¶ Intro / Opening
Started working with a group to help actually build our own internal model. to to help with our research. We have, you know, thousands and thousands of data points that we have in many different states. And right now we call it sticks and stones. It's like, oh yeah, I worked in that state. Remember we did this or oh we it was a matrix we did on that.
And so it's all over the place. And so what we're trying to do is build a platform internally, because we have such so much data that only we can see and not the outside world, so that our clients can obviously benefit from that. And then we started
finding out what other tools are out there. So it's things like Copilot or ways to kind of help organize your emails or other things like that. If we've experimented with Chat GPT or Gemini, so those are some of the things that we've been messing with. And on the administrative side, we've we're trying to figure out what is
to schedule clients easier and and faster. So we started using things like Zoom Scheduler. I mean all these different tools that are out there that used to take lots of, you know, human power to to kinda capitalize on
You too. Thanks.
¶ Firm Introduction and Practical AI Tools
Welcome to the AI Weekly Brief where we bring in business leaders that share practical AI tools.
as well as their one.
five and ten year predictions for how AI will shape their industry. Good afternoon guys. Thanks so much for joining me on the podcast.
Thanks for having us.
Thank you.
This will be my first recording with two individuals, two guests. Um we'll see how it goes. But should be pretty lightning fast here in twenty minutes. Could you um both share a little bit about maybe your background, your life story and I think you run your own podcast too.
Yeah, yeah. So uh my name is Michael Byrd. Uh uh other gentleman joining me is Brad Adato. We're uh founders of the law firm Bertadato. Uh Bertadado is a business law firm for health care practice. So we represent doctors and dentists and other providers throughout the entire country. A a little Brad and I've been working together for twenty years and um we uh give each other a hard time like people that have worked together for twenty years do.
But it's all in good spirit mostly'cause I'm better than Michael.
Yes. And uh and so uh we're kind of our genesis is that we're both sons of doctor. And uh and that's really what inspired our careers. Even before we met each other. That was our common connection. And um so my dad's a retired plastic surgeon, my dad's a retired orthopedic surgeon, and so we started Bernadotto. We have about fifteen attorneys and we're in uh uh we're all across the country and so we are definitely boutique but really positioned ourselves because of our
the the niche nature of our practice. And so you had mentioned our podcast. We have the legal one, two threes with Berta Auto. A huge part of how we built our practice going back twenty years is based on education. And uh so we speak at at continuing education events for for doctors and dentists and then we've started putting our own content about what five years ago or so the legal one two three support auto.
Just kinda keep uh everybody up to date on education and and letting everyone know um I mean, one of the nice compliments that we get often is that we're thought leaders in this area of law, so it's it's been a lot of fun.
That's great. That's awesome. Um well let's jump right into maybe the the heart of the episode. We'd love to hear how I guess you both use and your f your fifteen attorneys and and folks that your firm use. What sort of kinda AI tools or kind of technology has helped you enable your boutique firm?
So I I'll start on that one. Uh about two and a half plus years ago we started seeing the um these different AI tools that are out there and we're always uh we're entrepreneurs at heart um more than anything else. We're that's how we look at things from a business perspective. We want to figure out is this a tool that we can use and utilize how? And so we started exploring that and trying to understand, you know, what what is AI and how from a legal perspective as lawyers
Can't we utilize that? We attended a bunch of different bar association meetings and stuff to learn more about it, how other lawyers were using what they were using. So that was interesting just to kind of understand the language. And then about a year, almost two years ago now, we got actually started working with a group to help actually build our own internal model.
to to help with our research. We have, you know, thousands and thousands of data points that we have in many different states. And right now we call it sticks and stones. It's like, oh yeah, I worked in that state. Remember we did this or oh we it was a matrix we did on that.
And so it's all over the place. And so what we're trying to do is build a platform internally, because we have such so much data that only we can see and not the outside world, so that our clients can obviously benefit from that. And then we started
finding out what other tools are out there. So it's things like Copilot or ways to kinda help organize your emails or other things like that. If we've experimented with Chat GPT or Gemini, so those are some of the things that we've been messing with. And on the administrative side, we've we're trying to figure out what is
to schedule clients easier and and faster. So we started using things like Zoom Scheduler. I mean all these different tools that are out there that used to take lots of, you know, human power to to kind of capitalize on.
Yeah, I feel like the legal industry in particular is still in the exploratory phase of figuring out exactly what the right place is for AI and yet it's changing so fast that it It's crazy. Like we went to uh one of these bar association seminars a year ago for it was all managing partners and we were the smallest law firm in terms of size that was Uh but you had a swath of of leaders from all around kind of the Dallas area. And what struck us is how much
Every one of them is struggling with trying to fi solve that very thing. And one of the things that puts a lot of pressure on it. is that most law firms bill by the hour and so there is an inherent friction with AI doing something in seconds that they otherwise might do. in hours and hours of billing or days. Exactly. And uh so one advantage we have in that front is that we don't have an hourly billing model. We're a for lack of a better word, a subscription based model.
And and so we have alignment with our clients that if we can do better and faster that that'll work. That works well for everybody. The thing that is um you know, the exploratory part is that we're just we're you can't just throw an AI tool into your into the middle of it and figure that it's gonna solve everything. You have we have to have good systems in place and use it as a tool to support our clients. And so, you know, we've just like everybody struggled with
wanting to race really fast and then slow ourselves down. Feel like we're finding our right rhythm. I think the the enterprise version of Copilot has been a good bridge.
because it does have all of this data that we've collected over thirty years that we can use internally and it can help support us until we get this platform that Brad mentioned that we're having custom built for us. But It's definitely a a I've been practicing for thirty years and this is gonna be probably the biggest change that that uh I will go through and what the practice of law looks like.
¶ AI Limitations and Ethical Concerns
Yeah, and the fascinating piece is learning how to understand how AI speaks. And what I mean by that is unfortunately we don't litigate, but we have seen just bar after bar association coming out where they're disciplining attorneys. Who think AI is gonna help find a case for them or AI is gonna help build their brief and what they don't realize is AI at least this is my opinion, so AI may disagree with me.
But AI wants to plead the end user, so if you're asking it to go find a case similar to this case, it literally will make the case up, make up citations and give you quotes from a judge that doesn't exist. Lawyers who don't understand what AI is are actually some are either being disbarred and or suspended because they're going in front of courts and submitting false information, which sounds, you know, to maybe people who are used to AI tools are like, of course it
it doesn't know. And so I think that's the the piece that um a lot of lawyers have to be very careful. And now that we we work with a lot of people in the in the medical industry, same concept as it it giving advice on on the practice of medicine.
And being very mindful that this is still a tool, as Michael said earlier in the conversation, but it's not something that you have you can rely on. You actually have to double triple check it. You have to use your own common sense, your own knowledge base to make sure, is this making sense?
I know a long time ago we did an experiment, maybe three years ago now, with an AI tool to say, hey, go build us an employment agreement based on these criteria. And AI failed so miserably, it would have been malpractice. Now, that was three years ago, we haven't tried it again since then. But again, you have to be very mindful of of how that tool is being utilized. And I agree with Michael that that uh
There are ways that probably since we used it three years ago to where it is now, it's definitely improved. But I still see hallucinations and if you've been practicing as long as we have, you can see it. In real time, but if you you don't have that knowledge base, you don't even realize that it's giving you completely inaccurate information.
Think of it I think of AI right now as if you if you give it the right prompts, which is critical, that you can expect the performance of a first or second year lawyer to come out of it. And that helps us frame like, well of course you wouldn't send a first or second year lawyer out to to without some oversight to be doing really sophisticated things with clients. And so if you have that kind of expectation, then it really can be effective and useful tool.
Yeah, I agree. Sometimes it's a it truly is a a very green first year lawyer. Right, right. The hallucinations or not understanding the true impact of something and That just can and I'll say it this way for your audience to understand is we're looking at a law. We're looking at what does the statute say? How are they enforcing it? Have there been any case law or in our case medical board investigations? Have we had conversations outside?
outside of that with the boards, with their general counsels, and then we take all of that and put it together. And AI has yet to figure out how to do that. Andor there can be bad law out there or bad opinions or bad articles. have misinterpreted something and AI thinks, Oh, well, everyone's gone to this website I guess to read it, so therefore it must be good and it doesn't realize no, that's actually completely inaccurate and you're actually leading you're you're going the wrong direction.
Yeah, right. Like a a very green first year, like you're saying, you know.
¶ Data Privacy and Enterprise AI Solutions
that happens it happens very fast, so I guess if there is a little bit of time in reviewing afterwards, hopefully it's an output that is to your point, you know, like a little bit of it more of a first second year with some review for sure. And um I I I tell definitely agree with both you guys. had something that came to mind about uh I was gonna ask, either of you use Westlaw or Lexus Nexus, some of the AI tools that they've launched within those products?
We have not, um, but we do know that that's a big twitch right now with that because we don't litigate as much as most lawyers. We don't have that same need. We do have access to it, but unfortunately probably would we'd have to get one of our our younger peeps on this to answer that question.
Our our attorneys in the office use it all the time, we're just too old. It's basically what brought.
Which Michael's really old audience members can't tell just because we're off camera, but he's like nine hundred.
We don't we don't to that point. We don't do a lot of research at just at our role with the firm. But um I do believe that we the younger attorneys
Use that as a tool.
And what they have said actually is that we're not going to be able to They have even seen some of the summaries look really weird compared to how it used to look. And they're worried that some of the information they're getting might not be as in point as it used to be. So they actually again, because they've used it so much
When they see a summary that AI came up with, they're like, uh I don't know if that's correct. So even they are starting to realize it's a tool, again, but they're they've been practicing long enough that um they have some concerns when they see it.
Yeah. I've been hearing from attorneys that there's a bit more accuracy with some of those tools sometimes with less hallucinations. than you know, if you just use a Chat GPT enterprise license which'll have all sorts of stuff that'll come up. Um I have found that piecemeal type thing where you like to your point, you know, when you have a good prompt and piecemeal like a specific section
It will work way better with specific things like that. A whole document drafting and a whole review process. It doesn't matter if it's a legal document or if it's like, you know, speech or something that you're putting together. There's a lot of stuff in there that still needs tuning and I hate the the specific piece that's still hard coded in a lot of these LLMs and it's very much a design feature that it's meant to be agreeable. It wants to give you something that you want to hear.
Which I have to remind it over and over again when I use A High Tools. Don't do that. Give me basically without none of the nonsense that that I you know, it it's it's programming, sometimes it takes like two or three tries to really get it to give me the the straight facts.
Yeah, and a an important part, at least for our firm and hopefully most law firms realize this, is our policy you're not allowed to use Chat GPT or Gemini or even co pilot exterior. to analyze anything because if we receive a contract, so a physician employment agreement to look at
And you dropped it in Chat GBT, the entire world eventually would know everything that happened and what how much money that doctor was offered. And so, you know, having a closed circuit is the only way that can be used because of attorney client privilege. And I've actually had conversations with
Attorneys like, What do you mean it's everyone sees it now? I'm like, Yeah, everyone can see it because it's public. It's it's not private, it's not confidential. Plus I think some attorneys didn't even realize when they were dropping it in there, um they were basically advertising all the information out there.
Sure, yeah, unless you have the enterprise grade license with ChatGPT, the LMs will train on that data to your point. So absolutely right. Like do not use the free versions with confidential information. You have to have a specific enterprise grade license through OpenAI to and they claim that they're not training on it. Whether we believe it or not, that's a whole other question too.
True, very true. Good point.
Especially with some of the rug pulling that's happened for some of the early investors in OpenAI that thought was gonna be a non profit, now it's a for profit and there's obviously some weird, you know, back and forth stuff that some people to feel uneasy with, which is why I see some folks actually trust other companies sometimes more like anthropic. Some folks, you know, like you're saying, really like Microsoft Copilot and
¶ Future Predictions for Legal AI
Google Gemini for some reasons too. So very interesting. Well let's jump into like the lighting round here. The the very last part of the conversation. I'm really excited to to hear both of your hot takes here on what the world could look like one five in ten year. for how AI could shape the legal industry, but maybe we can even zoom in further and talk about the clients you work with too. You know, all the doctors, etcetera, all the practices. How does that world change as well?
Yeah, I w I I always feel like when I get the crystal ball out on something like AI that I'm gonna laugh at myself. Future me is gonna laugh at this uh recording. But I do feel like in the next year or so that it's going to be a better version of what we have now. So maybe you're in our world it is you're getting second and third year level production.
out of out of it with good prompting. As I start looking out into the future, I have to believe that there's gonna be continue to be big jumps in its learning and its ability and that, you know, AI is going to be able to do a lot of the tasks. that lawyers do, the projects, the it will be able to draft documents, et cetera. And so I think it's gonna change and I know it's gonna change the
the role of attorneys within a law firm. I think that there is a place for attorneys. Like there's Yeah, I feel like it's gonna put a premium on the human connection and the human element for all of us because everyone's gonna fa every industry is going to face this. Uh and so you that connection, the trust, the uh the nuance of how you lead a client through their journey for us in the business of medicine or the business of some other healthcare industry will
We'll still be there even ten years from now. But our role will be different. We'll have, I believe, just I can't even really fully project what level of what it'll look like to draft documents. It it may be something that's like the push of a button. I don't know what it is.
Yeah, I mean I I'll agree and then we won. one five ten was your question. I I think, you know, knowing w little time we have left, I think right now, in the next year, they it's gonna be better at maybe helping draft documents and research things more properly. That will be like that first layer. Second layer is and I don't know the terminology is
An email comes in, a question is asked, it can start developing responses more quickly. Ultimately there still needs to be an attorney to then analyze and make sure it's correct. So speed of responses, which will then start making clients more curious about hourly world. So I think the hourly world's gonna have a lot of trouble.
with that model all of a sudden, which is why going what you had said earlier, either fixed fees or the or subscription model become more clients gonna be start pushing back and asking for that. I know certain states if you are using for medical practices and others that if you're starting to use the an AI tool, you have to start having to disclose it. And then finally to Michael's last point, which is I think the the role was shift more of as a guide.
than anything else as an attorney because they're not gonna be expecting you to no, they're gonna want good documents and that you understand what those documents are and what do they mean, but they're not gonna expect every attorney to sit there and draft it from scratch to beginning. They know that
your guy, your your giving what's their vision, how do we incorporate that? How do we make sure they understand the risk associated with it and then deliver it? So it's a it's a a lot different role than, you know, when we first start practicing to where we are. That doesn't mean that lawyers will go away or most professionals will go away. I think it just ends up being
the expectation of the speed and eventually I think I kind of think of like email. When I first started practicing, um I was still getting faxes from other attorneys. And I'm like, what are you faxing me for? What's that's what email's for? Or
they we call it red lines. People would still write on a document and fax it to you. I'm like, No, no, just red line in a document and send me your changes. And that's so standard now. If if someone sent you a fax uh by the way, I don't even know if we have a fax number here, audience members, but
Like I wouldn't even know what to do with that. I don't know. I wouldn't even know how to fax you something. So that's that's a shift that has happened of fifteen, twenty years ago and I think that's where AI will eventually fall. Do you think it will get
get to in within ten or twenty years perhaps like a doppelganger of each of each of you that will be able to do some sort of tasks on your behalf.
Maybe so. I mean you're you're you know, that's the part that's hard to picture is the you know, the actual duplication or the doppling or And there may be a world for really small businesses that can't afford legal counsel, the people that go to a legal Zoom where they get a much better version of of that through uh you know through a a fake brat, a robot brat.
But yeah, I mean and you hear that in in the medical world and surgery, I mean that the you know, optimist is gonna be able to perform surgeries within three years, according to Elon.
Yeah. And I just want real quick and I know we gotta go in a second, but um the part that fac fascinating, Legal Zoom came out like fifteen, sixteen years ago, maybe longer now. And when it first came out, everyone's like, It's gonna kill your business. And Legal Zoom's been a boom for our business because people who don't go use it, they use it improperly. They set this up in the wrong document, the wrong entity, whatever it is.
And so going back to the guidance side, that's always gonna be there. And to Michael's point, if someone doesn't wanna hire a lawyer and they wanna try it, you know, do it themselves, they have exposure down the road and obviously the not obviously the job of a lawyer is to help reduce risk.
And however that's done, either via conversations, via documentation, via conversations with your medical boards or or for the federal government, that's our entire job. And so that's why I've never been intimidated by legal zoom and had a problem with legal Zoom.
And so it's just another tool that's out there that people can utilize. In fact, I've actually told a client once when he was trying to do something with his brother Uh he's like said I saw something on legal zoom where I could just have a a note between a family members like yeah just use it like it's it's not because we're guiding them and or anything of that sort that members we're guiding them but we don't they don't need us to draft that.
¶ Personal Insights and Recommendations
Super interesting. Well, last question before you hop off here. Favorite podcast or favorite book or both?
Wow. Favorite book was a book written by someone that we met actually was Twenty One Secrets of Million Dollar Sellers from me. It's a great book of it kinda solidified when we started our firm almost eleven uh actually eleven years ago, uh on February first.
that we had a lot of good visions about how to run a successful business and it solidified a lot of things and then gave us some good ideas. So I always remember that book really well from a business side. That was a great business read for us. I mean obviously the one two threes with Bernard is probably one of the best podcasts in the nation. Um but I love Joe Rogan. I've I've I I'm a big fan of
his. I would I'll I've been reading Strong Ground by Brene Brown and it's been a fantastic book on leadership and uh I have to go with something a little less heavy. I'll go with smartless.
That's low smartless.
Liz the podcast. Yeah. Nice. Well gentlemen, thank you so much for your time and join me on the show. Have a great rest of your day.
Yeah.