Embracing Innovation and Preparing for Change

In this episode, we dive into the evolving intersection of artificial intelligence (AI) and the legal profession. Our guest explores the benefits and challenges of AI in law firms, highlighting how AI can transform daily operations, from research and document review to billing and training new attorneys.

Transcript

REMINGER REPORT PODCAST ON EMERGING TECHNOLOGIES

Part II

ZBP      Zachary B. Pyers, Esq.

KG       Ken Gavranovic

 | ZBP | Now, I will tell you that we’ve talked, you know, not just me but the industry in general has talked a lot about how this is going to affect or change the way we practice law and, and, and I know a lot of organizations are going through this, not just legal organizations but a lot of other
| KG | I agree with you, yeah.
| ZBP | I mean, from doctors, accountants. When we, especially with professionals and I think of them, and I don’t mean to, but professionals sometimes tend to be very protective and especially do not like change because they don’t want anything that may affect or hurt the way that they earn money.
| KG | Probably.
| ZBP | I mean, there’s lots of, there’s lots of professions that are like that.
| KG | You pick, you pick the industry, yeah, absolutely.
| ZBP | Right, absolutely. And so as I think about this, you know, I, there are, I will tell you candidly, I’ve got some concerns for the industry, not, not so much that I’m afraid that I’m going to be out of work and that
| KG | Mhmm
| ZBP | AI’s going to replace all lawyers because I don’t, I just don’t see that happening.
| KG | It’s not gonna happen. It’s not capable of original thought.
| ZBP | Right, and, and so I do, I mean I do see, right, that as a lawyer I’m still probably going to be employed in the next 10 years, but I do think that there’s probably gonna be some changes to how things are done and how things, and so, as we kind of wrestle with some of these changes and/or downside, kind of, how does the legal industry take full advantage of this by and still kind of minimize the negative effects that may come about?
| KG | Mhmm. Make, makes sense. So I, I have some thoughts and I made a, made a couple notes I’ll go through. I think, you know
| ZBP | Yeah
| KG | First is this change is going to happen, right. So let’s, let’s start a culture of innovation vs. just hoping that it doesn’t change, right. I think that’s the first step is that to accept that change is coming and let’s have a culture of innovation which in particular, and again, I know people at, really, you know, 3,000 firm, you know, law firms, you know, innovation in law firms as far as how they operate usually aren’t two words that go together that frequently.
| ZBP | That, that is correct.
| KG | --laughs-- And so I think that’s the first part, is, I think you need to say, like, this is coming, we need to first start having a culture of innovation so we start looking at how can we improve our business, and it could be even to start off with your back office. Like, how can we just deliver high quality service to our clients faster, which in es, in essence should increase your gross margins in the near term, right, but it also gives you a competitive mode if suddenly you have pressure on your margins later. So I think culture of innovation is first. The next part is a lot of times people, when they get excited about change, they jump into it, you know, the, the deep end first. I always say don’t do that. Pick something small but strategic. So, like, I’ll give you a simple example. Maybe you’ve got a, maybe you’re a 2,000 firm, law firm. This happens all the time. I, I’ve got, I need specialists for this particular case that are licensed in this particular state. You’ve probably got it on, you know somewhere on your intranet usually or sometimes I’ve, I’ve literally heard of law firms where there’s, like, two people in the firm that just seem to know everything that everybody goes to. Like, who’s good at that, right. So that might be a very simple thing, is like, let’s just digitize our skill sets so that we can quickly deploy the right skill sets to the problem, you know, as an example of something very simple but strategic to start with, right. So you get the right people so you deliver the best quality service as quickly as possible. You know, upskill, upskill your teams. Like for example, if, if, if a lawyer hasn’t got used to the term of writing prompts, like some lawyers haven’t even used ChatGPT or Claude or anything like that, if they just start using it themselves, they will save a ton of their own time, right. So but how do you teach people, like, here’s how it works and here’s how you can do it and here’s how it can help you and here’s how you can load in, like, four different legal documents and ask it to cross-reference and check these things that you might have had to do manually before
| ZBP | Sure.
| KG | If that makes sense. So again, culture of innovation, small but strategic start, upskill your team. Sometimes there’s legal tech providers that might help you. I think ethical and risk frameworks can be important because the other part is putting all this data, that’s your secret sauce, so security and compliance. You already have that to a large extent, but you need to make sure that you really emphasize that and who has access to this because now it's really easy for someone, you know, whoever has access to this, it’s a very powerful tool so you want to be, make sure it’s very, very secure. And then the last part is, I always say, is if you think about, like, if you have a firm and you map I just got a client to all the things that I do with the client and that full life cycle. You know, if you think about it, the different types of clients. Map the life cycle and think about the internal processes of how it works today and then with that kind of mapping, because a lot of times people haven’t even thought about, like, all these parts that are very manual and that goes here and that goes there. Map it and then look at what can we use AI and this new approach to automate it or drive more efficiency in those processes.
| ZBP | So I think that a lot, I think that makes sense, and so I, I will tell you that at least, you know, as I sit here today, I see a lot of comp, well, I, I’ve, I sit on a handful of kind of industry groups where I, where I meet with other lawyers in Ohio and across the country, and I’m surprised that a lot of what I would consider to be the mid to large, not the huge, huge firms but the, kind of the 50 to 250 lawyer firms, a lot of them still are not necessarily using AI in a lot of different areas. I, I mean, they’re hardly using it at all.
| KG | Right.
| ZBP | And there are, there are some legal service tech providers, I think the big, and I won’t name names but I will say there’s, there’s two, at least two huge ones that have been around forever, and, and they give free access to kids and adults in law school to kind of lure them into using their products post-graduation. They’ve started to use it for a little bit of drafting work, a little, a lot of research, right, where they’ve started to kind of improve their research models, and it occurred to me as you said that, the prompt writing because when it comes to some of this stuff, especially the research, the prompt writing becomes so crucial.
| KG | Yep.
| ZBP | And, and if you use it like you would use a traditional legal research platform, you’re not gonna get the same results.
| KG | Right.
| ZBP | And you’re not gonna get the best results, and so there is a lot of that in there. What, what I, so it seems to me that a lot of firms as they’re dipping their toes in are dipping them in, in these commercially available research kind of platforms because it’s the easy one.
| KG | Mhmm.
| ZBP | But as, you know, you and I have talked. I mean, I, I can’t help but think about my back office as you’ve mentioned because I’m, like, oh, is our back office using this. I don’t even know because, I mean, I have very limited interaction with our back office. I kind of separate it. I have people, and all I know is that they send my bills out on time and tell me when a client hasn’t promptly paid them.
| KG | Right.
| ZBP | But I start to think about this and I’m like, oh yeah, like, this could be a real area of strategic use in, in that kind of framework, it could make them significantly more efficient in a lot of things. I even think of, like, I manually review all my bills and I’m starting to think is there a better way to do this now. Is there a way that AI could flag bills for me that are problematic or entries that need to be adjusted or changed because client guidelines won’t pay for it or
| KG | Mhmm.
| ZBP | There’s an error in a misspelling or there is a, you know, another, let’s say there’s just an incomplete statement. It says “Correspondence from” and then it’s blank.
| KG | Right.
| ZBP | Would it, could it help me, rather than me spending an hour or two every month flipping page by page through a bill to make sure that it reads correctly, could it help me with this process?
| KG | Yeah, I think that’s a really great question and one of the things that, like, okay, I’ll go back to when I do some work with private equity companies and we try to AI-ify their business. One of the, there’s certain areas, like for example, we know if we get developers to use AI to build software, for every three developers you get a full extra developer of velocitor capacity, like, the data’s pretty clear. And there’s other areas of QA and tech, but you know, going back to contract, there’s lots of other areas that a lot of times a company haven’t thought about it, so what we usually like to say is, a great way to start off might be, is to get somebody’s who’s tech savvy in your organization who’s really passionate about it and can kind of talk about the capabilities and, you know, demonstrate it to the team, and then do something like a contest internally. Hey, everybody’s in their thing and we want, we’re gonna have a contest - who’s got the best idea of something that we could quickly implement that would deliver value. And that way you get the people in the various functions not being afraid of AI because a lot of times if it’s, comes from on high, like oh what does that mean, but if it’s, if you feel like you’re a part of, like, hey this is not something against me. This is something that can help me be more efficient and help our company be more efficient and it’s pretty cool. The company’s asking me for the ideas. And then you’ve got your billing clerks. You’ve got, you know, maybe your executive assistants. You’ve got all of them thinking about how they could use it in a, you know, again, small step forward, and we’ve seen some really, really great things where great ideas that “leadership” never even thought of. Once you showed the team members the capabilities, they start to come up with really great ideas that you can implement relatively quickly.
| ZBP | Now I will tell you that there is a number of at least, I don’t want to say struggles, but people are predicting a number of challenges I will say with AI and the use in the legal industry as well as a whole host of other industries and professions. One of the things that I have heard or I have seen talked about is training young lawyers, and I will tell you that, and I, I see this kind of, I don’t want to say that I’m in the middle but I kind of see this from both angles in that I see younger lawyers who may not get the traditional training that I had because the tools are just different now and some of the stuff that I did, like for example, I, when I first started out, I had a partner that I worked with and I did a deposition summary of every single deposition he took so I was handed, after the, the deposition transcript would come in, he would hand it to me and I would review it and I would create a, a page by page summary of everything that the person testified to. Now I already know that that task is being replaced by AI because I’ve used it. I’ve, I’ve uploaded deposition transcripts and said please summarize this on a page by page, and in very short time, you know, you’re looking at a 200 page deposition which would have probably taken me three hours-ish, maybe four to create the summary for, it’s shooting it out in three to four minutes. And so, but, so that task may be replaced, which is helpful and efficient, but I reviewed so many deposition transcripts within the first year that I was clerking and then year that I was an associate. I probably reviewed a hundred transcripts and read them, not just of this lawyer but of all the depositions he was in, so there were all these other lawyers and I got a chance to review the questioning styles and tactics that these lawyers used. So I actually had the benefit of essentially sitting through a hundred depositions without actually having done it.
| KG | Mhmm
| ZBP | Now as we fast forward through AI, I think that these young lawyers, and like, they’re not gonna see those deposition transcripts because no one’s gonna pay them to review them and summarize them because you can use a machine to do it. So, and I know that’s a very kind of microscopic aspect of it, but I, one of the things that I think I worry about is how do we train the younger generation to still practice law since they, the normal training techniques we were using 10 or 15 years ago aren’t going to be there and how do we make sure that when I retire someday in the next 20 to 30 years, that there are some, there are some lawyers a generation behind me that are ready to kind of step up and take over.
| KG | Yeah, well, I think, first of all, I think that’s one of the dangers to society with AI in general, is if we don’t have to think and we could just get the answer immediately
| ZBP | Yeah
| KG | Then do our critical thinking skills decrease over time as a culture, and I think that’s a very valid question. Like sometimes, you know, like again, I’m a tech guy and I get excited about it but I think that’s very valid because the brain having to do that critical thinking is important. Now I think the good news is, I think we could probably develop ways for people to, to still get that knowledge because I’m just going back to that scenario and just brainstorming. If one of the things we determine, hey, like, let’s just say we modeled out how we train and on-board new attorneys is we historically, we had them do, review depositions and write summaries, we had them do this, we had them do that and we had these other things. Great. So then what I would say is, well how can we use AI to give them the same learnings. So it could be as, as simple as, like, you know, showing them here’s the, here’s the key patterns of effective questioning in depositions because then it could take through, through all of that, give it to them and you could even have AI create scenarios where it could be fully interactive and then score them based upon how they’re doing it.
| ZBP | You know, I, so I’m not gonna lie to you. I, if you can’t tell by my face, I’m getting a little excited as we’re talking about this because I started thinking, as good as the training that I had was, right, my reading all of these deposition transcripts, based upon what you just said, wouldn’t it be even better if the, if I would have been able to interactively train with the AI by doing mock depositions where it continued to change up on me where, and I would have to adjust or change my questioning style on the fly with that model in a practice setting and it saves you all of the quote/unquote “embarrassment” of having to figure it out in the real world because I, and I’ve, I teach a class in depositions at my alma mater, Capital University Law School, and usually what the class focuses on is, is roughly 50% of me telling them all of the stupid things that I did early in my career.
| KG | Right.
| ZBP | In hope, in hopes that they avoid those mistakes. But I started thinking about it as you were talking and I thought, that’s a really good idea. Why are we not using AI to do these mock depositions to help these young new lawyers get trained because I, I certainly think there’s a very real possibility for it to do that.
| KG | Right, and what’s great is, like, you could go back to, you could take is, you know who’s, who’s your best at, let’s just say insurance depositions, right, ‘cause that’s where you’re looking for little inconsistencies and so forth, and so you could literally then have the AI load that and you could literally tell the AI, you’re going to play the, the person receiving the deposition and I’m going to be the attorney and what’s, what’s crazy, if you notice how good the voice is getting on, like, Claude, ChatGPT. It’s almost, it’s almost conversational and it’s only gonna get better. You can literally then just have them say, okay great, here’s the scenario of this case, and it can mimic even a previous case and they could just go through the deposition and then they can score them based upon your best attorney’s depositions of what they missed, what they forgot to, you know. You can insert, you know, the false narratives. You could even have it run the exact same scenario that actually happened historically but allow them to live interface and, and run with it.
| ZBP | Yeah, that, I mean, this, the, the prospects for using this, I think, now, I think that’s a really helpful analogy, right. Is, is when we start talking about the problems of AI what you, I heard you say at least, is you flipped it and say well, AI has created this problem, how does AI also help us solve this problem.
| KG | Right.
| ZBP | And I, I don’t have good responses to why we’re not trying to utilize it that way. I mean, it would make a whole ton of sense to me because one of the problems that I think we’ve seen in the, in the legal industry in general over the last 5 or 10 years, 20 years, is that there’s a lot of things in the legal industry that have changed. Jury trials have gone away where people 20, 30 years ago tried a lot of cases. Now it’s like less than 1%, so one of the things people have been talking about for the last decade is how do we get younger lawyers trial experience. And there really hasn’t been a great answer. They, people have kind of come up with a hodgepodge of ways. Some, some law firms run a trial academy or a trial institute where they do a mock depo but, or I’m sorry, a mock trial, but it’s not real, so all of the things or the pressures and the anxiety of being in front of a jury and a judge who may or may not like you, may or may not like your client, some of those are removed. But it occurs to me that we really could start utilizing this really to start training in a way we just haven’t even thought of yet.
| KG | Yeah, I agree, and I think that to your point is, it’s never gonna be the same because, like, it’s not gonna mimic the pressure of being there and the political dynamics and the socioeconomic and all of those kind of things, but it can get them trained up on the practice of what good looks like. And then of course
| ZBP | Yeah.
| KG | They’re gonna have to learn to, you know, and that goes back to kind of internal controls, right, like how do we show up. That’s something that, you know, the, I think it’s a different type of experience, but I think the mechanical part of what’s a good questioning and how should you do it and those type of things, I think AI absolutely can power that, and again, going back to how would I go about it is, again, all of your processes including onboarding a new, a new young attorney, what, what do you, what do you have them do and why do you have them do that. And then is there a way that we can get that knowledge even faster and even of a higher quality into them so that they become up to speed, and I think in time you’ll probably see people get smarter. Like, I do think there’s a part, is the amount of knowledge that’s instantly accessible to everybody, you already, and I think we’re all in some ways getting smarter because now, we, if we have a question, we get to know it and we get to learn what the answer is in a definitive way. We get to see what’s good, what’s bad. We get to do reasoning. I think that that’s, you know, I think that if you can kind of think about that and reverse engineer each of your processes at the law firm, I think you can give most of that training to them.
| ZBP | You know, one of the things that I think, and, and I’ll dispute in general from the law firms that I’m aware of. I think one thing that I think the law firms are not great about is identifying necessarily the processes, and so sometimes the, there are processes, right, and they’re just, because these organizations have been around for so long, things just happen and nobody knows exactly how or why they happen but they just kind of happen and everybody knows that they happen, so, but nobody’s actually stopped to think, is this the most efficient and best way and are these processes that could be adjusted. I, I mean I think that, and I’m kind of zooming back out now, is one of the things that I think is gonna be challenging I think for a lot of mid to large but not giant law firms are gonna have to start really more closely examining the processes to see how they can best utilize and adapt.
| KG | I, I agree because you don’t wanna hire a consultant. Obviously that can, can do it, but again, we’ll go back to a best practice. What I would do is have a working group, like going back to, hey we’re gonna, we’re gonna, we’re gonna start to embrace a culture of innovation, to, you know, company wide. That’s what we’re gonna do. And we’re going to start looking at AI, is do, who’s passionate about it so we’ve got somebody that’s already passionate and can educate our force on the capabilities of it to the best of their knowledge. That’s a good start. And then, you know, again, like I said, that the competition or competition in working groups because the, the competition may give you the ideas of where you want to focus and then you could take the working group to then go execute on that change and again, don’t try to bite off the big thing first. What can we do that’s small, incremental and immediately delivers value which in the near term should increase your gross margins and in the long term protect your gross margins and in the long term protect your gross margins.
| ZBP | You know, so I, I will confess to you that I actually have to do a presentation in about two months on AI and how it’s gonna impact the legal industry, and I will tell you that I think I’ve now gotten about, it’s only a 20-minute presentation, but I think I’ve got about 19 minutes of my material just from my conversation with you here today. As we, and so I will thank you for that, but as we look towards kind of the future, is there, I mean we’ve talked a lot about how all of these benefits are gonna play out and how it’s really gonna help us improve. Do you see any downsides? Are there any risks that we should be kind of worried about, and I don’t just necessarily even mean the legal profession and the implementation of AI, but just kind of, you know, organizations in general. Is there anything that they should maybe say, hey, let's pause just a second before we move forward with this or is there anything along those lines?
| KG | Well, I think there’s the, there’s, we’re going through the practical and ethical, so on the practical side is AI doesn’t solve all problems. Right now, as I mentioned before, it’s not capable of original thought. There is people that believe they, they call it AGI, but that it will start to be able to do original thought in somewhere in the 3-5 year range, and you know, what does that mean? You and I both don’t know, but I think about, you know, you think about humans are a resilient species and we react to change, and this is gonna be one of the biggest changes that’s happened across all aspects of how we operate in business and life, and so if you believe that’s gonna happen, then you have to lean into it so you don’t get steamrolled by it. On the ethical side, you know, the part that, you know, like, I’ve got people I know that are in customer service. There’s certain job roles where their whole job is to take a set of data, listen to a customer and then follow a set of rules. Like, that’s gonna be tough because AI’s gonna be very good at automating those type of roles out of existence, and so, you know, what does that mean. You know, you don’t know, I don’t know, but it is going to happen and I’m just a believer that, you know, when you think about all of the innovations that have happened. You think the same thing happened when the computer came out, the same thing happened when the internet started. It’s going to change and you know, but if you don’t change, like if you never got a website, you’re probably out of business, right. If you, and I think this is gonna be the same thing is if you don’t change, it’s gonna be there. I, I wonder what it’s gonna do to some historically high gross margin businesses. I think, you know, the, the, I think there’s a real risk that it, it changes some of those businesses because it, really it’s an amazing driver of efficiency.
| ZBP | Yeah. You know, I, I remember reading years and years ago and I cannot tell you which economist it was, but this was an economist probably a hundred years ago that predicted that by the end of the 20th century, we would be working a total of 4-8 hours a week and that would be the average work week in the United States because, you know, we would see such technological innovations, and I do think, right, you know, by the time they made that prediction, you know, it was essentially a hundred year span. We saw wild technological advances occur during the 20th century. Those, I mean wild, but I’m not working 4-8 hours a week or 4-8 hours a week. I, you know, I’m still working a full work week and, and some would argue that the work weeks may have even gotten longer in some professions.
| KG | Yeah, and our output and our efficiency, you know, like, I, I would say, like, just what I use, Claude and ChatGPT, I would say I’ve got in, in throughput probably 20 hours a week in, in additional things I can accomplish.
| ZBP | Right
| KG | So I think that, you know, I think there’s probably a question over the longer, longer term, maybe 10-20 years and what does that do, but I think in the near term, this is an incredible tool for us to be so much more efficient in what we do and if you think about most of the stuff that is automating, it’s not the high value work that we do, it’s the low value work. That’s what, what AI today is automating for us.
| ZBP | Right. Well, and that’s, you know, and to your point when you were talking about hiring developers and the utilization of AI and essentially if you have three developers, you utilize AI, it adds like a fourth one, and I was thinking about that because that’s one of the areas, and I don’t mean to pick out software developers, but that’s one of the areas
| KG | Yeah
| ZBP | That I’ve heard is that, that AI can be used really effectively to write kind of, for lack of a better term, kind of the intro coding. Not, it’s not great at doing more sophisticated levels of coding, but if you, if you already have some code, it can repeat it and help you build it out where if a developer would have had to do it manually in, in years past, it can kind of already copy and do it. It’s the fine tweaking that really needs the developer’s experience or expertise on kind of the, the, the higher end stuff if that makes sense.
| KG | It, it does but I will say is if you know the questions to ask like as the developer it is, because I’ll, I’ll tell you this is the true life. Venture capitalists already are seeing this, is it used to be, hey we invested $100 million, they built this software, it’s got a proprietary mote that’s very hard to duplicate, but what you’re seeing is smaller companies with dramatically less capital be able to generate tremendous amounts of revenue. Like, there’s a company that I might do some stuff with and join, but in three years with 50 people, it’s gone from 3, it’s an AI company, from 3 to 20 to $59 million in revenue, and, and I’ll just give you one other example. There’s another company, that there’s a thing called No-Code, I don’t know if you’ve heard of it, but like, this is a lot of law firms, insurance companies have bought these systems that were kind of drag and drop to build applications.
| ZBP | Okay
| KG | And it was easier because I didn’t have to hire really sophisticated developers. But they’re already starting to see that they’re getting crosswinds because now those developers can actually build very sophisticated applications because they can just ask AI how to do the thing they didn’t know how to do, and it gives it working code to do that exact thing. So, and, and that’s the true life example of a company that was probably $60, $70 million in revenue that over the last couple years has gone down to $30 million in rev.
| ZBP | Yeah.
| KG | Because now it’s, AI’s made that problem they were solving that’s hard to code these type of solutions, you know, essentially go, well go away slowly.
| ZBP | Right, right. Well, Ken, I really appreciate you taking the time to talk with us today. I can at least tell you that I learned a lot and it kind of opened my eyes and kind of the way that I think and kind of adjusted some of my thinking on some of these topics so I greatly appreciate it. I appreciate you taking the time to speak with me and our listeners. We greatly appreciate it.
| KG | No, Zach, I really appreciate it and if there’s anything, I’m a big believer in just kind of, you know, do good things and the world, the world does good things to you, so if there’s anything I can do for you or your listeners or they’re like, hey, you know, we’re trying to put together one of these working groups, do you have any ideas on how we should get started, feel, they could feel free to reach, reach out. You can find me at kengavranovic.com or, you know, just search for my name on LinkedIn. It’s kind of unique so you’ll probably me pretty quickly.
| ZBP | I was gonna say, I Googled you and it was not hard to find you very quickly.
| KG | Yeah, you can Google me, too.
| ZBP | Ken, I really appreciate it. Thanks again.
| KG | Alright, have a great day. Take care.