.png)
Law Have Mercy!
Law Have Mercy! isn’t just about the law anymore—it’s about life, business, health, and everything that sparks curiosity. Join Personal Injury Attorney Chaz Roberts as he dives into candid conversations that mix legal insights with lifestyle tips, entrepreneurial wisdom, and personal growth. From breaking down complex legal issues in simple terms to exploring the challenges and triumphs of health, business, and beyond, Chaz brings his unique perspective and passion to every episode.
Whether you're here to learn, laugh, or find inspiration, Law Have Mercy! has something for everyone. Just remember: the opinions of our guests are their own, and nothing on this podcast is legal advice or creates an attorney-client relationship—it’s all about entertainment, exploration, and empowerment. Let’s make it fun!
Law Have Mercy!
The Power of Legal AI: How Alexi Ai Transforms Law Firms
Artificial intelligence is revolutionizing legal practice, but as personal injury attorney Chaz Roberts and Alexi Ai Director of Revenue and Former Personal Injury Attorney Daniel Diamond discuss, the transformation comes with both tremendous opportunities and potential pitfalls. While generic AI tools might hallucinate false case citations, specialized legal AI platforms like Alexi deliver research in minutes that once took hours, with verifiable sources lawyers can trust.
The conversation takes us back to the early days of Boolean searches and stacks of paper files, contrasting them with today's natural language queries that instantly extract relevant precedent from vast legal databases. Through practical examples, we see how AI transforms critical tasks: from finding jurisdiction-specific shoulder surgery verdicts to extracting chronologies from thousands of medical record pages, creating motion drafts with proper citations, and even uncovering contradictions in expert witness testimony across multiple cases.
Perhaps most compelling is the philosophical question at the heart of this technological revolution: what remains uniquely human about legal practice when AI handles the repetitive tasks? The answer becomes clear as Diamond and Roberts explore how AI frees attorneys to focus on relationship-building, persuasive advocacy, and strategic thinking—the elements that truly define great lawyering.
Beyond the practical applications, the conversation takes unexpected detours into creative AI uses, fitness tracking, sauna recovery techniques, and how technology enables better work-life balance. Whether you're a solo practitioner overwhelmed by research demands, a mid-sized firm looking to scale efficiently, or simply curious about how AI is reshaping professional services, this episode offers valuable insights into working smarter while preserving what matters most.
Ready to transform your practice with legal AI? Discover how domain-specific tools can enhance your efficiency without sacrificing quality. The future of legal practice isn't about replacing lawyers—it's about making them superhuman.
Check out Alexi Ai and how it can change your law practice in an instant!
You can watch most full episodes of Law Have Mercy on YouTube!
For more FREE legal tips, check out our social channels:
Instagram - @chazrobertslaw
Facebook - Chaz Roberts Law
TikTok - @chazrobertslaw
LinkedIn - Chaz Roberts
If you are in need of legal guidance, visit our website: https://www.chazrobertslaw.com/
This show is co-produced by Carter Simoneaux of AcadianaCasts Network, Chaz H. Roberts of Chaz Roberts Law and Kayli Guidry Bonin of Beau The Agency, and Laith Alferahin.
Not all AI is created equal. Right, don't throw the baby out with the bathwater. It's like saying, well, because some cars get in car accidents, I'm not ever going to drive a car, I'm going to walk to work. I'm going to walk 40 miles to work every day and not take the more efficient route of getting in a car and driving to work because I'm worried about car accidents. Well, I mean, that's one way to do it. But if you drive safely and other drivers drive safely and we have a licensing program and cars are built safely and you know what to use it for and you don't drive it into a lake, then you're going to do all right.
Speaker 2:All right, I'm out of breath right now. How about you?
Speaker 1:Yeah, I'm out of breath from doing more pushups than you, Chaz.
Speaker 2:Chaz Roberts back again with another episode of Law have Mercy. I have my newest friend, Mr Daniel Diamond. He is the head of revenue.
Speaker 1:Director of revenue, head of revenue, it's all the same.
Speaker 2:Both good titles, Both good For Alexi, which is a legal artificial intelligence company. Ai You've heard about it. It's something that I've been super passionate about how can AI help our law firm? And you know, it seems like every day you hear about AI with ChatGBT and Grok and some of those other. It seems like every day there's a new company coming on board. Alexi is really cool because it's specifically built for lawyers and mid-sized to small law firms, especially like me. It's an affordable solution. So I enjoyed the product so much. I came out here, I met with Daniel, we talked about it, we chopped it up, and so we want to just talk about kind of what it does and what it's about, and maybe you can educate me a little bit more about how better to use it.
Speaker 1:Absolutely. Yeah, I'd love to. Ai is crazy. It's super cool. From the first time I saw it at Alexi in the legal context and the first time I played around with ChatGPT, it feels like magic. It really there are few moments in life where you see something you're like this is sorcery, and AI is one of those things, and it just continues to impress me. Every time a new feature is dropped and a new product is launched. It, um, it's wild stuff.
Speaker 2:Yeah, and I mean I'm. I feel like we're kind of late to the game. I'm as a consumer. I feel like it's only come in to my world within the probably last year, but this is something that you guys have been working on for years.
Speaker 1:Yeah, we've been around since 2017. I joined the company in 2020. And we started as a legal research product and then we expanded into some other use cases, like document review. But when I first saw what it could do for legal research a task that would take me, as a former lawyer, six, seven hours to find the right cases is now taking a few minutes. I type in a question, I get an answer. It's really pretty incredible.
Speaker 2:Yeah, so you're a lawyer and I'm a couple of years older than you are. I remember when I was in law school we had actual books and that quickly like went to was it LexisNexis Westlaw? But you still had to learn a whole class on how to find the cases and you have to plug in the jurisdiction and all kinds of different things. Ai kind of gets rid of all of that?
Speaker 1:Yep, yeah, exactly. So when it comes to legal research, the old way of doing things I guess, beyond books and libraries, once you move digital is Boolean searches. Right, that's what most people are learning in law school and it's how that word? That's the word I was looking for. Ooh, what most people are learning in law school and it's how to that's the word.
Speaker 2:That's the word I was looking for. Ooh, I haven't heard that in 15 years.
Speaker 1:It's a strange word and outside of legal circles and a few real geeks or nerds out there, not a lot of people know what Boolean is. And it's the art of structuring a search in a way to get the optimal result. And so there's exclamation marks and brackets and slash and numbers and and it's, it's. It's not quite coding, but it it feels like you're almost doing some sort of coding when you're doing a Boolean search and then it doesn't stop there because you get you know you might get 350 hits. You're a personal injury lawyer. You want to look up comparative negligence in your state. You punch in okay, comparative negligence plus car accident, and you know, maybe you filter it. Cases from the last 10 years. You got 350 cases. Great, what's what's next? Right? You got to filter through those. You got to figure out which ones are the most relevant. You got to.
Speaker 2:You got to spend hours and hours trying to find relevance basically, well, we're going to get into some of those functions later, but I can tell you this On the plane ride here I was working with Alexi on drafting a demand and I said give me shoulder surgery cases with arthroscopic surgery for a rotator cuff tear in Louisiana and it came in boom boom, boom and I was like I need higher awards. It said, okay, yes, sir, boom, boom, boom and I was like, okay, I got it. I mean, that is a long way away from what you're describing and I think the whole, like me, inputting the information took a matter of seconds. The how long. And I was on an airplane internet connection. I think I received results within about 45 seconds to maybe a minute minute 30. It was incredible.
Speaker 1:Yeah, yeah, it's wild and it's funny because you still get people who get upset. Oh, it takes a full minute. Well, the old fashioned way would have taken you, you know, three hours, but yeah, getting a response in 45 seconds, that's amazing.
Speaker 2:Well, the old-fashioned way. You didn't know if you had, you don't know what you missed out on because you didn't put the right search terms in. Like, maybe your formula was off and maybe you excluded some things that you could have used. Because, man, you're bringing back memories now with the old Boolean search. You don't know how accurate the information you have is because of how good you were as an inputter of the information. Now I'm just putting plain language and I'm having a conversation with the software. Like no, I want more, I want higher, I want, can you add in a forearm injury, and it just yes, sir.
Speaker 1:Coming right up. Yep, exactly, you talk to the AI the way that you would talk to a colleague or a coworker or a junior associate, right? That's the goal, because lawyers use natural language in their everyday life. They shouldn't have to become a computer programmer and learn how to speak Boolean right. When I was in law school, a professor of mine said dentists have their drills, doctors have their scalpels, but lawyers have their words right. So being able to use that muscle memory and just use natural language to describe what you want, that's what it is to be a lawyer.
Speaker 2:How did you get from, how did you go from being a practicing lawyer to joining Alexi?
Speaker 1:Yeah. So I worked at a personal injury firm for six years in Canada. I'd say I loved the first five years. It was fun. It's the ups and downs. I was bought into every moment. It was great.
Speaker 1:And then in the sixth year practicing I found it just wasn't the best personality match. I was bringing home that kind of combative energy into the home which I didn't like, and at that point I had three kids and it wasn't't like. It's not so much a team building, relationship building exercise. Every day you're in fight or flight mode as a PI lawyer and you have to have really thick skin and it wasn't a perfect match for me. So during that last year of practice I was thinking you know what's a good way out for? Is it to become a transactional lawyer and be drafting contracts? Is it to do some other area of law?
Speaker 1:And I kept getting the idea in my head that I would love to try my hand at sales. I think that's a really transferable skill as a litigator is the ability to build relationships with people, the ability to um and just that human to human contact that you have in law and in sales. And I came across Alexi and I loved what they were doing. I met with the CEO and he showed me the product and I was hooked and the rest is history. I joined kind of in a low level sales position, initially just making cold calls and sending cold emails, and five years later director of revenue and haven't looked back. It's been an amazing fun ride that the software enables you to do right, I sound like the salesman now I just I'm a fan.
Speaker 2:It's like what you just said is on the money. You know you, we spend time looking at the worst in every possible situation. It's actually malpractice to not look at the worst in every situation as lawyers and that's very draining. And you know you, I like to say as a personal injury lawyer, when you're sleeping people are putting sticks of dynamite in your case, including your own clients. Right, could be your one social media post away from going to zero, yeah. And so you know it's a lot of contact that you have to have with clients and you know your opposing counsel and your staff to help move the ball forward. And so it's like any advantage that we can get as lawyers to help lighten the load in other areas so we can concentrate on those relationships. It's a huge, it's a huge advantage, it's a huge win, and that's kind of why I'm so excited about this.
Speaker 1:Yeah, it's almost philosophical to think about what are the core aspects of what it means to be a lawyer. And now that AI is stripping away some of the peripheral stuff, some of that repetitive BS like legal research and document review and all the stuff that lawyers don't really want to be doing, what's left when you strip that all away? And that's the core of being a lawyer. It is relationship building, it's managing client expectations, persuading a judge or a jury or opposing counsel, and that's the stuff that most lawyers want to be doing. They don't want to be stuck in an office at 7.30 pm missing their son's soccer game because they have to look up what the law is.
Speaker 2:Yeah, especially not this lawyer, right? Because 4.30, I have to go. I have Laith likes to say, I manage a business at home and it's my family, so I'm at the ballpark every single day. That is a non-negotiable. And I coach most of their sports non-negotiable, and I coach most of their sports. And so you know I just I'm not going to stay behind the desk doing Boolean search terms till seven o'clock, right.
Speaker 1:There you go, you got your priorities straight.
Speaker 2:I think I know the answer based on that, but do you do you think that lawyers will ever be replaced by AI?
Speaker 1:It's a great question. I don't, and I think it does tie back into one of the core aspects of being a lawyer, and I think there's parallels here between practicing law, or at least litigation, and having a job in sales. I don't think AI will ever be able to do sales, either to take a certain course of action or to manage their expectations or to talk them off the ledge. That requires a human-to-human touch.
Speaker 2:It requires the ability to build relationships and to exhibit some degree of warmth and EQ, and so you can have AI do all the IQ stuff and all the knowledge-based tasks that you previously did, but you'll never be able to take away that human touch. Yeah, I never thought of it as sales, but from the moment we acquire a client, we have to sell them on our services. And then, all through the process, we have to sell them on trusting us, essentially, that we have the right path. Having that difficult conversation to settle or take it to trial is a sales pitch. Essentially. That sounds bad, but it is what it is right. And then ultimately to take it to a judge or jury, you're selling your story. You're selling your client's story, you're selling the case, and that requires a separate part of your brain than doing legal research.
Speaker 1:Agreed? Yeah, totally. And AI can tee up all those pieces of the puzzle for you. It can give you the knowledge that you need. It can give you the cold hard data, the objective facts. Right, here's what's going to happen if you try and settle your case now. Here's what's going to happen if you take the case to trial objectively, cold, calculated. But that human element of talking to your clients, sitting down and saying, hey, here's what your life's going to look like if you settle. Here's what your life's going to look like if you take this to trial. Here's how it's going to feel. You know, here's in the dozens of trials I've done before. Here's how it's going to play out. Right, and allowing the client to feel the emotional side of the decision that they need to make rather than just looking at the cold hard facts a legal specific domain like Alexi.
Speaker 2:Why not just use ChatGPT or GROK or any of those other platforms? Why use a legal specific program? Because those other ones are pretty good too.
Speaker 1:They are. Yeah, they're really really good for certain tasks. Just like you wouldn't ask a chef for medical advice, I'd say I wouldn't ask ChatGPT for case law. And it boils down to there are just different tools that are great for different purposes. If you're trying to do something creative, if you want an amazing banana bread recipe, if you want to write poetry, if you want just something in the realm of creativity I don't know if you've experimented with ChatGPT's image creation tool Maybe we should try it. It's really cool. But for creativity, for pulling something new out of thin air that's never been created before non-domain specific, you know, general purpose AIs are really really good for that. Ai's are really really good for that.
Speaker 1:But a lot of lawyers have learned the hard way that if you ask chat GPT for case law or anything that requires domain specific knowledge, we're not even in the legal space. If you're a doctor, let's say, using some AI tool trying to diagnose your client and you're using chat GPT, the stakes are just too damn high to rely on a tool like that. And so some lawyers have learned the hard way. They got fake cases from chat GPT, some cases that it spits out from the future. It's pulling cases from the future and lawyers have unfortunately relied on them and put them in their briefs without second checking these cases to make sure they're real. And they go in front of the judge and the judge says these aren't real cases and there are a lot of lawyers who are getting suspended and in trouble and you know, some have been disbarred, I think. And it's funny because that then causes the market in general to move in the other direction and it causes people to say, well, I don't want to use any AI. Ai has betrayed us. It's making up fake cases.
Speaker 1:Well, not all AI is created equal. Right, don't throw the baby out with the bathwater. It's like saying, well, because some cars get in car accidents, I'm not ever going to drive a car, I'm going to walk to work. I'm going to walk 40 miles to work every day and not take the more efficient route of getting in a car and driving to work because I'm worried about car accidents. Well, I mean, that's one way to do it. But if you drive safely and other drivers drive safely and we have a licensing program and cars are built safely and you know what to use it for and you don't drive it into a lake, then you're going to do all right, you probably won't get in an accident, and so, yeah, just to walk this back a little bit. So, with domain specific AIs like Alexei, it's trained on legal data, it's trained on authoritative information, it's trained on the law.
Speaker 1:Real cases Real cases Real laws, real cases, exactly. And so there are guardrails in the pre-training and the fine-tuning of these programs that make them such that they're not going to make up fake cases. They're not going to that term. Is hallucinations right? They're not going to hallucinate like a hippie at Woodstock. It's not going to happen, and one of the ways that a tool like Alexi will show that to our user base is the answers that we give. We're also going to give a hyperlink that they can click on. We're going to back up our answers with sources. So here's the answer to your question. Here are the cases. Click through it. You could see. This is a real decision. If we're answering a question about document review, you upload some documents. You ask a question about your documents. Here's the exact page of your own document set.
Speaker 2:That's a really cool function. I was playing with it earlier. Here's the exact page of your own document set. This is sweet.
Speaker 1:Yeah, yeah, it's cool, cause it's one thing to just say, okay, trust our AI, get used to it, you'll see that it's pretty accurate. It's another thing to say, okay, here's the answer. We, this is an accurate answer, but if you don't believe us, here's where you can go find the answer and see that this is legitimate.
Speaker 2:Yeah, so the the general domains, right, like the chat GBTs. They really want you to use it more and stay on the platform, just like every other platform, and so it's trying to satisfy you. If you say I need this, this and this, it's going to maybe even make it up so that you stay on the platform. And these lawyers that were suspended or even disbarred I mean how reckless, right to number one rely on just that, without checking it yourself. But you know, what we were saying earlier is that the way I use AI, including Alexi, is that I'm the driver. I am driving the car. Ai is in my passenger seat giving me the directions, but I am driving the car. Ai is in my passenger seat giving me the directions, but I am driving the car. I am in control at all times, and to me, that's a winning formula.
Speaker 1:Yeah, exactly, and yeah, it's like a GPS right. There's that scene from the Office where Michael Scott follows the GPS and it tells him to turn right and there's no road and he just turns right blindly and goes into a lake. And it should be the same with AI. If you're using chat, gpt, check everything that it's giving you. Don't just blindly follow what it's telling you.
Speaker 2:I was hoping it was the it's Britney bitch scene.
Speaker 1:That's a good one too.
Speaker 2:And, by the way, if you're listening to the podcast, you may want to check out the YouTube. Daniel is a very handsome guy and I don't say that about many men.
Speaker 2:I appreciate that Likewise. All right, so what do you think are the top three functions that help personal injury lawyers with? Let me do that again. Top three is it a function Use cases, use cases. You can say use cases. What are the top three use cases that benefit that you think will benefit personal injury lawyers? Daniel, what do you think are the top three use cases that could benefit personal injury lawyers?
Speaker 1:I'd say it generally falls into one of three buckets that our customers are looking for help with, and most people who are seeking out AI in the plaintiff personal injury space are looking for help with it's legal research, document review and drafting.
Speaker 1:Those are the three general categories and legal research we talked about a little bit. With legal research, lawyers are looking for the right cases and there's different types of research you can do with AI and it's not always just looking for case law. Sometimes personal injury lawyers want to know if they have an expert on the other side who's assessed their client let's say, an orthopedic expert and you want to dig up all the dirt on that expert as you possibly can, because you're going to cross-examine them at trial. You can come to a platform like Alexi and you can ask for case law where their evidence has been accepted or rejected. You can find any information on the open web through Alexi as well. So it'll pull any articles. It'll pull academic articles that maybe they've journaled and then what you can do is kind of cross-reference all that stuff and you can ask for legal reasoning as well and you can say, based on what you know about my case from all these documents I've uploaded, find me information that I can use to cross-examine this expert.
Speaker 2:Oh, dude, I'll be right back. I got to go do some work. I didn't even know about this function.
Speaker 1:Yeah, yeah, and I guess I'd put that under the legal research bucket. But it also kind of touches on the drafting function, because if you've got a cross-examination coming up, you can take all that information, the documents you've uploaded, the case law you've pulled, the open web search, that you've done and you can say, well, now draft me a list of questions for the cross-examination of this doctor that you think will hurt their credibility.
Speaker 2:All right, so I have a cool story. All right, so I have a cool story. The first time I used the software, we have a client who we represent the family of a woman who was unfortunately killed, and the person who killed her in an automobile is facing criminal charges. Okay, and so we really want to depose him. Take a deposition, take his deposition so we can move our case along. But he has a fifth amendment, right against self-incrimination. Well, we have a right to a jury trial, and so I put that information into Alexi and I said give me a legal memorandum on the interplay between the two and tell me if I have a right to take his deposition.
Speaker 2:I was a little rusty. I'm sure I researched this issue 10 years ago, right, and it spit it out Boom, you have a right to take his deposition and gave me all the cases and you know, if they assert their Fifth Amendment, right against self-incrimination, it's a. It can be used against them in a civil proceeding. You need the DA's approval. Blah, blah, blah.
Speaker 2:I was like this is fantastic, this, this was a legal memorandum that a law clerk would have taken a week to draft, including the research at least a week, and I don't know how much money it would have cost to get that memorandum least a week. And I don't know how much money it would have cost to get that memorandum. Then I said provide me a draft of a motion that I can file with this information. Boom, it spit out. Now, look, I got to take it, I got to put my magic in it, I got to put my words in it. But all the bones of the memorandum, the motion and the supporting memorandum, are there. I cross-referenced all the sites, all active good standing cases in Louisiana, and I thought that was the coolest thing in the world.
Speaker 1:Yeah, I'm glad you brought that up. I think the way that AI is meant to be used in the legal domain, or really any domain, is you called it. You know it gives you the bones right. We always say it's here to give you a jumping off point or a starting point for your legal research, for your document review, for your drafting. And there's a funny expectation.
Speaker 1:There's this irony where lawyers don't want AI to take over their jobs, but then they're also kind of disappointed that it can't entirely take over their job. When they try it, and as long as there's understanding going into using an AI, that this is going to get me from zero to seven, zero to eight. You know, is it always perfect? No, but it's going to give you a really solid jumping off point that you can then take and add your own flair to and figure out how to tweak it to the exact nuances of your case. And that's the goal. The goal is not to be an end to end solution and to completely replace everything you do so you can sit on the beach drinking margaritas. The goal is to just take some of those rote, repetitive, frustrating tasks, those day to day, and leave behind what it is to be a lawyer the core aspects of persuasion and legal writing and, you know, managing client expectations and stuff like that.
Speaker 2:Yeah, so I agree with everything you said. But it's not just the road, it's the novel, right? So this was a novel issue that I hadn't had in years and years and years. That you know. The problem is, a lot of times we put search terms that would bring up thousands of entries, because it's the same language that you see like self-increase. Fifth amendment, fourth amendment Well, there's millions of cases, but the more terms that I was able to put in by just explaining in plain language, it gave me exactly what I needed, and if I had to do my own research it would have taken me I mean, I don't even know, I'd still be there.
Speaker 2:So, to me it's the road stuff, but also, like the novel stuff. You know the jumping off point that you alluded to.
Speaker 1:Yeah, that reminds me of a story that one of our customers shared with us that he was arguing a motion in court and opposing counsel raised a novel issue and said because of this novel issue, we're going to have to adjourn this motion.
Speaker 1:You know, maybe three, six months, and you know court systems backed up everywhere this is up in Canada and Ontario and I think another date for the motion would have been six months. Down the road and our customer was pretty pissed off about that and the judge asked this lawyer who was raising the novel issue well, do you have any cases to back up what you're saying, any cases to back up what you're saying? And he didn't have any cases and he didn't have Alexi. And our customer, um, who was on the other side, said you know what? Your honor, give me five minutes. I'm going to see if there's any cases on on this issue. And he took a break and he came back into the courtroom after using Alexi for five minutes, found some fantastic cases exactly on that issue and it supported his case and completely destroyed the other party's case for the fact that they needed an adjournment. And he won the motion and they didn't get an adjournment.
Speaker 2:They argued it right there and then and he won the motion he had access to cases in the span of a couple of minutes that opposing counsel didn't yeah, and in my career, the judge like within my career within the last 15 years what normally happens is the judge says hey, why don't you guys brief the issue and submit a brief within 30 days and then I'll reconsider? Now it's five minutes. Bring your laptop to court and you can pull something up in five minutes. Hey, judge, you have a printer. I could hook this thing up to.
Speaker 2:Yeah, exactly that's really cool. I feel like I get advertisements constantly with other types of legal AI and I was curious what makes Alexi different than some of those other providers.
Speaker 1:Yeah, that's a good question. So there's so many players in the space right now, there's a lot of really small players who are springing up out of nowhere, and what a lot of them are doing is taking chat GPT and building kind of a layer on top of it to mask the fact that it's really just chat GPT, that they've done maybe some fine tuning to, but we call that a GPT wrapper and there's hundreds, if not thousands, of GPT wrappers out there and it makes it very confusing for the consumer.
Speaker 2:There's a lot of smoke and mirrors, so they're using the platform in the background, but they're putting something else over it to appear as that they're their own AI Exactly and there's no underlying knowledge base that they built up as far as, let's say, case law or legislation.
Speaker 1:There's no retrieval systems that they built to find the best, most relevant, most authoritative answer to a question. There's repackaging, chat, gpt, so there's a lot of those. It's pretty easy to suss those out. And then you have some of the bigger players in the space who there's kind of two different approaches. There's the Swiss army knife approach, which some companies are taken, which is okay, let's try and do everything. Let's try and do legal research, document review, drafting, transactional stuff, contract redlining, contract drafting, right for litigation and transactional. They're trying to be everything to everyone.
Speaker 1:And then you have really really niche companies who have AIs that are just tailored to personal injury firms or just tailored to tax law firms.
Speaker 1:And then you have companies that are somewhere in the middle and I'd say Alexi is kind of somewhere in the middle where we have a nice breadth of use cases that we do, but we've really honed in on the two or three that we do and we do them very well.
Speaker 1:So we do legal research exceptionally well, we do document review and answer questions about your own documents exceptionally well, and drafting is something that in the coming months, we're going to get really, really good at. The cool thing is the way that we built our software. So when you start with legal research, you have an AI that understands the law. Then you build some functionality that it can ingest documents and now we can understand the facts of your case. It can understand the evidence. Then the third piece of that puzzle, which is what we're building now and we have some of it in place, it's getting way better in the coming months is drafting, which has really taken both those pieces of the puzzle. It's taken the law, it's taken the facts of your case, and now we can turn that and communicate it into some sort of draft that articulates it in the best way possible.
Speaker 2:Yeah, so currently it's assisting you more with breaking down the chronology, breaking down the law, breaking down even. It read a police report and told me the operative facts of the police report.
Speaker 1:But as it evolves it's going to get better at putting all those elements together in an actual draft Exactly, exactly. Yeah, you mentioned chronologies. That's really cool use case as well for personal injury lawyers.
Speaker 2:So it takes a lot of time, and it takes a lot of time to draft a chronology for sure, or very expensive to hire nurses to do it or a paralegal to do it.
Speaker 1:Yeah, yeah. And you've got tens of thousands of pages of hospital records. And I mean, if this is an accident from half a decade ago, I'm sure you've got treatment records in dozens of different boxes, mountains, mountains. And reviewing that the old fashioned way, trying to flip through page by page. Okay, you know, the client said that they had a brain injury. Let's find evidence of that. They said they had a fracture. Where is that in the records? Well, now you can upload that to Alexi and you can say, across all these tens of thousands of pages of documents, please tell me if there were any fractures identified and when and where, and give me a chronology of of all the treatment. So it's very, very cool stuff. Do y'all have?
Speaker 2:lawyers that are doing quality assurance and checking on the responses.
Speaker 1:We did for the first five or six years of the company. When we started with legal research, every research memo we provided had an additional review from a human lawyer until the AI got to the point where it could autonomously answer questions and provide instantaneous answers. And now, for privacy and security and data security reasons, we don't have visibility into everything our customers are doing. But the way that we train our AI, we train it on synthetic data. So we do have a team of lawyers not looking at actual customer questions and answers but asking their own questions and answers and trying to anticipate what our customer is going to ask and feeding that synthetic data into the AI to continue training it.
Speaker 2:One thing that I do appreciate is at the bottom right-hand corner of the platform it says all of this information is confidential, and I don't think, like on a chat, gbt, the stuff that I input, that's not confidential, it's stored somewhere and that scares me.
Speaker 1:Yeah, it should scare you. It's a scary topic because they're. So I think if you're a consumer looking into AI platforms and if you're a lawyer and you want to comply with law society obligations, you have to be asking the right questions to the vendors that you work with. And just like you go to the grocery store and you can look for certifications like gluten free, dairy free, whatever it, the same thing does exist in the data security space, and one of those is called SOC certification or SOC 2 certification. At Alexia, we're SOC 2 certified and that's a question you can ask your vendor Are you SOC 2 certified?
Speaker 1:So that you don't have to become an expert on data security and ask 50 different data security questions. You can just outsource that to a company that puts their rubber stamp on it like SOC 2. And so that's a question that you should always be asking. And some other questions that you can ask the vendors you work with are you know where is the data stored? Is it getting used to train your AI? Is it getting sent to any third parties to train those third parties? And you have to be really cautious about confidentiality and privilege, and especially when you're a lawyer and you're dealing with really personal privilege documents. That's something you got to be thinking about and HIPAA violations.
Speaker 2:Absolutely, medical records, I mean, that's the highest level of security, that privacy that you can expect. People and you're you're dealing with their medical records, yep. And so if you upload them on some AI platform that's not certified secure, you could be you could be getting yourself in hot water. Yeah, that's pretty scary, yep. What's your favorite thing to do non-legal related with AI?
Speaker 1:Sometimes at night, when I'm laying with my kids and putting my six-year-old to sleep and I don't have the creativity to come up with a story on my own, I can just pull out, chat GPT and ask it to come up with a bedtime story. Or really cool use case is coming up with pictures, just like wild anything you can think of, just the silliest pictures in any style. You want to do it in the style of Claude Monet? You want to do it in a watercolor? You want to do it in Picasso style? We should try that, yeah let's do it, you want to do it yeah.
Speaker 2:How do you do it in Picasso style? We should try that. Yeah, let's do it. You want to?
Speaker 1:do it? Yeah, how do you do it? Yeah, teach me, teach me wise one chat GPT. So I'm going to, I'm going to dictate here. Um, I'll do one, then you do one. Make me an image of a lawyer holding a briefcase on the back of a T-Rex, riding through outer space and holding a laser gun in one hand, battling space aliens. That's quite the imagination. I can guarantee you this has never been created before. And that's, you know, the hallucination topic. Hallucinations are actually a feature of chat GPT for non-legal purposes. It's not. It is a flaw if you're using it for legal, but it's a feature. You want it to create things that have never been created before.
Speaker 2:Is that what's the definition of hallucination? Whoa.
Speaker 1:Can you see that lathe?
Speaker 2:That is cool. Yeah, yeah, you want to try one Sure. Just hit the dictate button. I'm a lawyer so I'm not very creative. Just the dictation button.
Speaker 1:Yeah, that should work, I think.
Speaker 2:Yeah, that should work, I think, in outer space, while holding an artist, a paintbrush.
Speaker 1:Do you have a shaggy dog?
Speaker 2:I do. Holding an artist and a paintbrush Creating image, we have something Damn. This is good, this is.
Speaker 1:Oh my gosh. Tap on it to go full screen. Yeah, show the camera.
Speaker 2:Is that insane?
Speaker 1:Yeah.
Speaker 2:And so you can just spend hours playing with your kids, changing up the scenario, making a story to go with it. You can say now give me a short story, a three paragraph story in the writing style of Ernest Hemingway, to explain what's going on here?
Speaker 1:Yep, exactly, and that's exactly how a tool like ChatGPT, a general purpose AI, is designed to be used. Don't ask it for legal research, don't ask it for case law. Don't ask it for domain specific, authoritative knowledge. Ask it to do silly, creative tasks like these, and it'll do them really well creative tasks like these and it'll it'll do them really well.
Speaker 2:I had something this is this blew my mind we went to, we went skiing for Mardi Gras um, the Mardi Gras holiday I don't think you guys celebrate that, but it's um, it's a good time. We like to get away from Louisiana because things are kind of crazy back home with all the partying. So we go skiing. It's a great time on the slopes. And so I'm in the back seat of the shuttle going to the mountains and I'm like hey, make me the perfect itinerary to spend time with my wife, my kids, and here are the inputs Boom, boom, boom, boom.
Speaker 2:And I was like my wife and I plan on skiing the third day. The kids are gonna be in ski school on Monday and Tuesday. We can all ski together. On the third day we plan on eating at these. The AI said you should really consider skiing with your wife on the second day, so you have alone time together on the mountain. And it was right. My wife and I had the best time ever that second day, and then me and the boy. She was only skiing one day, so me and the boy skied the third day and we cut up the mountain had a blast and my wife was chilling, packing like enjoying her day, shopping or whatever. But my wife and I had a great day on the second day, based on the AI recommendation Wow.
Speaker 1:That's a great story. Yeah, chatgpt can make pictures of space aliens and it can save marriages, can save marriages.
Speaker 2:You're fit, you're in shape. Do you ever use it for workout plans, diet plans?
Speaker 1:I haven't, actually I haven't tried that. That's a really cool idea and I should, but I've used it for like structuring spreadsheets. And if I'm having a problem in Excel and I don't know how to, how to format something or say I, I'm, you know, I'm in sales. So we go to a conference and we get a list of attendees at that conference, but the Excel sheet is all the emails are mixed in with the names and it's all. It's all jumbled together. I could throw that into chat GPT and say hey, can you please extract this into three separate columns First name, last name, email and it'll spit that back out. It'll reformat an Excel spreadsheet for me. So there's a whole bunch of cool stuff that it can do. Haven't tried that though.
Speaker 2:Daniel, tell me about the spreadsheet we were talking about earlier please. Okay, Would you mind?
Speaker 1:Not at all.
Speaker 2:This is a little embarrassing, not for all this is a little embarrassing, but I Not for me I want a copy, go for it.
Speaker 1:So I lost about 35 pounds in the last 16 months or so and part of how that happened was my wife bought me a daily health and wellness journal and I would like a physical journal that I would put pen to paper and document. You know, how many pushups did I do today? How hydrated was I? How many glasses of water did I drink? And so that kind of morphed after I filled out that entire book three months of filling that in but it got too detailed and I was putting all sorts of different activities in it. So when I was done with that, I created a spreadsheet for myself and I have about a dozen, maybe 14, different columns in it. So when I was done with that, I created a spreadsheet for myself and I have about a dozen, maybe 14, different columns in it for different activities throughout the day. How many pushups did I do? How many sets of weights? How many healthy meals? How many units of junk food did I consume? How many units of alcohol did I?
Speaker 2:consume. Did you log the 60 pushups we did before the pie?
Speaker 1:Oh, they're going to get logged, baby, you know it. They're going to get logged. And at the end of the day. So I have a formula that I've fine-tuned over the span of a couple months and it's still fine-tuning that assigns me a score and I know if I'm getting like 7.5 and up, that was a healthy day, if I'm getting anything less than that, not a healthy day. And how many hours of sleep did I get? That's another column on there. And then I can just look at my metrics every month, every quarter. I could see how I'm improving in any given category. And I started by doing like 70 pushups a day. Now I'm at like 110 pushups a day and uh yeah, I don't know how we got on this topic.
Speaker 2:I don't know, but we should call it Daniel hard and we should we should this. This could be the next business this is. Rob Dyrdek stuff. I love it. Yeah, I would say. The only thing I would add to it is like a happiness indicator, like one more column where you subjectively put on a scale of one to ten how happy you are.
Speaker 1:That's on there. It is, that's on there. Emotional well-being, I thought.
Speaker 2:I was reinventing the wheel here. It's on there, baby. That's cool. That's cool. Yeah, I'd love a copy of that. I'm glad you appreciate it. Thank you, yeah, I'm, I'm into all that man, yeah, and then you're also in a sauna, you're? You like a sauna? Um, what do you think? What do you think the sauna does for you?
Speaker 1:Oh, I love the sauna.
Speaker 1:I have a funny story about that after you tell me yeah, so I try to do sauna in the evenings. If I'm getting a workout in like 7, 8 PM, I'll cap it all off with a sauna and then, if it's you know, within two, three hours of bedtime, I sleep like a baby. It's fantastic. I love the social element of the sauna. So the one I go to fits about 15 people it's at my local gym. It's always the same crew of like older guys and you know they're talking politics, they're BSing, they're just that banter. Um, I really, really enjoy that. But on a scientific level, I mean, if you listen to people like, uh, andrew Huberman is that his name Andrew Huberman.
Speaker 1:Yeah, um, you know. They talk about the heat shock proteins that it generates in your body. It's resting cardio, so your heart starts beating. All you're doing is sitting there, but that last five minutes feels like you just ran a race. Your heart is beating out of your chest. So it's just a great way to end a workout and create like pre-sleep activity too.
Speaker 2:Yeah. So I've done sauna my entire life and I didn't really know what it was about. I just go in there, heat up, get a little sweat in and I'm good. I used to do it when I would drink. It was a good hangover remedy. But I never really understood it until Running man. So I did the Running man Festival and there's a huge outdoor sauna. Talk about social aspect. There's a DJ booth in the sauna, but there's these little mini saunas.
Speaker 2:So we run a one course, a marathon, and a one mile course off road uphill. It was brutal. My neck was sore from watching every step. I took rocks, cracks in the, in the dirt uphill, lots of elevation. It was brutal. I didn't know what I was signing up for. I didn't realize it was like an ultra course. Right, I thought it was. I was only used to running on the road, so it just tore me to shreds, like I can show you pictures. There is pain in my face. Next day I wake up. We slept in yurts in the middle of the grounds, in the middle of the one mile course. The next morning I wake up I can barely move.
Speaker 2:Where there's a sauna and I was like, well, let's just go sit in there. Me and my friends sat. We had our. We just had shorts on, no, no shoes or anything, cause I didn't care. I was. I was a broken man. I sat in the sauna for about 20 minutes. When I walked out I felt like a whole new man. It's like it was lube from my joints, man. It was like someone like the Tin man, and someone came and oiled me up and I was like now I understand the purpose of saunas. I had never had that experience before.
Speaker 1:Yes, it's so good. It's so good. Do you feel like you carry some of that health and wellness stuff into your law practice? How does it impact your law practice?
Speaker 2:Well, first of all, I think doing hard things like the marathon, running and stuff builds me up mentally, Mental fortitude, mental strength, and it's long days, it's grueling days. It's grueling because there's 50 challenges that come in the course of a day and so you will wear yourself down and a lot of lawyers turn to drugs and alcohol, bad eating habits, unhealthy sources, to cope with that amount of stress. We were talking about the stress earlier, man, and so the way I deal with the stress is I go get a good workout in.
Speaker 1:That's the right way to do it. I love that.
Speaker 2:And my kids observe me. Man, I've run four world major marathons and they're there and they see me running and they're at the finish line and it's a whole experience we take in the city. We do all things and so I try to make those healthy habits part of our lifestyle and so my kids just naturally gravitate towards that. They say kids don't learn by you telling them things. They learn by through observation. And so I'm always pushing myself and they suck that up, man.
Speaker 1:You and I are going to get along, man. I love that philosophy. I think the same way. And my father was a lawyer, is a lawyer too, and he ran a personal injury firm for like 30 years and, uh, he was so immersed in his work and I'm so grateful for that because he you know, we never had to worry financially. He, he always provided.
Speaker 1:But the consequence of that was he didn't really focus on health and wellness or exercise or any of that stuff. And then, after you retire, I mean maybe we can edit this. I don't want my dad hearing this stuff, but just between you and I, like he he's, he turned 70 this year and he's not a healthy person. He never stacked those habits. He never got to a point where, like in retirement, he's now going to the gym and exercising. He just he doesn't have that habit. So I'm trying to build that up. I want my kids to see it, I want it to be something I do forever into retirement.
Speaker 1:So yeah, it's a, and it's funny because people will like bifurcate that and they'll say, well, no, health and wellness stuff and exercise is totally separate from your professional obligations. I think one feeds into the other. I think you have a professional obligation. If you want to be the most optimal employee that you can, you want to be the best lawyer you can, you should be hitting the gym, you should be exercising, you should be trying to do hard things like sauna. Physically. Lawyers are doing hard things every day mentally, emotionally, but getting that physical aspect to complement it as well, it carries into the work.
Speaker 2:You have to condition yourself. And I have a question and I'm curious You've seen your father practice law. You practice law. 20 years ago it was fax machine and no cell phones. Do you think it's harder or easier to practice law now Because in some respects the technology makes it so much easier and more efficient, especially with AI, I mean, this is a whole new ballgame, but the accessibility to the person, there's no detachment. Do you think it's harder or easier to practice law now?
Speaker 1:That's a really good question. I think it probably goes both ways. I think it's so. If you look at the business of law and let's say you're using cutting-edge AI tools to automate a lot of your practice, what does that mean on a business level? Well, it means that you don't have to bill your clients as much. Let's say, you're not in personal injury where it's contingency fee or flat fee, but you're billing by the hour. Well, now you're spending less hours on legal research and document review. And so for the business of law, that means you're probably billing your clients less, so lower margin practice, but that enables you to be a higher volume practice.
Speaker 1:So, in the one sense, there's some stress I guess that comes with running a high volume legal practice but you're replacing one stress with the other, because now you don't have to worry about the stress of performing frustrating, repetitive tasks where you're doing lower volume but you're spending way too much time on BS that you shouldn't be doing. So I think you're just replacing one stress for another. Ok, do I want to become a higher volume practice? Or you can find that line where you're able to take on exactly the number of cases you want to take on and be home with your kids and wife in the evening and have work life balance, and you're able to choose what that level of work life balance is and what that level of volume of files you're taking on is. And you have that option now because of tools like AI.
Speaker 2:Yeah, I think it's more, even more clear cut on the plaintiff side. How many cases do I want to take? Do I want to vacation three months out the year and take on 25 cases or 50 cases? Or do I want to make millions and millions and millions of dollars and still operate eight to five or eight to seven and use those capabilities and take on hundreds of cases, right?
Speaker 1:Yeah, exactly, and with tools like AI, you now have the option of lost my train of thought. Give me a second. Let me redo that.
Speaker 2:What did you just say? On the plaintiff side, it's more clear cut. You have more versatility on whether you want to take on a lot of cases or not. A lot of cases, right?
Speaker 1:And if you use cutting edge tools like AI, you now have more predictability around what that file is going to look like. So instead of being hit with an emergency motion and saying, oh crap, I got to go into the office now on the weekend and do legal research, Now you have predictability around. Ok, I know, even if unforeseen things happen in the course of this file, I'm able to handle them with finesse.
Speaker 2:Just a few years ago I would get a box of files from another lawyer and it was a folder with paper indexed, written, and it was boxes of files. Now I'm able to ask AI what a shoulder surgery with arthroscopic assistance for a rotator cuff tear is worth and get an answer in two minutes. How did we get here so quickly?
Speaker 1:Good question. It happens quick. Ai develops quickly. Well, the funny thing that a lot of people don't know is that AI has actually been around for a while. It's been around since the 1960s, the early stages of AI.
Speaker 1:But the advancements that had to happen at a hardware level, at a microchip level, at a computing level, to be able to process so much information only happened more recently. And that's how AI works right. It has ingested so much more data and information than any one human possibly could. You could try a thousand cases in your lifetime, but if you're an AI that has analyzed the transcripts from millions of trials, tens of millions of cases, then it's just going to know the law better than you do Now. Is it going to be a better lawyer than you are? No, but yeah, a lot of this legwork that went into building AI happened very, very gradually and then it happened all at once. It felt like it was happening all at once because the advancements in the hardware and the compute power and companies like GPT and OpenAI just enable these really rapid advancements in the technology in recent years.
Speaker 2:You know, I just canceled my shredding contract, so we haven't generated paper ever. I was paperless since the time I was formed. I don't even get paper from other lawyers anymore. Wow, Just to show like the Michael Scott paper company was a terrible idea. He really he had no shot. Yeah, no shot. I got a question for you Would you invest in NVIDIA? Question for you Would you invest?
Speaker 1:in NVIDIA. I don't know anything about investing man At this point because of DeepSeek, it seems like they took a big hit, so I don't know. I feel like it's probably past its prime, but I don't gamble in the markets.
Speaker 2:I don't play around with that stuff. It was a loaded question, because it seems like these chips are getting more and more intelligent and efficient in their use and it's requiring less time, less capacity, less energy. It's just wild. It's growing at an almost uncontrollable rate.
Speaker 1:Yeah, it is, and who knows where we'll be in 10, 20 years. I think it's not too difficult to predict the next year. To predict the next two years five years out is even a stretch, right. Five years ago, we didn't have chat, GPT, People didn't really know what AI was. When I joined Alexi in 2020, trying to sell this technology, I was getting hung up on all the time. People didn't really know what AI was. They didn't care. And then, when it became popularized through a tool like chat, chippy T now everybody's jumping on the bandwagon and they want to adopt AI. Um, so who knows where we'll be in five, 10 years? But it's going to be an interesting and wild ride.
Speaker 2:Do you like any football teams? American football teams You're going to hate me, but I don't follow football.
Speaker 1:What's your sport of choice, dude? I don't really follow sports. I, if anything, basketball. Because my kids are really into basketball. So they got me into it Toronto Raptors a little bit. But now we're a big Cavs family living in Cleveland and we go to a couple couple Cavs games a season. We go an hour and a half early. My kids try and get all the signatures from Donovan Mitchell and all these guys. Yeah, they love the Cavs.
Speaker 2:I was really upset. They're on the West Coast tonight. Oh are they? I was going to try to catch a Cavs game. I was like we need to do a better job planning next time game. I was like we need to do a better job planning next time. Yes, we will do that, man. Thank you so much for being on the podcast. You're awesome and um, I'm I call you a friend now.
Speaker 1:You're an awesome dude, send me that spreadsheet, bro.
Speaker 2:I will. I feel the same way, chaz. Thank you so much, absolutely Thank you, bro.
