Maximum Lawyer

Tyson Mutrux (00:00)
Good morning and welcome back to Maximum Lawyer Live. I'm Tyson Mutrux. Reporting live from Rolla, Missouri. So hopefully everyone's having a great Wednesday. So last night I was at a soccer game, my son's soccer game down in, you know, middle of nowhere, Missouri. It was like all the games are seems like an hour and a half away, two hours away, three hours away. But I had a really interesting conversation with a retired attorney, and that is what I'm gonna talk about today. It was, he said something to me that was really interesting. He goes, because he retired in a wait, okay, so, before e-filing ever started occurring in Missouri, like we had federal e-filing, which is has never has actually not changed since then. For the most part, Missouri's got it to me as a very advanced e-filing system. I've got it to compare Missouri versus Illinois versus federal. Missouri's is very good. It's way better than the Federals, the Pacer and everything. It's way better than the Illinois system. That's not what this is about.
Tyson Mutrux (01:12)
But he said something to me. He goes, I don't understand why everyone's getting mad about people writing briefs using AI. And I was like, it's an interesting comment. It's interesting because he's been in the world, but now he's outside looking in. So he's got the inside about it, but he's not in the world anymore. He's not licensed to practice law anymore. I mean, he's retired, retired. For him to make that comment was really interesting. And then it got me thinking about... There is a recent bill on Gavin Newsom's desk in California. The bill is SB 574. It passed on August 31, and I believe the vote was unanimous. It would ban delegating the practice of law to generative AI, require you to keep client data out of open models, personally verify citations including AI generated ones, and disclose AI use on court filings. So, I mean, I'll kind of go through each of these individually. So, do I think you should delegate the practice of law to generative AI? No. But, it depends on what you mean by delegate, right?
Tyson Mutrux (02:26)
Like, you're not going to you're not going to delegate your judgment. No, I agree with that. Like, but if you're saying that, like you can't delegate any parts of it, that's insane to me. Like drafting a, a brief, motion, or pleading a basic, super basic pleading. Technically that's delegating the practice of law, but you already do that with paralegals and legal assistants. Anyways, that seems kind of absurd. you could even say arguably if you have forms pre-built forms you're delegating the practice of law to forms that seems absurd completely absurd when it comes to if you get to require to keep client data out of open models okay depends on what you mean by open models, because many of the quote unquote open models have it where if you have a paid plan, the data stays inside of your app. It doesn't go outside into other apps. It doesn't go to the company. It's closed off.
Tyson Mutrux (03:31)
So that seems like if the bill is exactly what I'm reading, and this is from a Bloomberg article, then that's a misunderstanding as to what the models actually do and how they store your data. Okay. Let's set that aside though. Personally verify citations. We personally, personally verify citations right now. Like we, we have like, we do pay people to draft things for us. And I, I've got an healthy amount of skepticism as to whether or not they are personally drafting those. I don't, I fully understand that they might be using AI to draft these things. As long as they do a good job, I don't care. As long as they are good, I do not care. But I have a massive amount of distrust for AI. And I wouldn't say massive. I'd say healthy. I've got a healthy amount of distrust for AI when it comes to citations.
Tyson Mutrux (04:32)
So we personally verify every single citation, every single one of them, to the point where we even had one thing I mentioned on the show where there was we got one of the motions back from this person that we pay to do it. And we couldn't find the citation. We Googled it. We couldn't find it anything. And so we were we were actually about to fire the person. We were really angry about it. And they we let them know, like, we can't find this site. And they knew how much we cared about this.
Tyson Mutrux (05:11)
So like they they check too like they but they understand like hey we are like this is something we take very seriously so he went and he found it he sent us the case and he it was he found some obscure case it was actually a favorable case for us but we couldn't verify it anywhere it was kind of a weird thing he found it existed it was a real case he sent us the pdf of the case and so it was it was all fine everything was fine so i i have like we we take it very very seriously Like, that's how seriously we take it. But I do foresee a time where, like, I'm guessing, like, AI right now is going to... It's so good now that you could actually catch if it's a bad citation. I don't see the hallucinations that I saw, you know, a year ago, six months ago. You don't see that like you did before.
Tyson Mutrux (06:18)
So, I mean, I did... i posted this in the Association recently where it it was unrelated this was not even i don't think i was doing anything legal related but it flat out told me it made something up and i was it's like what it just flat out because i i asked for it i i knew it was whatever it told me was completely wrong i can't remember what it was but it was i knew i remember it being completely wrong and i and i just called it out right away and they said Let me be upfront with you. You're right. I made that up. And I had never seen that, that it just flat out said it made it up. That was kind of interesting. So you just have to be careful about it. But I do foresee a time where the personally verifying sites is not going to be necessary anymore. As of right now, I'm still doing it, but I'm guessing it's not going to be necessary. And then disclose AI use on court filings.
Tyson Mutrux (07:31)
I think that that is such a ridiculous, unnecessary thing. Like, what... Okay, first of all... First of all... When it's... If I... Like, you're not disclosing... if a paralegal drafted something for you or like you, what does it do for you? I don't understand what the point of it is. What does it do for you if I tell you that it was drafted with AI or if you used it on any part? Like these rules exist across the country already, by the way. So I'm already aware that it's not just California, it's other states. They already have these rules in place that you have to disclose it. What does it do for you if I tell you it was drafted with AI? Does... Does the court have to tell us that they drafted an order with AI? No. No, they don't. What benefit does it do? If anything, the only thing it really does is it creates some sort of doubt about a document that may be one hundred percent accurate. That to me is insane.
Tyson Mutrux (08:44)
So I guess the whole point of this, just the show is, I don't understand why our profession is pushing back. It's weird. I feel like a lot of the people in Maximum Lawyer, the Association, they're embracing it where on the other one, we have this profession that's regulating it. That they have a misunderstanding as to what the AI does. And so they're creating these rules because they don't understand it. And we have to fight back against this. We really have to. When you see something come across that is a proposed rule, we have to go and we have to argue against these things. when they don't make sense. Like I if we go back to this, I don't I don't disagree. I think that, yeah, there's maybe there should be some guardrails, but generally banning the delegation of the practice of law to generative AI. That's not specific enough. We have to be more we have to be more specific to that. Keep requiring you to keep client data out of open models.
Tyson Mutrux (09:55)
I mean, That rule needs to be refined quite a bit. I'm personally okay with that one. I'm personally okay with that one. But the disclosure of AI use on court filings and the fact that this was unanimous in the vote is kind of ridiculous to me. That many people cannot have a misunderstanding as to what AI does. I find that really, really, really hard to believe. But I think maybe that's the world that we currently live in. So I guess what we really have to, as law firm owners, and we are leaders in our firms, we are leaders in this industry. So we have to be able to distinguish for regulators, like the legitimate caution from AI from like what's actually ridiculous like because it's becoming to a certain point it seems for some people an excuse not to learn not to learn it not to understand it And that's kind of sad because there's a massive amount of usefulness.
Tyson Mutrux (11:01)
What's interesting is this attorney, to kind of circle back a little bit, this attorney that I was talking to, he was not like , I'm not going to drop his name. He wasn't some guy that , just you know had no involvement in regulation like he actually was involved he worked for the state of Missouri for a while he worked for an insurance company he was in really integral in getting unfortunately the oh seven tort reform passed in Missouri so he had a say in things okay this is a guy who understands regulation but he's also understand he sees the usefulness of AI he can see it And he is against the regulation of it. So it's a really interesting thing how I think sometimes the regulators, they get so sucked into creating rules that they don't stop to think about, do these rules make a lot of sense? So that's part of what I want to discuss here. It's one of those things like...
Tyson Mutrux (12:09)
We as leaders, we've really got to step up in the legal community and advocate for what rules need to be advocated for, but also advocate against ones that shouldn't. Because there should be responsible adoption of AI. We have to determine what does that look like in a law firm. Right. Because just letting our agents run rampant is not the best thing. We don't we need to put guardrails from a business standpoint. We need to put guardrails on these things. Does that necessarily mean that we need to have regulation on it? No, it doesn't. But. We should not be letting AI, these agents just go run rampant on all of our systems without any control and without any internal regulation. We absolutely need to make sure we include these guardrails. That is really, really important. Just as if you were going to put guardrails on your home. You're going to have guardrails on these agents as well. That's a really important concept too, because it's not just, it's not just good business.
Tyson Mutrux (13:19)
It's just one of those things where like from a process standpoint, right? Whenever it comes to your process, you've got to have controls over things. So things don't run haywire and you, you have to really think through your processes, think through your systems in advance before you let these things go. You really do. And so they need to have really finite tasks that they do. And they have to stay within those boundaries. They can't go outside those boundaries. If they go outside those boundaries, you need to tighten the boundaries a little bit. And there's also, I really want to hear from people, if you were listening to this and you have not really adopted AI, I'm really curious as to why.
Tyson Mutrux (14:10)
The amount of work that I can get done today is... times at least as great as what I could two years ago it is we are we are sprinting every day we are sprinting it's what we're doing becomes a litigation pre litigation when it comes to leads all that we are sprinting when we I feel like the the firms that are that are not using AI right now. Like they're not even walking, like you're crawling versus these other firms. So you need to really embrace it. I'm very curious as to those that are not, why not? I don't think it's gonna change. And there is a element that I guess I wanted to talk about with this too is I really enjoy talking to this guy. I really did. And it made me think will there be part of the practice of the law that it just becomes where like we become like work in the seventies. Okay. We're, I do feel like by necessity, people had to talk to each other more.
Tyson Mutrux (15:21)
They had to, they had to communicate on the phone and talk to each other and all that. There was an email that you could fire off a snark email or sending a letter was just take too long. You had to pick up the phone or you had to go visit them at the office or whatever. You had to talk to them in court. I do wonder if there's a part of the practice of law that's going to become more like that because by necessity, where the agents are doing all the work or most of the work and the stuff that really matters to get the deal done. Those of you that have ever done a deal of any kind, you know what I'm talking about, where you pick up the phone. Hey, let's cut the chase. Let's cut to the chase. Let's figure this thing out. that that being done in person or on the phone or, you know, at court, you know, where you're getting it done.
Tyson Mutrux (16:30)
It makes me wonder if like there's a part of the practice of the law that's going to shift back to that. I kind of hope it does. I like that part of the practice of law. I think most of us do like we enjoy that part of the engagement. We don't enjoy the drafting of boring letters and the really basic pleadings. I do kind of like drafting a really strong motion. I do enjoy that. I enjoy drafting a really strongly worded hammer letter. But there's the other letters that I just don't. And the AI can draft those for me in a matter of seconds. So I think there's... I do wonder, I don't know. I wonder what your thoughts on this, because I do wonder if there's a part of the practice of law that will shift back to where, to where we used to be when all this other stuff is done by AI, because it can be done by AI really easily.
Tyson Mutrux (17:38)
That's, that's one of those things I, it's one of the questions I kind of, kind of pondered as I was thinking about it yesterday. I was with this long drive home. So I had this, you know, this really good conversation with this attorney. And then, you know, I had this hour and a half home just to think about it. It was really kind of interesting to think about, but, I want to talk a little bit more about the, the, the citations though. I wonder like, does it, is it going to be necessary? I don't, I don't think it is. I really don't. I think it's going to be, it's going to be at this point where, It's almost like this operational debate in my head where my legal hat, my attorney hat, I really want to verify every single lead because I think it's necessary. But I did, and I didn't mention this before. I meant to. That's why I want to bring this back up. But I had taken a brief that I had actually drafted.
Tyson Mutrux (18:50)
And I sent it through AI. I said, I need you to check every single citation and find every single case and give me a link to every case. And what it did is it listed every single case that I had in a hyperlinked each one so I could verify that each one was an actual existing case. And then I could look to see if the quote was accurate, too, because I had to verify the quotes. And It had picked up a quote that was not fully on point. It was off a little bit. And so it told me where my quotes were off and what I was, what I was quoting wasn't quite what the case said in certain areas because I was, you know, I was being an advocate. So I was kind of pushing it as far as I could. And, but it did a really, really effective job where every single case was, every single citation was verified.
Tyson Mutrux (19:57)
I linked each one, told me where I was off and based a little bit. took seconds to do was very, very quick. So if it already can do it, So from an operations standpoint, I really want to be able to say, okay, we can run all these through. And I think what we'll probably do is we'll just require each brief, each motion, everything, every response, whatever, is run through AI one time to give us that one sheet, to look at the one sheet to see, okay, do these arguments make sense? You can have all these different list of things that you wanted to do. But do these arguments make sense? Do the citations or do they actually exist? Give me a link to the citation. Add those to the file if you haven't already added them. Because a lot of times what we'll do is we will cite to something. You know you do this. You'll cite to it. You never pull the case up. You never read the full case.
Tyson Mutrux (21:07)
So you can have it actually pull the cases for you, add to the file, and then whenever the judge asks to read it, you can actually just quickly find it for the judge and give to the judge. So you can, you can do those things and you can do it from an automatic standpoint. Like, so that's probably what we'll end up doing and then still require a human's eyes to go over it. Especially if there's a rule on it, that's something we'll, we'll have to do. So, we're going to continue to use it to our advantage because I think it's, it is a massive advantage. I really want to encourage people to push back on some of these ridiculous rules. Not if there's a good rule, support the good rule. Absolutely. But, if it, if it is not a good rule, we have to step up and fight these things.
Tyson Mutrux (22:09)
Otherwise it is going to one, depending on the state, if your state has these really restrictive rules, there are going to be other states that are just going to dominate and they will come into your state. These firms will come into your state and they will dominate, right? The Morgan and the Morgans of the world, will come in and dominate you because they're able to operate in multiple states. And so they can just easily come to your state. Their central office is located elsewhere. So they're not going to be following the same rules. You know, they're not, they're just not going to be. Cause the, the opportunity here is the better work. It's not just faster typing, not just getting things. It's just better work. There is, I used it to help me file a response to a motion for summary judgment. And it took me far less time and more, I have created a Claude skill where it has taken my work and created it in my voice. So I can run it through my voice.
Tyson Mutrux (23:22)
And it has all of our pleadings rules that we have in the firm that we give to every single attorney. I ran through that and it, It wrote a far better document than I could have. It just did. It did. It's way better. It used my voice in many parts, but I read through it. I was very impressed. I was a little shocked by how good it was. It's better. It's getting better and better, too. But it is much of what is out there is better than the work that we will do. OK, it just is. So we've got to embrace it. Don't don't fight back on it. Embrace it. And that's not what this episode is about. I've talked about that before. This episode is really about as the rules are coming through, we have to fight them, the bad ones. We have to support the good ones, fight the bad ones. Don't pick a fight. This is one of my big pet peeves in litigation. Don't pick a ridiculous fight just to fight, okay?
Tyson Mutrux (24:34)
So if we're wrong on something, this is litigation. If we're wrong on something, just accept like an object. Let's say they object to something because the question's bad. Okay, just let it go. Okay, let it go. We're wrong about it. Let it go. Set a different set of discovery, okay? If the rule is good, let it go. It's a good rule. Okay, let's just support it. Be supportive of it. If it's a bad rule, we got to go. We have to fight against it. Okay. That is all I have for us. I just want you to embrace learning AI. Embrace it. Don't have blind reliance on it, though. There are legitimate risks, and those deserve very specific controls in our firm. It shouldn't have some sort of automatic veto on every application. AI should just should. Make sure that you keep responsibility visible. Make sure that the information that you're putting into is approved by the firm. You have some sort of meaningful verification when it comes to human verification on things.
Tyson Mutrux (25:45)
Make sure you name the reviewers and how they belong inside the workflow, whether it's an agent or it's not. Make sure you have all that documented so you can easily track things down. measure better work, not just faster drafts and faster results, like count the full cost of the review and the correction of documents, then decide whether the process deserves to expand like something we have inside of Foxy, which is our case management system that we built. It actually has we track how much we spend on AI in the system. We can actually look at I can pull up a dashboard and it shows me how much of AI we're using on specific things. And I can click on yesterday, and it shows me all the things that were done with AI, which is really, really helpful. I understand that not everybody has access to that, but you can easily build it on Dashboard. You can do that. So it doesn't matter if you're using Clio, Filevine, or Smokeball. It doesn't matter. create a dashboard that tracks that stuff.
Tyson Mutrux (26:59)
If it's not one of your major KPIs, don't worry about it. It's more just information. So have that stuff first before you do anything else. But you got to measure, the whole point is measure better work, not faster results, measure better work. That's really what you want to, what you want to, All right, that's all I have for you today. Have a wonderful weekend. Make sure you check out BeckInLess.co. If you're interested in the conference, go to MaxLawCon.com. If you're in the Association, make sure you fill out the Association survey. Really appreciate you doing that. And we will see you all next week. Have a great week, everybody. Bye.