EP 914- Using AI in Your Practice

Episode Notes

Transcript

Announcer:

Welcome to Heels in the Courtroom, where the trial lawyers of the Simon Law Firm break down what it takes to win in the courtroom and in life.

Elizabeth Lenivy:

Hello and welcome back to another episode of Heals in the Courtroom. This is Liz Lenivy, and today I’m joined by Katie St. John and Elizabeth McNulty. Hello, ladies.

Elizabeth McNulty:

Hi.

Elizabeth Lenivy:

Hello. Today we decided to tackle the topic of AI, right? I feel like it’s something people talk about now on a daily basis. It seems to be the headline of a news article, at least every other day for me. And the reason I kind of thought about it is because, and I don’t know if you guys have experienced this lately, I feel like I’ve had more and more clients recently tell me in conversations, “Well, ChatGPT told me this or AI told me this about their case.” And so it kind of got me thinking about how we utilize AI and what this may mean for the future of our profession, but also just generally, and maybe this is a little bit too heavy of a topic, but humanity in general. It’s all crazy. I don’t know if you all saw this, but there was some articles posted about an MIT study that they did where basically, I mean, and the headline was, “Is AI making us dumber?” And the idea behind the article and the research that these folks at MIT did was that is utilizing AI stifling or hindering our creativity and our memory.

And so what I naturally did was I went to ChatGPT and I said, “Hi, ChatGPT. Can you tell me about that recent MIT article where it found that utilizing AI is making people less creative and dumber?” And so it spat out a response that explained a little bit about the study. And so it said that researchers had 54 adults aged 18 to 39 write SAT style essays using ChatGPT, Google, or no aid over multiple sessions. EEG caps monitored brain activity showing that the ChatGPT group had markedly lower neural engagement, especially in regions tied to creativity, memory, and executive control compared to the other groups. And that has now led to people thinking, “Well, when the computer’s thinking for us, we don’t have to use our brains the same way.” And I know there’s a lot of attorneys that are now utilizing AI in our line of work to obviously try to improve things, but I also wonder if it’s maybe trying to shortcut things.

And obviously I’m sure you all have seen articles about hallucinating cases and attorneys getting in trouble for that, but I thought it might be an interesting conversation to have about the ways that we’re utilizing AI to improve our representation for our clients and improve the jobs that we’re able to do. So Katie, I’m going to throw it to you. I’ll first ask, do you use AI at all in your work?

Katie St. John:

Yeah. So I actually hadn’t used AI that much at all until I came here and I started working with Johnny. And one of the first times he had me use it was when we were talking about different topics for a specific corporate representative. And I don’t know if this happens all the time or if it was just extremely helpful, but basically it’s an interns company. We’re taking their corporate representative and ChatGPT gave us the name of the specific document management platform that the insurance company used and pulled this article about that the insurance company had posted, outlining why they switched to this platform. And so it actually was very helpful in that instance. But prior to that, I hadn’t had a whole lot of experience with it, so I’m still getting my feet wet for sure.

Elizabeth Lenivy:

Okay. It sounds like for research.

Katie St. John:

Yeah.

Elizabeth Lenivy:

Have you used it at all for, for example, you talked about a deposition. Have you used it at all to help generate topic ideas when you are taking a deposition, whether it’s a corporate rep or maybe a doctor or maybe a topic that you may not necessarily have a ton of background in?

Katie St. John:

Yeah. So that’s the most I’ve used it. And I’ve also used it for in medical malpractice cases in the context of breaking down a certain report or medical language into layman terms to make sure I understood what I was talking about.

Elizabeth Lenivy:

Okay. Elizabeth, what about you? Have you used AI or ChatGPT in your practice?

Elizabeth McNulty:

Yeah, so I’ve used it some, not to a large extent like I know some other people at our firm have. I know it’ll generate demand letters for you and things like that. And I tried it once and it almost felt more cumbersome than just starting over on my own just because a lot of those, especially in auto cases, it’s pretty much the same kind of letter you’re going to send out. And it didn’t seem as easy as I thought it was going to be using AI. So I just ended up doing it myself mostly because I didn’t think that there was that much information to synthesize for me. But I use it on some maybe, I don’t want to say softer skills as part of our practice, but if I think I’m going to have a difficult conversation with maybe a client or opposing counsel, I’ll kind of talk it through with ChatGPT to see if there’s any areas where I can soften my approach or especially if it’s like letting down a client, like a case that we can’t take.

And I found that to be pretty helpful, but I do see the point where I am the human in this and it’s like that taking out that human elephant that I’m using AI to help me have a very human conversation with a potential client. So I do definitely see how it’s stifling our own creativity, but I also think it’s creating so many efficiencies in our work. So it’s just a trade-off and it is also about just like the user. If you let it stifle your own creativity, then that’s what’s going to happen. But if you’re aware of it, then you can potentially try to do something about it. But I think it’s great for low hanging fruit tasks that we do have or educating us on topics that would take a really long time to educate us on, or it would be costly because then you have to have your expert teach you instead of just using ChatGPT.

Elizabeth Lenivy:

Yeah. I did that recently in a corporate representative deposition where there was a ton of business entities and I could not figure out how they were all related to each other. And so I just plugged into ChatGPT. I have these following companies. I’m looking at this particular product. I provided a link to the product and I said, “Can you explain to me what information is available online that can identify how these companies are related to this particular product?” And it gave me, and I will say, the next day I took the corporate representative’s deposition and it turns out ChatGPT was not 100% correct, but it did give me at least enough of a foundation that I could go into that deposition, not feeling foolish and actually having a little bit more confidence because I had some basis on which to ask my questions. And I felt like I could ask more educated questions because of it.

But I’m really curious about what you’ve talked about, Elizabeth, in using it to sort of develop conversation points. I don’t know if you remember off the top of your head, but when you were putting in your prompt about a difficult conversation, do you remember the kinds of things it was telling you to say or what sort of advice it was giving you on how to handle that conversation? I’m just curious.

Elizabeth McNulty:

Well, it’s interesting because I think when you explain the scenario, you can even name the people, you can give them different names if you want to protect the confidentiality of it, and then you tell it what you’re going to say, and then you can give it potential scenarios on how you think they’ll respond, and it’ll give you ideas for how you should respond if X says this, then try doing this. And it’s almost like having a conversation with a colleague, because that’s something that we would do, and it’ll give you good ideas for how to approach the conversation given different scenarios. So I mean, I do think it’s helpful. I don’t remember the specific prompts it was giving me, but I found it after the conversation that I had with the actual person that I did utilize some of the tips and they were helpful.

Katie St. John:

Kind of on that same line when you were saying about you have these conversations with a colleague or in the context of helping using AI to teach you something, that’s something that you may have had to have a conversation with an expert and it cost the client more money in the end. So for me, I also think the other great thing about ChatGPT is just the availability of it. If you find yourself late night preparing for a deposition, that is a situation in which the expert may not have been available. So that happened to me in a case that we have right now, it’s a medical malpractice case and I was going to take the defendant doctor’s deposition and I found myself really getting into the weeds on the medicine side of everything, but it was just like a conversation. It’s just me and my computer and it was like, I don’t know, 11:30 at night and just talking about the different medical topics.

But that is a really nice feature of AI that I feel like our practice can utilize. It’s just like the availability of it and it is flexible with your schedule. So I found that helpful.

Elizabeth Lenivy:

That’s what I’ve been using it for as well, especially whenever … If you’ve gone through medical records, you find that oftentimes there’s a lot of shorthand used or they’ll report something in there, but obviously the doctors or the nurses, whoever is documenting this is not necessarily explaining everything that they’re writing in there because it’s not meant for non-medical providers to read. It’s meant for other people who are going to understand that language. And so what I’ve been recently doing is I’ll just type into ChatGPT again without identifying anything client specific because, and we’ll talk about that a little bit more, but just in general, I’ll say, “If so- and-so read this blood panel, if these are the results that come with the blood panel and here’s everything else that’s going on with the patient, what does this mean?” Or I’ll type in whatever the shorthand was that used and I said, “Can you basically, this was in someone’s medical record.

Can you translate this for me? ” And then the other thing I will ask because I don’t entirely trust what I’m getting from the AI. And part of this is because I think the AI, really what it’s designed to do is sometimes give the user what it thinks it wants to hear as opposed to what is the correct answer. So I will ask the question, I’ll get the explanation, but I will ask ChatGPT, “Can you cite your sources for me? Can you cite to an article? Can you cite to a website? Give me a book chapter, something I can try to find and look at and confirm for myself.” And that’s been really helpful because again, it’s giving me a basic overview so I can feel more educated on the topic, but then I can do that next step, which is actually confirming that the information I’m getting fed is correct.

Katie St. John:

In that same sense, when there are resources cited, especially in a medical malpractice case, I have also found it helpful maybe if it’s something that the opposing counsel’s expert is citing to take that article and you can actually upload it to ChatGPT and ask for a summary. I felt like that’s helpful also. And since it’s literature that’s already cited, I think that’s a unique feature that can help you kind of … I know in some cases there’s so much medical literature out there to kind of pick out what you want to focus on.

Elizabeth McNulty:

The concern that it’s just going to start making up stuff to tell you what you want to hear is a really valid and we hear about it all the time. It’s like one of the biggest concerns I think we as lawyers have using it. Everyone’s heard of, as you mentioned, the lawyers who cited made up case law from ChatGPT. One way to try to avoid that is I think you can tell it, or I know you can tell it, like a closed universe of sources. So only cite to medical journals or whatever. And I think that that is one way to get around that in this context. I’ve heard another use of it. I haven’t tried it myself, but I was talking to a tax lawyer and there was some new line of cases or something that changed the law and he uploaded a really long Supreme Court opinion and he told it to create a podcast and he just listened to this podcast that ChatGPT created about explaining this new case law to him.

And it even had two different voices and it was like a legit podcast that explained him the law. And so you can just do that while you’re getting ready, going for a walk, working out, whatever. I mean, I think that’s like, I

Don’t know. I feel like the Jetsons, we’re just living in the future, but it’s crazy. I don’t know. I mean, it’s helpful though.

Elizabeth Lenivy:

That sounds bananas to me. And I joked before we started recording and we were just sort of prepping that I feel so tinfoil hat and I’m that weirdo that says, “Hi, ChatGPT and thank you ChatGPT because I want the robot to remember me and like me in case things get out of hand.” But that’s crazy. Although I still, something about that would just make me nervous. I need to still read the case because what if it’s getting it wrong? What if it thinks like, “Well, I know Liz is a plaintiff’s lawyer. Maybe she wants to hear it with more of a positive spin for other plaintiffs.” Or what if there’s something in there like a footnote or one line that someone wrote somewhere? It stresses me out.

Elizabeth McNulty:

Yeah. But also that could be part of your prompt. I want a neutral discussion of what this case law is saying.

Elizabeth Lenivy:

That’s true. That’s

Katie St. John:

True. I was just worried that people were going to stop listening to our podcasts and start creating their own ChatGPT podcast because Elizabeth just told them to.

Elizabeth McNulty:

You’re right. Not only are they coming for our jobs as lawyers, but they’re coming for our jobs as podcasters. And this is where we take a stand.

Elizabeth Lenivy:

Well, maybe this is where we reveal to the audience that we’ve actually been ChatGPT the whole time. But I mean, it is funny because I feel like all the time I’m seeing people talk about, they’re coming for our jobs, the AI is coming for our jobs. And I am trying to envision a world where an AI is doing the job of a plaintiff’s attorney in the capacity that we have. I mean, I guess it’s possible. Can you have an AI lawyer? Some guy tried. He got in trouble for it, actually. That’s

Elizabeth McNulty:

Too

Elizabeth Lenivy:

Bad.

Katie St. John:

Our jobs aren’t going away yet.

Elizabeth McNulty:

No, I don’t think so. But I mean, I do see … I don’t know that the outcomes would be as good, or would they just be more equitable for everyone, meaning both sides, because a computer’s just kind of deciding. I’m thinking auto cases, pre-suit negotiations, how quickly an AI lawyer could just settle a bunch of those. But would the … I don’t know that the plaintiff would be better off, but I also don’t know that the insurance company would be better off. Would it just be a more even split? And I’m not saying I’m for that. I just, I don’t know.

Katie St. John:

I think that’s an interesting thought, but I think that there is so much of our job that is captured by our ability to advocate and to advance certain arguments at different times, given whatever’s going on in the case, that is something that can’t be captured quite yet by AI. But I think as it continues to develop, it will definitely be interesting to see how it goes.

Elizabeth Lenivy:

Yeah. I was thinking about this in terms of how often we have to come up with arguments on the fly and also how part of our job in advocacy is appealing to the human element and the human side of things. And part of this was me thinking about the last rebuttal I had to give in a closing. And I was truly sitting there and I was like, “I know I get to stand up and I got five minutes left to argue. What am I going to say? What points am I going to address? How am I going to reach this jury and explain to them why my client’s case matters and why her injury matters?” And I went up there and I was truly just speaking from the top of my head, but also from my heart and hoping that I was connecting with these people.

I’m not saying an Ai can’t do it, but I don’t know if a jury is going to connect the same way. And also, I’m imagining now this weird world in which we have still a human jury, but with an AI lawyer, which feels very silly and futuristic, but also probably not completely implausible. And I’m looking back at what I learned about this particular MIT study and one of the things that they found with the study participants, they were measuring these essays generated and the things that people were writing. And they found that the essays generated via ChatGPT tended to be formulaic and homogenized, lacking a personal voice and creativity. In fact, up to 80 to 83% of participants couldn’t even recall what they’d written shortly afterwards. And it’s this idea that it’s just going to be this same argument over and over and over again that is not personalized to a client.

And I feel like for clients, they want their personal story told because every injury, every case is different. And I don’t know if an AI is capable of doing that for people in the same way that human beings do, but this is also why I think it’s important when lawyers are using AI that they don’t let it replace their brain and they don’t let it replace the skills that we have to have and that we have to hone in order to be able to do our jobs effectively.

Katie St. John:

I think that’s a good point. And that’s why, I mean, I really just use it right now. It’s kind of like a sounding board or if I find myself at a dead end with something and I know that I need to list more topics and discovery asks some more questions, just kind of using it as a sounding board right now, I feel like is one of the most helpful. I haven’t dabbled in having it actually draft anything, but I’m still a newbie. I’m a couple months.

Elizabeth Lenivy:

I’ve not used it to draft anything, but I did recently, I knew I was going to be making an argument in front of the court and I had all of my arguments ready, but just to make sure that I wasn’t missing anything, was I leaving something on the table that maybe I just hadn’t thought of yet? I plugged in the scenario, I plugged in what arguments I had and I just asked ChatGPT, “Am I missing anything else? Are there any other arguments regarding relevance or why this particular privilege…” It was on a motion to compel, why this particular privilege shouldn’t apply for what the defendants are claiming that I’m not entitled to. And it gave me a handful of ideas. I think maybe one or two of them were worth at least tossing into my outline and just having sort of in my back pocket in the event I needed something else to say or something else to throw out there.

But ultimately I found that at least that the skills that I have and that I’ve been trained on were still superior to what ChatGPT was giving me, but I do like it for an additional idea generator, but it’s similar to what you said, Elizabeth, where I’m treating it more like a colleague as opposed to replacing my brain with it. Something we talked about or we mentioned earlier was though client confidentiality and how we are securely using it when it is something applicable to a particular client as opposed to maybe just a general broad, tell me about this area of law or tell me about this case. And so our firm has been utilizing some AI, not specifically ChatGPT, but some AI platform to help organize cases. And Katie, I know you had a really interesting question to this particular platform, and we won’t name them for their confidentiality, I guess, but you had an interesting question about how are we protecting client information or how is this different than ChatGPT?

So can you tell us a little bit about that?

Katie St. John:

Yeah. So I think just because I’m newer in the artificial intelligence world, I was concerned this specific platform we’re using, what is the difference between using this platform, which is like an AI generated platform to organize different things versus just putting it into ChatGPT. And I think the difference is that, Elizabeth, you’ll correct me if I’m wrong, but that’s considered a closed AI versus ChatGPT, which is just open. And so I mean, I was talking along the lines of like, is this something that we need to let clients know we’re utilizing? Is this something that we need an additional authorization for? And the answer was essentially for the platform we’re using, the answer is no, because it is the closed, so it’s a secured server versus AI, which is open. And so the rep I was talking to advised me not to put anything client specific into ChatGPT if it were medical records, for example, to not drag and drop a medical record into ChatGPT because that’s open.

And then potentially you are exposing your clients to different risk there versus a platform that we utilize that it’s okay because it’s closed.

Elizabeth Lenivy:

Okay. So that’s, I think, something important for any practitioner that may be considering using an AI platform for document organization or for quick summaries, things like that. You need to consider whether or not you are potentially exposing your client’s personal information to the greater worldwide web. Well, I feel like we’ve kind of hit the nail on the head as far as what we are utilizing AI for. And I think it’s important for any attorney to get comfortable with it. I mean, it’s just the next step in technology. This feels like no different than whenever we went from using law libraries to Westlaw. You got to get comfortable with it. It’s the way of the future. And we’ve talked about the importance of idea generation, maybe summaries, quick lessons in whatever complicated topic you may need to quickly educate yourself on. Curious, is there anything else you guys are using ChatGPT for?

Katie St. John:

Yeah, I think if you really just want to get your feet wet before jumping into using it as a practicing attorney, you could just … It also helps plan vacations. If you got small kiddos and you’re driving, you can give ChatGPT your route and say, “Where should I stop along the way? Or where is there a park or a nice restaurant or something?” So might be a good way to just try it out and see how you feel.

Elizabeth Lenivy:

Yeah. The last thing that I wanted to say, just because I found it very funny when I typed into ChatGPT to tell me about this MIT study was that it felt like ChatGPT got a little defensive of itself. It didn’t feel completely neutral. And some of the things that it felt like it needed to give me, oh, this is what it called it, crucial caveats from the researchers. With just 54 participants and results shared before peer review, the study is preliminary. The authors stress the need for more extensive research. Did they? Did they? Despite sensational headlines, the MIT team specifically asked journalists to avoid terms like dumb or stupid because their study didn’t prove a drop in innate intelligence. And then context matters. Some studies suggest well-structured AI tutoring can boost learning. The concern arises when AI becomes a crutch, especially for creative or critical thinking tasks.

So I think that that’s an important takeaway. Also felt, again, a little bit weird that the AI got defensive of itself. And it’s a question of are we approaching singularity? Tinfoil hat back on, but in all seriousness, I do think it is the way of the future. I think lawyers, we need to get comfortable with it. We need to understand it, but we also need to make sure that it’s not replacing our skills as attorneys. And with that, thank you so much for joining us today. Remember, new episodes drop every other Wednesday, of course, unless you want to go to ChatGPT and say, create a specialized heels in the courtroom podcast just for me. Maybe try, maybe I might try it. Actually, I’m really curious what it would spit out, but thank you again for joining us. And if you would like to contact us, you can reach us at [email protected].

Thanks so much.

Announcer:

Thanks for listening to Heels in the Courtroom. At the Simon Law Firm, we know that trial success isn’t just about experience, it’s about strategy, resources, and the power of collaboration. That’s why attorneys across the country partner with us to strengthen their cases and deliver justice for their clients. If you’re interested in working with our team of seasoned trial lawyers, call 314-241-2929. And if you enjoyed the podcast, be sure to subscribe and send us your thoughts at heelsinthecourtroom.Law.