out a way to make them not dreaded or else they may leave and fire, you know, like retention and all these things happen. We have a problem with that in dentistry as you know. So sorry, I didn't mean interrupt. That's exactly right. Because you know, you know, for four months, I was like in the the depths of running a dental practice, not what I expected to do. I always feel like I was the most overqualified dental receptionist for four months. I was the office man. I was doing everything, right? And you just you see the the dreaded has on the team and how it affects the patient experience. And it's like, I've spent the past 10 years watching my mom come home so tired from this gunfight, feeling like she just didn't have a chance to win. And like, the best part of my day is like, I feel like I actually get a chance to help. Do you feel like you have a chip on your shoulder? Of course you do. Huge one. I love a chip. I want to do a podcast just on chips because Craig wouldn't be sitting here without a chip. I sure it would not be sitting here without a chip. Like chips are just a superpower. And I always tell people like go get a chip. And so like, you have some pain because you saw your mom go through the pain. You're like, it shouldn't be like this. I'm going to show the industry kind of thing. Let me show the powers that be. Yeah, I hope you guys know a good chiropractor. I went to one and he said he can't get that chip off my shirt. Yeah, you don't want to pry it off. Leave that, leave that chip like a barnacle. Craig, anything else? Did you learn a little bit, Craig? Oh, for sure. Yeah, no, I loved it. I really love it. I think it's an exciting time to be in dentistry. And these tools are going to not only be tools of the big boys, but every single practice. Yeah, it levels the playing field. I love this because technology, you know, oh, I can't get a 50 person call center. For instance, stress, we, we had a conversation with gigantic DSO and they had what Craig remediation team was 20 people that would listen to phone calls only to follow back up with the people who didn't call. And so you're listening to this as a provider. Yeah, 20 person team that only listened to the phone calls that didn't end in an appointment being booked. So, so me as a solo provider, I'm thinking like shit, talk about a gun, right? Like that's a gun fight. Like I don't have the either the financial means or the or the ability to do that. So like that's an unfair advantage. Quote, unquote, maybe not unfair is not the right word. But I love where I can see like just hey, onboard our project, our product, you know, and like let's level the playing field so that you get to go focus on dentistry as opposed to feeling like you get your ass kicked by, by, by unfair advantages, if you will. And the rising tide is going to lift all the boats, right? That's what love about this. Like, you know, eventually if we're doing things like appeals and we're helping figure out like, hey, actually, across the whole country, Sigma continues to like have this weird, you know, trick, if you say, you know, we manually reviewed these files, it increases the likelihood that their AI catches it and approves the claim. We want to know that we want to share that knowledge, like you only get that when everyone's kind of getting to be on the same team, which that's my hope is that we create a product that allows even if you're a small individual dentist to feel like, hey, I'm actually, I'm part of this larger network and sharing the benefits of open sourcing the way to it. That's amazing. I love it. I love it. Yeah. Anytime we can see, you know, I mean, honestly, the genesis of this pod, as I told you before, was really just to help help dentists, you know, Craig and I help each other. And then we're like, dude, how can we kind of impact at scale? And so anytime I hear like where the good guy can win, not that there's a good guy and bad guy, but, you know, we are all grinding. The average guy, I think it is, the average, we're all like your, we're all like your mother at some point in our career, right? And it's overwhelming. And so I just, anyway, but I'm really blessed that you came into my life and Craig's life now. I'm really blessed that you're speaking at Summit and I just, I'm so glad that our paths crossed and I'm just a huge fan. I want, I want nothing. I want you to really succeed. I want you to win big time because that means dentistry wins big time, in my opinion. I really appreciate it. And thank you guys so much for welcoming into the community and to the family. Really, buddy, your buddy. All right, everyone, that's the pod. We'll see you next time.