Latest / AI Tools for Practicing Lawyers / Episode 026 Volunteers and Hostages

Transcript

Episode 026 Volunteers and Hostages · AI Tools for Practicing Lawyers

• 0:03
  SPEAKER_00: Welcome to AI Tools for Practicing Lawyers.
• 0:07
  SPEAKER_00: Practical, no nonsense guidance to help attorneys put AI power to work in their practice right now.
• 0:15
  SPEAKER_01: There's a lot of talk about what AI might do to the practice of law.
• 0:20
  SPEAKER_01: But this episode is about what happens when somebody actually does it.
• 0:25
  SPEAKER_01: My guest today is Chad Van Horn.
• 0:28
  SPEAKER_01: His firm is the consumer bankruptcy version of big law, and he isn't experimenting with AI on the side.
• 0:36
  SPEAKER_01: He's putting it into production.
• 0:39
  SPEAKER_01: And some of the results are remarkable.
• 0:42
  SPEAKER_01: Chad built an AI petition review process that he says has reduced amendments in his firm by 80%.
• 0:51
  SPEAKER_01: He's using AI to analyze bank statements, review pleadings, check local rules and judges' procedures, and train lawyers and paralegals.
• 1:00
  SPEAKER_01: And he has a simple philosophy for rolling it out.
• 1:03
  SPEAKER_01: In his firm, he keeps volunteers, not hostages.
• 1:08
  SPEAKER_01: If you don't want to work into these great AI tools, you don't have to.
• 1:12
  SPEAKER_01: At least not in his firm.
• 1:14
  SPEAKER_01: But this conversation gets much more interesting than just a list of AI success stories.
• 1:20
  SPEAKER_01: Because Chad also thinks that AI is making us dumber.
• 1:26
  SPEAKER_01: We talk about what happens to young lawyers when AI starts doing the work that used to teach them how to practice.
• 1:34
  SPEAKER_01: We talk about why the mundane work still matters, why bankruptcy is incredibly easy to file, but extraordinarily dangerous to get wrong.
• 1:44
  SPEAKER_01: And why human judgment may become more valuable, not less, as AI gets better.
• 1:50
  SPEAKER_01: And Chad makes a prediction that should get the attention of every bankruptcy lawyer listening.
• 1:56
  SPEAKER_01: Firms are going to need AI to handle the coming increase in filings, and lawyers who refuse to adapt may eventually find themselves unable to compete.
• 2:07
  SPEAKER_01: So this isn't an episode about whether lawyers should use AI.
• 2:11
  SPEAKER_01: It's about what the practice of law looks like once they actually do it.
• 2:17
  SPEAKER_01: Welcome to another episode of AI Tools for Practicing Lawyers.
• 2:21
  SPEAKER_01: I'm Ron Dresher.
• 2:22
  SPEAKER_01: I'm Heather Gardner.
• 2:24
  SPEAKER_01: And today's guest is Chad Van Horn, the founding partner of Van Horn Law Group and a dual board-certified business and consumer bankruptcy attorney.
• 2:36
  SPEAKER_01: Chad has served as attorney of record in more than 11,000 cases and leads a multi-office firm with a team of more than a hundred people.
• 2:50
  SPEAKER_01: Already I have a headache.
• 2:52
  SPEAKER_01: Last week, Chad presented at NACPA's Me, Myself, an AI webinar where he showed lawyers how to turn repeatable firm procedures into reusable AI skills.
• 3:04
  SPEAKER_01: Today we're going deeper, where AI genuinely helps a bankruptcy practice, where it does not belong, and what responsible adoption requires, but it has to work across an entire firm.
• 3:18
  SPEAKER_01: Chad, thank you so much for being here.
• 3:20
  SPEAKER_02: Ron, Heather, so happy to be here with you.
• 3:23
  SPEAKER_02: I'm glad you got to me quickly, too, before I forgot everything that I presented last week.
• 3:28
  SPEAKER_01: That's all right, because we have so much to ask you.
• 3:32
  SPEAKER_01: I I don't expect we're going to get finished today.
• 3:34
  SPEAKER_01: We're going to have to have you back again.
• 3:36
  SPEAKER_01: But you've handled more than 11,000 cases.
• 3:41
  SPEAKER_01: What first convinced you that AI was useful in an actual bankruptcy practice instead of it's just an interesting experiment?
• 3:49
  SPEAKER_02: Well, it actually took me a while.
• 3:51
  SPEAKER_02: I'm a relatively late adopter for early adopters, if that makes sense.
• 3:56
  SPEAKER_02: Um I started with ChatGPT, loved what it was doing for my personal life.
• 4:00
  SPEAKER_02: I was at uh I was utilizing, as most people first utilized it, you know, help draft emails, help with social media posts, check different things that you want to purchase.
• 4:10
  SPEAKER_02: And then I heard about Claude, and I thought Claude was just for nerds, uh, which may be right, but I I uh was a little nervous to get involved.
• 4:19
  SPEAKER_02: And then I finally got involved with Claude and I saw the power that it really had to make a difference, especially with mundane tasks or tasks that humans aren't really that good at.
• 4:29
  SPEAKER_02: And uh I was hooked from there.
• 4:31
  SPEAKER_02: So I uh I met with my uh develop my developers earlier today.
• 4:35
  SPEAKER_02: I have in-house developers that work on different projects for me.
• 4:38
  SPEAKER_02: I told them I'm dreaming about different projects that we're working on now.
• 4:42
  SPEAKER_02: So I think AI might might be just taking over my whole body at this point.
• 4:46
  SPEAKER_01: What did you see from Claude that convinced you that you could go to a higher level?
• 4:53
  SPEAKER_02: So uh the analysis that that it would do of large amounts of data that we're putting in because for for some reason nowadays our our clients and maybe most individuals have so many bank accounts.
• 5:04
  SPEAKER_02: And that's where it started is with with bank accounts.
• 5:07
  SPEAKER_02: I I I hated going through with my highlighter, going through each bank account.
• 5:11
  SPEAKER_02: Of course, each bank account has a hundred transactions, but it's all like$12,$5,$7.
• 5:17
  SPEAKER_02: And you're just going through it, and it seems like wasted effort, but but there's always that one that you have to catch.
• 5:24
  SPEAKER_02: There was a transfer, a cash withdrawal, or something.
• 5:27
  SPEAKER_02: So I put that into Claude.
• 5:28
  SPEAKER_02: I said, Can you do something with this?
• 5:30
  SPEAKER_02: And it gave me this beautiful report that I manually checked it against, and it actually did a better job than my report.
• 5:38
  SPEAKER_02: They said, okay, that this definitely has something.
• 5:40
  SPEAKER_02: And especially in the bankruptcy world that we're in, it's dealing with a lot of data and a lot of repetitive tasks that we have to do over and over again.
• 5:49
  SPEAKER_02: So if you can build out those skills and then refine those skills, then you can, then you're you basically have a highly, highly trained associate or paralegal that of course you still have to double check, um, but will give you some detailed reporting.
• 6:02
  SPEAKER_01: So let's talk about the bank statement example.
• 6:05
  SPEAKER_01: How did you actually go about building something sufficiently robust to support your hundred-person team?
• 6:15
  SPEAKER_02: Well, initially, I just built it uh for me.
• 6:18
  SPEAKER_02: So the way the way that I work is I'll build my skills personally and have Claude help me build out those skills.
• 6:25
  SPEAKER_02: And then I test them myself.
• 6:27
  SPEAKER_02: And then once I get them to a point, then I'll then I'll share the skills and I'll train the team on how to utilize those.
• 6:33
  SPEAKER_02: We we have weekly meetings uh with the team members.
• 6:36
  SPEAKER_02: Not everybody has access to AI on on the team.
• 6:40
  SPEAKER_02: So we still have people that are are manually doing things and and that and I like seeing the side by side because I want to make sure obviously that the quality levels are are either staying the same or increasing, and then uh the ability to handle more data obviously is happening with AI.
• 6:57
  SPEAKER_02: So so the way I I roll it out, it's it's much slower than you could roll it out, but that's more from a quality control standpoint because uh you know we all see AI gone wild doing things, and and still uh when it reviews reviews some petitions and different things that we have it doing in-house, it'll it'll return information that isn't quite where we want to go or accurate.
• 7:20
  SPEAKER_02: So I'm taking that out and I'm teaching the skill.
• 7:23
  SPEAKER_02: I said, don't do this again, and it's it gets better at uh gating that information.
• 7:28
  SPEAKER_01: Do you think that your AI adoption is a work in progress?
• 7:33
  SPEAKER_01: Or do you feel like like it's pretty well set for your firm right now?
• 7:37
  SPEAKER_02: I think it's a work in progress and it always will be because AI is ever evolving and it's gonna get and it's gonna get better at some things.
• 7:45
  SPEAKER_02: And I do a weekly update on my skills.
• 7:47
  SPEAKER_02: So I say, okay, let's let's look at the last week of things that we did.
• 7:51
  SPEAKER_02: Let's look at my feedback that I provided, and let's update this skill to make it more efficient and just better moving forward.
• 7:58
  SPEAKER_02: And uh sometimes, unfortunately, when I do that, it actually the skill regresses, and then I have to roll it back to the previous generation because I'm like, I know I like the other stuff that you were doing.
• 8:10
  SPEAKER_02: I just wanted you to fix A, B, and C and become more efficient because you know those tokens are extremely valuable.
• 8:17
  SPEAKER_02: And if we can use uh lower versions of AI that uh cost less tokens, then that'll save you some money over the long run, especially when you have you know a staff of a hundred utilizing Claude or or its AI.
• 8:31
  SPEAKER_01: How do you decide which team members have access to the AI?
• 8:35
  SPEAKER_02: So initially, I I want the ones that want access to the AI.
• 8:40
  SPEAKER_02: So I I asked, I want them to be excited about this.
• 8:43
  SPEAKER_02: Isn't something that I want to force upon people?
• 8:46
  SPEAKER_02: So we've we've had discussions internally and said, okay, do we give this to our highest performers, uh, which could make sense?
• 8:53
  SPEAKER_02: Do we give this to our lowest performers?
• 8:55
  SPEAKER_02: Which so that that can help them compete with the highest performers.
• 8:59
  SPEAKER_02: I can see an argument for that.
• 9:01
  SPEAKER_02: For me, I wanted the volunteers.
• 9:03
  SPEAKER_02: My thing is I want volunteers, not hostages.
• 9:06
  SPEAKER_02: And then let's work with those volunteers to show how it really can benefit them.
• 9:11
  SPEAKER_02: And then what what happens, and inevitable, inevitably this happens.
• 9:16
  SPEAKER_02: Hey, uh my legal assistant wants access to it now because it's it sees how it made me more efficient in this, ran that report, helping me with my emails, and then that gets other people excited about it.
• 9:27
  SPEAKER_02: And now I have a power user that can help train another user, and it's and it's already in that department, it already has shown that it has significant value.
• 9:37
  SPEAKER_02: So I would recommend look for look for volunteers, look for people that are excited about it, and we'll provide feedback.
• 9:44
  SPEAKER_01: Well, all right.
• 9:45
  SPEAKER_01: What would you tell to a solo practitioner who's got one team member who is not about to volunteer, but wants to bring AI into their practice?
• 9:55
  SPEAKER_01: How would you help that lawyer adopt over reluctance?
• 10:01
  SPEAKER_02: Well, I actually have that scenario in my firm because you know we we have attorneys in my firm that are in their 20s to in their 80s.
• 10:09
  SPEAKER_02: So we have that whole gambit, if you will.
• 10:11
  SPEAKER_02: And just because you're in your 20s doesn't mean you're tech savvy, and just because you're in your 80s does not.
• 10:16
  SPEAKER_01: That's true.
• 10:17
  SPEAKER_02: So I I find that very interesting.
• 10:19
  SPEAKER_02: I you would many people make the mistake thinking that uh, you know, the next generation really knows about computers and tech.
• 10:26
  SPEAKER_02: And that's not always the case.
• 10:28
  SPEAKER_02: So I have that scenario in in my firm.
• 10:30
  SPEAKER_02: So what I do is I have them, when they have certain things that I know AI can help them with, I have them use me as their AI agent, if you will.
• 10:39
  SPEAKER_02: I said, hey, whenever you have that issue, just email it to me, and then I'll run it through AI and send it right back to them.
• 10:45
  SPEAKER_02: And initially they're like, man, you're turning around this really quickly.
• 10:49
  SPEAKER_02: This is really well done.
• 10:50
  SPEAKER_02: And then I finally break the news to them.
• 10:53
  SPEAKER_02: It wasn't me.
• 10:54
  SPEAKER_02: I'm you're doing the final review.
• 10:56
  SPEAKER_02: I do a final review of what Claude has provided, but this is actually Claude doing it.
• 11:00
  SPEAKER_02: And they're like, oh my gosh, all right, I need to learn this.
• 11:03
  SPEAKER_02: So you really have to show proof of concept.
• 11:06
  SPEAKER_02: And honestly, it it's it's better for me, even if I have to do that for forever, it's still going to be more productive than having the individual try to figure it out on their own.
• 11:16
  SPEAKER_02: Especially for some of these, it's you know, a two, three hour task that's done in seven minutes.
• 11:22
  SPEAKER_01: You know, what I'm hearing is that your firm doesn't have a tech stack.
• 11:28
  SPEAKER_01: I'm hearing that you've got multiple tech stacks within the firm, depending upon who's on the AI team and who's in the dugout to get into the AI team.
• 11:39
  SPEAKER_01: Is that a fair assessment?
• 11:42
  SPEAKER_02: So we we uh we invested heavily in a company called Glade, as far as the time and effort uh that is a practice management system, but also bankruptcy software, so it helps build the schedules.
• 11:53
  SPEAKER_02: So everybody in in our firm is on GLAD.
• 11:57
  SPEAKER_02: And and GLAD has AI tools built within it.
• 12:00
  SPEAKER_02: Some utilize it, some don't.
• 12:02
  SPEAKER_02: And then on on top of that, then we have Claude on and we have Google Workspace, which has Gemini built in, which Gemini is is good for certain things and and it and it can be helpful.
• 12:15
  SPEAKER_02: But I think for more complex tasks, you know, we have to use Claude.
• 12:18
  SPEAKER_02: So the the short answer is yes.
• 12:21
  SPEAKER_02: We we have multiple uh stacks based on which department, uh, but it all filters into uh Glade.
• 12:29
  SPEAKER_01: You know, I'm really glad you brought up Glade because on the Facebook consumer bankruptcy lawyer groups, every three weeks there's a lone voice in the wilderness that says, Has anybody tried Glade?
• 12:44
  SPEAKER_01: And you know, and nobody seems to answer because I don't really know why.
• 12:50
  SPEAKER_01: So I got two questions.
• 12:51
  SPEAKER_01: First of all, what was the adoption into Glade like?
• 12:55
  SPEAKER_01: And and why do you think lawyers are so unwilling to talk about their experience with Glade?
• 13:01
  SPEAKER_02: Um so the adoption, I think we're I don't know, 18 months into the adoption at this point.
• 13:07
  SPEAKER_02: Uh and this is the first time I've ever seen software developed, especially this complex software developed from infancy to where it is now.
• 13:16
  SPEAKER_02: What I love about Glade is it is a one-stop shop.
• 13:20
  SPEAKER_02: Uh before before we had Glade, you name it, we had it Clio, Lidify, Salesforce, Hubs, HubSpot, kind of all Frankenstein together to uh best case, Sin Compass, Easy Filing.
• 13:34
  SPEAKER_02: I can go back, go back forever.
• 13:36
  SPEAKER_02: We we tried them all, and uh none of them have has worked as one unit together, like Glade.
• 13:45
  SPEAKER_02: And what I really like about Glade now, more than even when we started, is the growth that we're seeing because the AI is getting better and it's built for that, right?
• 13:56
  SPEAKER_02: And I think that we're we're seeing in in the market, the older technologies that weren't built for AI are struggling to say, okay, do we completely scrap that and build a new AI version of it?
• 14:10
  SPEAKER_02: Or how do we build AI into it?
• 14:13
  SPEAKER_02: And Glade does not have that issue because it was built for AI.
• 14:17
  SPEAKER_01: All right.
• 14:18
  SPEAKER_01: So, what has been the most exciting thing that you've seen in your firm as a result of adopting GLADE?
• 14:25
  SPEAKER_02: So much.
• 14:26
  SPEAKER_02: So uh I love meeting with the different departments and seeing what their wins are.
• 14:31
  SPEAKER_02: Uh, even like I met with my accounting department, and they used to have to manually input all these payments every day.
• 14:36
  SPEAKER_02: And now that's been automated with with our QuickBooks, seeing the income organizer uh really organize all the different pay stubs that are coming in automatically, and then with a paralegal and lawyer review on top of that.
• 14:50
  SPEAKER_02: So my paralegals and I when I started, I I did all the inputting as well.
• 14:54
  SPEAKER_02: I said easy filing earlier.
• 14:56
  SPEAKER_02: That was my favorite software before Glade.
• 14:58
  SPEAKER_02: It was it was the most basic software.
• 15:01
  SPEAKER_02: Basically, you just inputted the information, but it was frustrating because you'd have to input the information multiple times over.
• 15:07
  SPEAKER_02: So, my most exciting is just seeing how much more efficient that we're getting, and my team can spend more time with the clients or really solving problems instead of just doing basic data entry.
• 15:19
  SPEAKER_01: So you mentioned that you've got Glade and you've also got Google workspace in your firm.
• 15:25
  SPEAKER_01: Where does where does one leave off and the other one pick up?
• 15:28
  SPEAKER_01: Or how do they how do they coordinate with each other?
• 15:32
  SPEAKER_02: Yeah, so and the integrations have got gotten much better.
• 15:36
  SPEAKER_02: So, you know, Google Drive, Gmail, uh Google Calendar, we're they're working on something now where basically every email that comes in, it automatically will tag to a file and say, do you want to send this email to the file?
• 15:50
  SPEAKER_02: So it'll all go to centralized notes.
• 15:52
  SPEAKER_02: So that that's the that's the good part about it.
• 15:55
  SPEAKER_02: And that's actually a newer integration.
• 15:57
  SPEAKER_02: So uh it's good to have backup systems because Glade has its own calendar.
• 16:02
  SPEAKER_02: It's keeping all the CMECF events.
• 16:04
  SPEAKER_02: We also have that on our Google Calendar as as well.
• 16:07
  SPEAKER_02: So it would be nice to have that fully integrated.
• 16:11
  SPEAKER_02: It's not there yet, but I think it's good to have uh two systems in that case.
• 16:16
  SPEAKER_03: Where has AI saved you the most employee time?
• 16:20
  SPEAKER_02: Oh my gosh.
• 16:23
  SPEAKER_02: So, really throughout the whole process, we we used to have this intake process, and and this is going back years where we had like intake forms that would be would be filled out, and then that form would go into uh the system, and then we'd have to do the engagement letter.
• 16:41
  SPEAKER_02: And then basically what would happen previously was such a manual process, and now somebody goes to my website, they sign up for a consultation, all the information is already taken in as far as their contact information and already put into GLADE.
• 16:57
  SPEAKER_02: They can fill out a basic questionnaire that gets us that information.
• 17:01
  SPEAKER_02: So then when I'm doing the consultation with the client, that saves a lot of time because they've already put in a lot of the information.
• 17:08
  SPEAKER_02: My notes go directly into GLAD, and then it will automatically generate the engagement letter and email it to the client directly.
• 17:17
  SPEAKER_02: They sign up, it'll it'll work out where it's setting them up with a payment plan right from the beginning.
• 17:22
  SPEAKER_02: So there's no human interaction necessarily needed, and then it'll tell them what documents that that are needed and it will follow up with that as well.
• 17:30
  SPEAKER_02: So it it really has solved the duplicate entries and the need for humans to keep emailing, hey, we we need this document, or hey, we need that.
• 17:41
  SPEAKER_02: Obviously, we still send people still send text messages, we still call, we still have in-person meetings.
• 17:47
  SPEAKER_02: But for somebody that just wants to get it done, they can basically go beginning to end with AI and us working together instead of having all the different meetings.
• 17:59
  SPEAKER_02: So I for me, the biggest, biggest win and time saver is just not having that duplicate entry, like now they retain, now they got to go into QuickBooks.
• 18:07
  SPEAKER_02: Okay, now they got to go into best case, now they got to go into Lidify, they got to go into Clio, whatever it is.
• 18:13
  SPEAKER_02: And we never found that way to completely automate that.
• 18:17
  SPEAKER_02: And even when we did find ways to automate portions of it, if one of the software is changed, then it would break the connection, and then we'd have to figure out how to reconnect it.
• 18:26
  SPEAKER_02: Uh and so that would take time as well and create headaches.
• 18:31
  SPEAKER_02: So now it's just one long chain with what which I love.
• 18:36
  SPEAKER_01: So where does where does how do you coordinate between Claude and Glade in your office?
• 18:43
  SPEAKER_02: So Claude and Glade actually have a connector.
• 18:47
  SPEAKER_02: So if if uh so I can go into Glade and and I can ask uh Glade questions, I can have it create reports, I can have it, I can say, I, you know, I need you to make this edit for me in the file.
• 18:59
  SPEAKER_02: I need you to add a note.
• 19:00
  SPEAKER_02: And then, but also I can have Claude go in there and and I can ask Claude information about Glade and it connects directly into Glade.
• 19:08
  SPEAKER_01: All right.
• 19:09
  SPEAKER_01: So let's talk about your NACBA presentation.
• 19:12
  SPEAKER_01: You compared writing a skill to giving instructions to a new paralegal.
• 19:17
  SPEAKER_01: I what must those instructions contain before they can be trusted across a firm?
• 19:23
  SPEAKER_02: So again, it would be just like training a paralegal.
• 19:26
  SPEAKER_02: If I give a paralegal an assignment and they bring it back to me, I don't say, oh, it's good to go.
• 19:30
  SPEAKER_02: You know, I'll do a detailed red line review of that and give that back to them and explain to them why it needs to be that way.
• 19:37
  SPEAKER_02: Same thing with the Claude skill.
• 19:39
  SPEAKER_02: If it gives you back something that you don't like, I was just meeting with one of my attorneys.
• 19:43
  SPEAKER_02: They said, I loved everything, but I didn't like the font that it came back in it.
• 19:47
  SPEAKER_02: I said, Well, did you tell it what font you wanted?
• 19:50
  SPEAKER_02: He said, No, I thought it was good enough.
• 19:52
  SPEAKER_02: I said, No, you got to tell it.
• 19:53
  SPEAKER_02: Hey, this has to be 12 point times new Roman fault uh font moving forward.
• 19:58
  SPEAKER_02: I want to see that every time.
• 20:00
  SPEAKER_02: And then and then it'll update the skill based on that.
• 20:03
  SPEAKER_02: You hit update.
• 20:04
  SPEAKER_02: It's it's a very quick process.
• 20:06
  SPEAKER_02: So run you run the skills, you review the results, and and then if you don't like the results, you tell them what you don't like so that they can improve.
• 20:15
  SPEAKER_02: That's the same thing that I would expect from any of my employees.
• 20:18
  SPEAKER_02: If they do something wrong and I don't tell them what they're doing wrong, it's not their fault if they continue doing something wrong.
• 20:24
  SPEAKER_02: It's mine because I need to train them.
• 20:26
  SPEAKER_02: Same thing with AI.
• 20:28
  SPEAKER_02: If if it's doing something wrong, you're giving it the wrong prompts, you're giving it your and and you're not updating it the correct way because eventually it will get there.
• 20:36
  SPEAKER_02: And sometimes it will be frustrating, but sometimes it's frustrating with the employees as well.
• 20:43
  SPEAKER_01: You talked a little bit about a pleading reviewer that can check statutes, rules, local rules, judges' procedures, negative notice language, service lists, and proposed order formats.
• 20:56
  SPEAKER_01: Now, those things are evolving all the time.
• 21:00
  SPEAKER_01: How does the system keep up with them?
• 21:03
  SPEAKER_02: Yeah, well, well, I I drop in where to look for for that information.
• 21:08
  SPEAKER_02: And uh in in bankruptcy, we are it's a federal code, so it's it's all the same, but it is so different, judge to judge, uh, district to district, trustee to trustee.
• 21:21
  SPEAKER_02: Like there's even in in one of the districts that we practice in, like the order has to specifically have the signature line in one place, and if it's in the other place, it's gonna get kicked back to you.
• 21:33
  SPEAKER_02: And you need to know that.
• 21:34
  SPEAKER_02: And uh, and the only way that you can know that obviously is if you practice there regularly.
• 21:39
  SPEAKER_01: And had one and had one kickback.
• 21:41
  SPEAKER_01: And you probably had one kickback, exactly.
• 21:44
  SPEAKER_02: So so you can you can teach AI.
• 21:46
  SPEAKER_02: And and the good thing about the AI as well is it'll actually look through and see kickbacks, and it'll say, Hey, we noticed that an order was kicked back that was similar to this for this, this, and this reason.
• 21:59
  SPEAKER_02: Would you You like us to update based on that this the skill because this judge wants it that way.
• 22:04
  SPEAKER_02: It said, actually, yes, I I would, because you know, we don't want to waste the judge's time.
• 22:09
  SPEAKER_02: We want to make sure that we're we're putting the right information there.
• 22:12
  SPEAKER_02: But we also point uh the AI to the judge's websites and uh to the local rules, to the local forms, and we tell it to make sure that you go there and you're updating your information based on that.
• 22:24
  SPEAKER_02: Fortunately, it's really only updated about once a year every year or a few years that we see.
• 22:30
  SPEAKER_02: But obviously, there's new judges coming in, and the judge can update update things whenever.
• 22:34
  SPEAKER_02: If I see something come in, because usually they'll email us, I will feed that to Claude as well so that it'll it'll know that.
• 22:42
  SPEAKER_03: Chad bankruptcy filings have increased quarter over quarter since 2022.
• 22:46
  SPEAKER_03: We all in the industry know this.
• 22:48
  SPEAKER_03: With chapter seven's up significantly, I think around 29% since last year.
• 22:52
  SPEAKER_03: We've had over 50% increase in some chapter fives in the first half of this year.
• 22:56
  SPEAKER_03: If it continues rising, and we all believe it will at this rate, can the bankruptcy ecosystem even absorb this?
• 23:03
  SPEAKER_03: And can AI be a multiplier and help that capacity solution?
• 23:08
  SPEAKER_02: Yeah, that's a wonderful question.
• 23:10
  SPEAKER_02: And uh that was a big question at the national conference in Boston.
• 23:15
  SPEAKER_02: And I know it'll come up again in Orlando.
• 23:17
  SPEAKER_02: For whatever reason, the next generation of bankruptcy attorneys really haven't come through.
• 23:23
  SPEAKER_02: Uh I'm I'm still considered uh a young bankruptcy attorney, if you will.
• 23:27
  SPEAKER_02: Uh, I don't hide it well with the gray in my beard because just the next generation really hasn't hasn't come through, especially for consumer bankruptcies.
• 23:37
  SPEAKER_02: So, what I and and many poor people in the industry, what we're gearing up to do is be able to handle more of these cases by utilizing AI tools.
• 23:47
  SPEAKER_02: And we're we're getting close.
• 23:49
  SPEAKER_02: Actually, I shouldn't say we're getting close.
• 23:50
  SPEAKER_02: We we're already doing it, but we're not near where we're gonna be.
• 23:54
  SPEAKER_02: And where we were a year from now, if you like I would have never guessed it could have grown as quickly as it did.
• 24:01
  SPEAKER_02: Because I I remember a year ago, you know, trying some of these tools, and I'm like, I it can't really draft, it can't format, it can't, you know, do all these things.
• 24:10
  SPEAKER_02: Now it can do all of that.
• 24:11
  SPEAKER_02: And uh especially if you uh put it with co-counsel, which is Westlaw, which is another connector that it has, it does incredible, incredible work.
• 24:21
  SPEAKER_02: So the legal industry as a whole is gonna change, but especially our sect within bankruptcy is gonna change significantly over the next few years.
• 24:30
  SPEAKER_02: And and practitioners, it they're gonna have one of two options, really.
• 24:35
  SPEAKER_02: They're gonna have to get on board or they're gonna be forced into retirement because they they just won't be able to compete, unfortunately.
• 24:42
  SPEAKER_02: Um, and it's a shame because most bankruptcy practitioners are solos, and you know, they're in communities where people really need them.
• 24:51
  SPEAKER_02: So I I appreciate podcasts like this.
• 24:54
  SPEAKER_02: That's why I'm so excited to be on here, where we're getting the word out that you know, this is here, it's not as difficult as it seems.
• 25:02
  SPEAKER_02: Start with a small task and just make yourself more efficient.
• 25:07
  SPEAKER_02: It's much easier if you're you're a solo to just learn this, set up your skills.
• 25:12
  SPEAKER_02: I'm happy to answer any questions and people can email me directly and just get yourself set up where you can handle double, triple the amount of work, but put in the same hours.
• 25:24
  SPEAKER_01: Well, I'll tell you what, I'll definitely give you the opportunity to drop your email address into our show notes.
• 25:31
  SPEAKER_01: And then you'll decide if you regret giving that very welcoming invitation to our listeners.
• 25:38
  SPEAKER_01: But, you know, before Heather asked the question, I was just getting ready to ask a follow-up to something you said where you point the AI to judges, websites, et cetera.
• 25:50
  SPEAKER_01: And how do you do that?
• 25:52
  SPEAKER_01: What's what's somebody's gonna say, I want that, I want that capacity, I want that ability.
• 25:58
  SPEAKER_01: What do they have to do to get there?
• 26:00
  SPEAKER_02: All you do is you go to the link and they and you drop it into Claude and say, reference this link every time that you review a pleading.
• 26:07
  SPEAKER_03: In the skill, right?
• 26:08
  SPEAKER_03: In your skill.
• 26:09
  SPEAKER_02: Yep, in your skill, right?
• 26:11
  SPEAKER_02: Reference these local rules, reference these local forms, reference this judge's page, and just drop it into Claude and say, and then say add to skill or update skill.
• 26:22
  SPEAKER_01: All right.
• 26:23
  SPEAKER_01: So you also talked about your the petition review.
• 26:27
  SPEAKER_01: That that was really interesting to me.
• 26:29
  SPEAKER_01: How you you said that that increased your you had an 80% figure in there.
• 26:35
  SPEAKER_01: Could you could you talk to us a little bit about what that was?
• 26:38
  SPEAKER_01: Because that was a very impressive number.
• 26:41
  SPEAKER_02: Yeah, so uh we we started doing a quality control where we do the paralegal prepares the petition, reviews it with an attorney.
• 26:49
  SPEAKER_02: The attorney does an independent review, and then they review it with the client.
• 26:53
  SPEAKER_02: What I added was a step where a quality control step where before the petition is filed, it's sent to me.
• 27:00
  SPEAKER_02: I run it through a petition reviewer, and then I send it back to the attorney and the paralegal so that it's updated, and then they update and they get the get the client to sign off on it.
• 27:12
  SPEAKER_02: What this has done is it has reduced amendments by 80%.
• 27:17
  SPEAKER_02: So meaning that we're saving so much more time because the information is accurate and and everything is filled out accurately on the schedules so that amendments don't need to be done in the future.
• 27:29
  SPEAKER_02: And what this is doing on top of that, it's training my attorney and it's training my paralegal on things that they might not know or things that they might miss.
• 27:39
  SPEAKER_02: And what I do is I have it automatically create an email that goes back to to uh to them, and I'll go through and I'll verify everything against the petition, and then I'll write little notes in there that that explain the why, and I'll have them talk to me if they if they have any pushback.
• 27:55
  SPEAKER_02: And I and I always encourage uh them and anyway, anybody to push back on anything that I say or anything that Claude says, because I'm certainly not always right and Claude isn't either, but it should should at least be utilized as a starting point and uh as a conversational.
• 28:11
  SPEAKER_01: You know, you you've got a big firm.
• 28:13
  SPEAKER_01: And you know, Jenny Doling, who was on your panel at the Me, Myself, and AI, she talked about putting in place an AI policy even before you have AI.
• 28:26
  SPEAKER_01: Have you done something like that?
• 28:28
  SPEAKER_02: For sure.
• 28:29
  SPEAKER_02: As soon as Gemini came out, rolled out to Google Workspace, we put the AI policy in place because as we discussed on on during the seminar, you may be using AI and not even knowing it.
• 28:41
  SPEAKER_02: So if you're on Google Workspace, you're using AI.
• 28:45
  SPEAKER_02: And it's just called Gemini.
• 28:47
  SPEAKER_02: And you might feel better about it because you're familiar with Google, but you're still utilizing AI.
• 28:54
  SPEAKER_02: So you you you need an AI policy.
• 28:56
  SPEAKER_02: You probably need something about AI in your engagement letter uh sooner than later.
• 29:01
  SPEAKER_02: And and the good news is if you truly aren't using AI, then it's not gonna hurt you to have that in place.
• 29:09
  SPEAKER_02: But the way things are going, you're gonna be utilizing AI more and more moving forward.
• 29:16
  SPEAKER_01: So you talk about your Google workspace and Gemini.
• 29:21
  SPEAKER_01: Do you have any set workflows that that use Gemini as opposed to other tools because you're a workspace user?
• 29:30
  SPEAKER_02: So we we do, and we we've set up gems for uh for the team because again, it's it's free, if you will, because everybody has access to it based on having access to Google Workspace.
• 29:42
  SPEAKER_02: So as far as different sheets that we use and automating, automating different things, and even within the emails, uh, you know, Gemini will give prompted responses.
• 29:52
  SPEAKER_02: I I don't find their prompted responses extremely helpful other than the ones that are very basic, like thumbs up or will do.
• 30:00
  SPEAKER_02: You know, it's really uh and I don't understand why.
• 30:03
  SPEAKER_02: I if I would if I were a betting man, I would have said Gemini would be winning by a large margin by now just because they had more data than anybody else when it started.
• 30:12
  SPEAKER_02: Obviously, I would have lost a lot of money if I would have made that bet because they are not dominating uh the space like I thought they would have.
• 30:20
  SPEAKER_02: So as far as what we're using Gemini for, it really is just for anything within directly within Google Workspace.
• 30:27
  SPEAKER_02: And even with that, uh, we're finding Claude is more efficient at reviewing the data within Google Workspace.
• 30:34
  SPEAKER_02: So I'll I'll always fall back to utilizing Claude.
• 30:38
  SPEAKER_01: I'll I have a theory.
• 30:39
  SPEAKER_01: I and I and I've said this on the podcast before, and I I've said it a little bit more off of the podcast, and I I wouldn't uh step away from that Gemini bet for a couple of reasons.
• 30:52
  SPEAKER_01: First of all, if you want to get to the enterprise level, Gemini is far and away the cheapest bet out there.
• 30:59
  SPEAKER_01: Last I checked, you could get an enterprise license for$36 a seat per month, and nobody else comes close to that.
• 31:08
  SPEAKER_01: First of all, and second of all, I'm not convinced that we have reached the real economic level of the cost of tokens for companies like anthropic and open AI.
• 31:24
  SPEAKER_01: And I'm wondering, well, there's gonna come a time when when the you know the the the money runs out, the the support runs out, and and you've got to raise the price of the tokens to actually be profitable.
• 31:36
  SPEAKER_01: Am I way off base, or do you you think we're we're probably in a pretty good position for the cost of tokens moving forward?
• 31:45
  SPEAKER_01: Yeah, I I I could see that.
• 31:46
  SPEAKER_02: So it I think there's gonna be two forces that push it.
• 31:49
  SPEAKER_02: I think tokens will become cheaper just because as as it evolves, but but eventually they're gonna be pushed to to make profits.
• 31:57
  SPEAKER_02: So you're basically the tokens that you're utilizing today in theory will be cheaper and more efficient later, but you might they might stop passing along those savings.
• 32:09
  SPEAKER_02: Does that make sense?
• 32:10
  SPEAKER_02: It does.
• 32:11
  SPEAKER_01: And they may they may not support the lower tiers after a certain point, right?
• 32:16
  SPEAKER_01: By the way, ChatGPT has started putting out documentation that looks remarkably like what Claude puts out.
• 32:26
  SPEAKER_01: And they're still not where Claude is, but they're getting pretty close to the quality of their output.
• 32:35
  SPEAKER_01: Uh do you play around with Chat GPT at all?
• 32:38
  SPEAKER_02: Yeah, ChatGPT is my personal friend still.
• 32:40
  SPEAKER_02: So I I have everything.
• 32:42
  SPEAKER_02: I have a seven-month-old daughter, so that Chat GPT has helped me raise.
• 32:46
  SPEAKER_02: Uh, my my wife and I, ChatGPT has helped us raise our daughter.
• 32:50
  SPEAKER_02: Obviously, you know, we we uh we have our our our doctors and everybody else as well.
• 32:54
  SPEAKER_02: And it's funny, we you know, we had our our daughter had something going on, and uh my wife put it in Chat GPT.
• 33:01
  SPEAKER_02: I said, Did you text the doctor though?
• 33:03
  SPEAKER_02: Like maybe we should start with that.
• 33:04
  SPEAKER_01: So the doctor before ChatGPT.
• 33:07
  SPEAKER_01: All right, here's a tough question.
• 33:10
  SPEAKER_01: Is is AI making us smarter or stupider?
• 33:14
  SPEAKER_02: Overall, I would say dumber for sure.
• 33:16
  SPEAKER_02: And and it the the problem that I'm seeing is it's attacking critical thinking.
• 33:22
  SPEAKER_02: And that is the biggest issue.
• 33:24
  SPEAKER_02: I don't think it's just AI.
• 33:26
  SPEAKER_02: I think it's been going on for a while.
• 33:28
  SPEAKER_02: But the the biggest talent that I look for in a new hire is somebody with critical thinking skills.
• 33:34
  SPEAKER_02: Like I want problem solvers.
• 33:36
  SPEAKER_02: And yes, Chat GPT, Gemini, all these things can help you solve problems.
• 33:42
  SPEAKER_02: And there's always been things that can help you solve problems for thousands of years, right?
• 33:47
  SPEAKER_02: There have been things that can help you solve problems, but you need to actually go out and figure out how to solve the problem, not just come up with a theory about how to self solve the problem.
• 33:57
  SPEAKER_02: You need to be the one that takes it to 100% where this problem is figured out, it's done, completed, and and I can tell Chad it's it we're we're done with that.
• 34:07
  SPEAKER_02: We can move on to the next one.
• 34:08
  SPEAKER_02: And unfortunately, what I'm seeing more and more of are people will come to you with ideas.
• 34:15
  SPEAKER_02: And ideas are great, and I can get ideas from Chat GPT, Gemini, Claude, but I need people that can think through the ideas, pick one, and then execute on that idea.
• 34:27
  SPEAKER_02: And unfortunately, I I think that we're losing some of that because we're just going to ChatGPT and say, okay, step by step, what should I do?
• 34:35
  SPEAKER_02: And then we're just putting our hands up and following it and hopefully it works out.
• 34:40
  SPEAKER_03: That's there is there's a real paradox around that.
• 34:42
  SPEAKER_03: The AI brain rot conversation, similar to when industry started making, you know, technology made physical labor less and people had to start going to the gym and, you know, eating different with AI, bankruptcy professionals, especially, if AI is going to help bankruptcy professionals handle the increase in filings.
• 35:02
  SPEAKER_03: However, those of us that have worked in the industry for many years became experts by doing thousands of cases.
• 35:08
  SPEAKER_03: If AI starts doing more of that foundational work, how will the next generation of bankruptcy professionals gain their expertise or understand the really vital nuance that you need to understand in bankruptcy matters?
• 35:21
  SPEAKER_02: It's a great question.
• 35:22
  SPEAKER_02: And and I still have uh when I have new attorneys that start and have no bankruptcy experience, I still have them filling out schedules to start.
• 35:30
  SPEAKER_02: Like that, that is, that is foundational to me.
• 35:33
  SPEAKER_02: If you don't understand the schedules, you don't understand bankruptcy.
• 35:36
  SPEAKER_02: And actually looking through the documents and understanding, you know, everything that you have to look through to create those schedules is is extremely important.
• 35:44
  SPEAKER_02: So I I as much as we want to take shortcuts, uh, I don't think it's good, especially for practitioners and professionals that really want to be the best at their craft, because I think the value of a top employee at any level is understanding that nuance and how to and and how to again solve any issue that comes up, because uh AI still is going to struggle with that.
• 36:10
  SPEAKER_03: And when you're doing a full AI lifecycle, for instance, when the intake is AI for you know what we talked about earlier.
• 36:16
  SPEAKER_03: One thing I remember from, you know, when I was really actively working cases daily is how important those early conversations are for picking up red flags and the nuance.
• 36:26
  SPEAKER_03: And so with AI taking over all of that, it can't understand nuance.
• 36:30
  SPEAKER_03: How are you and your firm handling that?
• 36:33
  SPEAKER_02: So we we still have obviously, you know, human in the chain, as they say, human humans have to be involved.
• 36:39
  SPEAKER_02: So we still have a multi-step process where where the clients are dealing with humans.
• 36:44
  SPEAKER_02: And that's for multiple reasons.
• 36:46
  SPEAKER_02: You know, everybody talks about the person that's you know hiding or concealing things.
• 36:50
  SPEAKER_02: That's such a very small percentage of people that are looking at bankruptcy.
• 36:54
  SPEAKER_02: Those people, generally speaking, stay out of bankruptcy and they hide in the shadows.
• 36:58
  SPEAKER_02: But truly, people forget about things, and especially when they're stressed during the first consultation.
• 37:03
  SPEAKER_02: So during the first consultation, you're you're gonna speak to an intake person that's gonna get information.
• 37:09
  SPEAKER_02: You're gonna speak to the attorney, then you're gonna speak to the intake person again, then you're gonna speak to a documents person, then you're gonna speak to a paralegal, then you're gonna speak to another attorney.
• 37:18
  SPEAKER_02: So there's so many touch points where we're asking similar questions differently that AI may not do that because it said, why would you ask this question again?
• 37:27
  SPEAKER_02: We already have the answer, right?
• 37:28
  SPEAKER_02: It's like, well, but if you ask it this way, you might get a different answer.
• 37:32
  SPEAKER_02: So, and and those are that's such important information to get, obviously, prior to filing.
• 37:39
  SPEAKER_02: You know, my my favorite, and we get this especially a lot in South Florida, is like, do you own any real estate?
• 37:44
  SPEAKER_02: No.
• 37:44
  SPEAKER_02: Do you own any real estate anywhere in the world?
• 37:46
  SPEAKER_02: Oh, yeah, my house in Columbia.
• 37:48
  SPEAKER_02: Uh yeah, I got so so you own real estate.
• 37:51
  SPEAKER_02: Well, yeah, but no, like I don't own real estate here.
• 37:54
  SPEAKER_02: I'm like, well, no, that's real estate.
• 37:56
  SPEAKER_02: So the answer is yes.
• 37:59
  SPEAKER_02: So that's just a little example of uh how you change the question and then all of a sudden you get a different answer.
• 38:05
  SPEAKER_01: You know, there's a perception.
• 38:07
  SPEAKER_01: This is this is a complicated question.
• 38:09
  SPEAKER_01: I'll try to not mangle it too badly.
• 38:12
  SPEAKER_01: There's a perception amongst a lot of the world that consumer bankruptcy does not tax the intellect of the lawyer or the law firm, and that it really is paralegal work and it really is form filling out type work, and the the edge cases don't get a lot of attention from the public or the world.
• 38:39
  SPEAKER_01: So here's the question.
• 38:41
  SPEAKER_01: The question is how can lawyers use AI to combat that perception that lawyers don't have to do any work.
• 38:51
  SPEAKER_01: And in fact, now AI is going to make it even easier to do bankruptcy.
• 38:55
  SPEAKER_01: How do you address that particular kind of like tension?
• 38:59
  SPEAKER_02: Well, initially, my first answer is it actually is extremely easy to file a bankruptcy.
• 39:04
  SPEAKER_02: Now, to successfully complete a bankruptcy is extremely different.
• 39:08
  SPEAKER_02: And obviously, you know, a goal normally in a chapter seven is a discharge, chapter 13, a reorganization where you want to get to a discharge.
• 39:16
  SPEAKER_02: So filing the basic paperwork to say that I'm in bankruptcy is easy.
• 39:22
  SPEAKER_02: And the the biggest problem with chapter seven is once you're in chapter seven, you can't just get out, right?
• 39:30
  SPEAKER_02: And we get cases all the time, and um, we're reluctant to get involved with them where it's pro se and uh you know, and ChatGPT or Claude or whatever has helped them with the paperwork.
• 39:42
  SPEAKER_02: And I spoke with the local clerk of courts, and he said the paperwork actually is more organized now.
• 39:49
  SPEAKER_02: The problem is, like I just said, oh, I didn't know my house in Colombia needed to be listed because that's not real estate.
• 39:56
  SPEAKER_02: But when you're in a seven, you can't just say, you know, I don't like this anymore.
• 40:00
  SPEAKER_02: I'm out, I don't want to be done with this.
• 40:02
  SPEAKER_02: No, they're selling that house in Colombia.
• 40:05
  SPEAKER_02: So if the stakes were extremely low, where you know your worst case scenario is I'm out a couple hundred bucks for a filing fee, it might, it might be worth the gamble.
• 40:16
  SPEAKER_02: But in a situation where your mom might be sued for$50,000 by a Chapter 7 trustee, that's the bigger issue, is it's easy to file the paperwork.
• 40:28
  SPEAKER_02: It's but dealing with the Chapter 7 trustee, avoiding those issues.
• 40:33
  SPEAKER_02: You know, we we just had a case come through where they really wanted to file, but they got an extension on their taxes, and the taxes weren't gonna be dischargeable for a few more weeks, and it would have been tens of thousands of dollars of difference.
• 40:47
  SPEAKER_02: Though that's the nuance that individuals are never gonna get, unfortunately.
• 40:52
  SPEAKER_02: And and you you really have to be a practitioner.
• 40:55
  SPEAKER_02: I I tell all my attorneys and even paralegals, I said it's very easy to get from like zero to 70 or 80 percent knowledge of bankruptcy if you're really practicing day in, day out.
• 41:06
  SPEAKER_02: But then it's that 0.1%, 1% that you'll pick up over the next few years until you get into the 90%, and then it even becomes slower.
• 41:14
  SPEAKER_02: It's a slower drip because you get it from experience and uh and you know, understanding what your judges expect, understanding what the trustees expect, understanding what the U.S.
• 41:22
  SPEAKER_02: trustees are going to look for.
• 41:24
  SPEAKER_02: I mean, there's so much to this, and and there's criminal referrals that that the US trustees will will make, the Chapter 7 trustees will make.
• 41:33
  SPEAKER_02: There's there's such a significant downside if you mess it up.
• 41:37
  SPEAKER_02: And I think that one of the things I think that AI may help is if we become extremely efficient at AI, hopefully it does reduce the cost and and uh it makes it more affordable for for people.
• 41:48
  SPEAKER_02: And I I hope that does occur.
• 41:50
  SPEAKER_02: I, you know, I love my I love bankruptcy practitioners.
• 41:53
  SPEAKER_02: I I want them to make money and they have a right to make money.
• 41:56
  SPEAKER_02: And I think that uh they they and we are undervalued significantly.
• 42:01
  SPEAKER_02: But in a scenario where the AI can reduce the time and increase the profit margins, I hope costs that end up coming down for the consumers.
• 42:09
  SPEAKER_03: That is amazing you asked that because it's in my list as well.
• 42:13
  SPEAKER_03: But it's a but it's more of a it's more of a statement and a philosophy I have.
• 42:17
  SPEAKER_03: I was gonna ask to say this to Chad.
• 42:19
  SPEAKER_03: My theory is about 70 to 80 percent of a bankruptcy workflow and process in case is process.
• 42:25
  SPEAKER_03: It is, you know, the organizing, the calculating, it's the document collection, it's the bane of every bankruptcy professional's existence.
• 42:32
  SPEAKER_03: I think 20 to 30 percent is legal judgment and the attorney, but in my opinion, that 20 to 30 percent is the most consequential piece of the entire process because it determines the case outcome and allows the consumer to become, you know, to have a successful discharge.
• 42:51
  SPEAKER_03: I mean, do you agree with that?
• 42:52
  SPEAKER_03: Or do you guys think it's a different percentage?
• 42:55
  SPEAKER_01: You're both I I told Chat GPT we're working on software that lawyers make their money in the 10%.
• 43:03
  SPEAKER_01: You think it's about 10%?
• 43:05
  SPEAKER_01: I that's that's what I think.
• 43:07
  SPEAKER_01: But you know what?
• 43:07
  SPEAKER_01: That's a lot.
• 43:08
  SPEAKER_01: That 10%.
• 43:09
  SPEAKER_01: Maybe it's 20%.
• 43:11
  SPEAKER_01: I mean, it's enough to kind of yeah.
• 43:14
  SPEAKER_02: Well, and I mean you have to you have to have your mind about Yeah, I I I think that uh for any legal proceeding uh you you should have representation, especially one that can be as significant as as this.
• 43:27
  SPEAKER_02: Just and it it's if if if the trustees weren't as aggressive as they are, and they have to be.
• 43:34
  SPEAKER_02: I mean, it's their job, right?
• 43:35
  SPEAKER_02: That's how they that's how they get paid.
• 43:36
  SPEAKER_02: The and they and they have people that they have bosses that they have to answer to, they're the ones that are that are really gonna keep keep consumer bankruptcy attorneys busy because you you're you hear those stories all the time.
• 43:49
  SPEAKER_02: Like I I see at 341 meetings, I there was one the other day my attorney told me about.
• 43:54
  SPEAKER_02: Uh the trustee said, Okay, when are you turning over your car?
• 43:57
  SPEAKER_02: They said, No, there's a hundred percent Vehicle exemption exemption and I need it for work because it's a toll of the trade or something.
• 44:05
  SPEAKER_02: So, well, no, there's a$5,000 vehicle exemption in Florida, and you have$20,000 worth of equity, that person is losing their car and may lose their job.
• 44:15
  SPEAKER_02: Like this is like ruining life scenario.
• 44:19
  SPEAKER_02: You know, that and I think that's the bigger issue is would you rather pay a couple thousand dollars to get it done right and really minimize uh the chance that something like that happens?
• 44:30
  SPEAKER_02: Because obviously it's never zero percent, but you're gonna minimize it significantly by paying you know a few thousand dollars, or would you rather roll the dice and say, no, I really don't have anything?
• 44:40
  SPEAKER_02: And then they they do a four-year look back and you're like, oh, you don't have anything now, but you did, and you gave stuff away, and now we're gonna sue some people for that.
• 44:49
  SPEAKER_01: Right.
• 44:50
  SPEAKER_01: That's a big problem.
• 44:51
  SPEAKER_01: The downside is so steep that most people don't see.
• 44:57
  SPEAKER_01: But this leads me into our practice signal segment, which is a time when we have a chance to hear from lawyers in the wild about problems they're having and how maybe AI can help with the problem.
• 45:12
  SPEAKER_01: Maybe it can't.
• 45:13
  SPEAKER_01: And they all come off Reddit.
• 45:15
  SPEAKER_01: In fact, you know, I had ChatGPT build uh a uh a tool where I push a button and I get the top five stories off Reddit's various legal uh subreddits to give me these uh these practice signals.
• 45:33
  SPEAKER_01: So today we were talking about solo practitioners.
• 45:36
  SPEAKER_01: This one deals with a solo practitioner.
• 45:40
  SPEAKER_01: A newly hired legal assistant at a solo firm recently described being asked to build the practices intake, calendaring, billing, and file management systems from scratch.
• 45:55
  SPEAKER_01: The kind of thing that I loved doing when I was in practice.
• 45:58
  SPEAKER_01: But this, like many other solos, they don't they don't want to spend their lives sitting in front of a screen.
• 46:06
  SPEAKER_01: And so he's giving her very little direction.
• 46:09
  SPEAKER_01: So the question is when a lawyer says I need systems, where should the lawyer's responsibility end and the staff's responsibility begin?
• 46:19
  SPEAKER_01: Chad, what do you think?
• 46:20
  SPEAKER_02: Well, the question is So I I might not have a popular opinion on this.
• 46:25
  SPEAKER_02: Just because I have the Esquire at the end of my name doesn't mean I'm smarter than anybody.
• 46:29
  SPEAKER_02: Like it doesn't mean I'm better at systems.
• 46:31
  SPEAKER_02: It might mean I'm worse at systems, actually, because my education isn't necessarily built for systems.
• 46:38
  SPEAKER_02: So it it really depends on the people that you're bringing on.
• 46:41
  SPEAKER_02: I I'm fortunate to have incredible people that are much better than me at building systems.
• 46:48
  SPEAKER_02: So it really depends on the lawyer skill set and the individual skill set.
• 46:53
  SPEAKER_02: Now, would I bring on a legal assistant with no experience in doing this and say figure it out?
• 46:58
  SPEAKER_02: Probably not.
• 47:00
  SPEAKER_02: Could they, if they really understood Claude, could they build something out?
• 47:03
  SPEAKER_02: Maybe, but it probably won't work for their specific firm because you really need input from that attorney of what they're gonna want and what they're gonna use, because systems are only as good as the people that are in the system.
• 47:17
  SPEAKER_02: So if you put a perfect system together with bad people, it's not gonna work.
• 47:22
  SPEAKER_02: So you really need input from the people that have the experience.
• 47:26
  SPEAKER_02: And I'm I'm all for outsourcing or hiring people that do things better than you.
• 47:31
  SPEAKER_01: So, Heather, what do you think?
• 47:33
  SPEAKER_01: You know, you you've got one of the members of your team on an attorney you've just onboarded, and your member sends you a Slack or something, saying, Help, the lawyer wants me to build out his tech stack.
• 47:46
  SPEAKER_01: Are you on the phone with that lawyer in five seconds?
• 47:49
  SPEAKER_01: Or what's your perspective?
• 47:52
  SPEAKER_03: I mean, my perspective would be you you could absolutely do that on our hourly basis, as long as the attorney is guiding you and supervising and you are getting the information from them that they want involved.
• 48:03
  SPEAKER_03: Yes.
• 48:04
  SPEAKER_03: I I don't I do think that hiring a brand new legal assistant with no legal background and asking them to do that would be really dumb.
• 48:14
  SPEAKER_01: I think it's a terrible, terrible, terrible idea.
• 48:19
  SPEAKER_01: First of all, you're the lawyer, you're gonna have your practice into your 60s, however long you're gonna have 10 years, 20 years, 30 years.
• 48:29
  SPEAKER_01: You don't know that this assistant is gonna be with you for the duration of your practice.
• 48:34
  SPEAKER_01: So you can't trust your systems to an employee that's not as deeply involved in the success of your practice as you are.
• 48:45
  SPEAKER_01: So it's just a terrible idea.
• 48:47
  SPEAKER_01: I totally agree with Chad.
• 48:49
  SPEAKER_01: You want to spend?
• 48:50
  SPEAKER_01: You know, look, call David AI, right?
• 48:53
  SPEAKER_01: David AI, I don't know, Chad, if you know this.
• 48:56
  SPEAKER_01: They're a uh we had them on the podcast.
• 49:00
  SPEAKER_01: They will come in and they'll do a build for your firm that's custom to your firm.
• 49:06
  SPEAKER_01: Call them.
• 49:07
  SPEAKER_01: It's worth the money.
• 49:08
  SPEAKER_01: I mean, I have a tech stack that you can trust.
• 49:11
  SPEAKER_03: But that's wouldn't that be similar?
• 49:13
  SPEAKER_03: Having David AI build it, someone else is still building it.
• 49:17
  SPEAKER_03: If you're saying the attorney needs to be the one in, and just you know, Ron, you are a different breed of attorney because you do like to build your own systems.
• 49:24
  SPEAKER_03: I start working with attorneys all the time who are in their 60s who have no idea how they got the systems they have and uh need us to help in untangle all of that because they have had past legal assistance and paralegals building this things.
• 49:37
  SPEAKER_01: You know, Heather, you bring up a really good point that we have discussed in the past, but maybe not as deeply as it merits, which is vendor reliability.
• 49:48
  SPEAKER_01: And that's a real big issue that's coming up in AI right now.
• 49:52
  SPEAKER_01: I think 2026 is gonna go down in history as the year that AI really changed, you know, especially in the legal field.
• 50:00
  SPEAKER_01: And you don't, you know, all sorts of vendors are coming forward.
• 50:04
  SPEAKER_01: I've got this great new tool, this great new AI tool, and who knows if they're gonna be in business in two years.
• 50:10
  SPEAKER_01: So that is that is a factor that is worth considering.
• 50:15
  SPEAKER_01: But I I I agree with Chad, you know, outsource it.
• 50:19
  SPEAKER_01: You know, if you don't want to sit, if the only time you want to sit in front of a screen, uh besides checking pace your dockets, is to play that new Cat Sudoku game that uh that they're showing on social media.
• 50:32
  SPEAKER_01: Well, okay, then hire somebody to do it for you.
• 50:37
  SPEAKER_03: I have got to look up the cat sudoku.
• 50:39
  SPEAKER_03: I have not caught that, but I am not on social media enough these days.
• 50:43
  SPEAKER_01: Well, and the cat sudoku will uh help us segue into our final segment.
• 50:48
  SPEAKER_01: Uh, you know, Chad, we've had two seasons.
• 50:52
  SPEAKER_01: The first season was where we really kind of dug down into ways that Flintstone's level lawyers and Simpson's level lawyers and Jetson's level lawyers could bring up AI.
• 51:03
  SPEAKER_01: Season two, we've modified that a little bit, where we were talking about how to build.
• 51:09
  SPEAKER_01: If you want to start building AI skills in your practice and you're of the different levels, how do you approach it?
• 51:16
  SPEAKER_01: And we we've gone through five different weeks of gradually getting to the point where we can finally bring AI in.
• 51:24
  SPEAKER_01: And so if you're a Flintstones lawyer, we've been dealing with engagement agreements where you try to have a document that you're going to bring AI magic into.
• 51:36
  SPEAKER_01: So if you're a Flintstones lawyer, upload your engagement agreement and its markdown companion into an AI project like Claude.
• 51:44
  SPEAKER_01: Then tell the AI, use these materials to help me prepare engagement agreements, identify missing information, don't invent facts and flag anything require attorney judgment.
• 51:56
  SPEAKER_01: That's the Flintstones level where we're talking about specific documents.
• 52:00
  SPEAKER_01: On the Simpsons level, we're talking about upload your client fact and tell the AI, answer client questions using my firm's explanation, not generic answers.
• 52:13
  SPEAKER_01: And then finally, on the Jetsons level, we were talking about building systems into your firm.
• 52:19
  SPEAKER_01: And if you're a Jetsons lawyer, upload your workflow and tell the AI, guide me through this process, identify the next step, and tell me what's missing.
• 52:31
  SPEAKER_01: So, and then you have to test the system, right?
• 52:34
  SPEAKER_01: Test it, give the AI a fictional client or a matter, see what it gets right, correct what it gets wrong, test it again.
• 52:42
  SPEAKER_01: And that's and that's the cycle.
• 52:44
  SPEAKER_01: Upload, instruct, test, improve.
• 52:48
  SPEAKER_01: And that's the payoff from the last six weeks.
• 52:51
  SPEAKER_01: You haven't just organized documents, you've built the raw materials for AI systems that reflect your law firm, not someone else's.
• 53:02
  SPEAKER_02: Chad, what do you think?
• 53:04
  SPEAKER_02: I I think it's wonderful.
• 53:05
  SPEAKER_02: And uh, you know, my my message to all every attorney I speak with, start with one.
• 53:10
  SPEAKER_02: Start with one skill, uh, whether it's helping you respond to emails, whether it's reviewing documents, retainers, just start with one and then and perfect that skill.
• 53:19
  SPEAKER_02: Get that skill working for you over and over again.
• 53:21
  SPEAKER_02: And start with the the the most annoying task that you have, the repetitive, annoying task that that unfortunately every every uh lawyer has to deal with.
• 53:31
  SPEAKER_02: Uh, we have to deal with administrative things that we'd rather not have to deal with.
• 53:35
  SPEAKER_02: But if you have if it's a repetitive task that you have to do over and over again, work up that skill and then you can start building that into systems.
• 53:43
  SPEAKER_02: So I think that's great advice that you that you've given.
• 53:46
  SPEAKER_02: Get involved now.
• 53:47
  SPEAKER_02: It's it's there, it's it's been there, and it's only getting better.
• 53:50
  SPEAKER_02: Like I said, 12 months from now, 18 months from now, we're not even gonna recognize where we are now.
• 53:55
  SPEAKER_02: And then if you're if you're just starting at that at that point, you're gonna be at such a disadvantage.
• 53:59
  SPEAKER_02: Um, but start whenever you can.
• 54:01
  SPEAKER_02: I would say start now, start small, and then you're gonna see extreme growth over the next 12 to 18.
• 54:07
  SPEAKER_03: What do you think, Heather?
• 54:09
  SPEAKER_03: I think that's amazing that you did that for everyone listening.
• 54:12
  SPEAKER_03: It was a great, great, you know, toolkit you gave them.
• 54:15
  SPEAKER_03: I think that going straight to Jetsons, Chad mentioned something early on in this episode that we're all gonna see is a huge need in our industry, and that is bankruptcy teams need to be able to handle two to three times the filings that they're currently handling.
• 54:30
  SPEAKER_03: We're not going to in the next 18 months or two years, when we're seeing these cases continue to rise over the next several years, we're not gonna all of a sudden produce thousands more experienced bankruptcy attorneys, professionals, court personnel, trustees.
• 54:44
  SPEAKER_03: So it's going to be necessary for people to understand and as he's Chad has talked about throughout this, really need to get on board because we need the AI to take away a lot of that repetitive work around the expertise and allow bankruptcy professionals to help more consumers.
• 55:02
  SPEAKER_01: That's so true.
• 55:03
  SPEAKER_01: I mean, we do need more people in the consumer bankruptcy field.
• 55:07
  SPEAKER_01: It's it's an opportunity, and I hope that law schools start uh understanding that.
• 55:13
  SPEAKER_01: Uh, you know, the whole idea of the Flintstones, Simpsons, Jetsons demarcation has come up in a manner of speaking in a Facebook chat that I've been having with somebody.
• 55:25
  SPEAKER_01: And this person, this lawyer that I've been chatting with, who is a delightful person, by the way, has expressed frustration about using AI.
• 55:34
  SPEAKER_01: And I said, you know what?
• 55:36
  SPEAKER_01: If I can get to 80%, I'm happy.
• 55:39
  SPEAKER_01: If I can go from zero to 80% using AI, uh, that's great because um now I only have the last 20% and I can manage that.
• 55:48
  SPEAKER_01: And this lawyer said that if if it's not 98%, the that lawyer is frustrated and and and loses interest.
• 55:57
  SPEAKER_01: And that represents the difference between the Jetsons lawyer and the Flintstones lawyer.
• 56:02
  SPEAKER_01: The Jetsons lawyer expects it to fail.
• 56:05
  SPEAKER_01: They expect the technology to fail until they work through it to get it up to where it's a manageable.
• 56:13
  SPEAKER_01: Is it gonna be 98%?
• 56:14
  SPEAKER_01: Sometimes.
• 56:16
  SPEAKER_01: Sometimes it will be, sometimes it'll be 100%, depending upon how detailed and specific the instructions are.
• 56:23
  SPEAKER_01: But if you're not willing to fight through the failure, you're gonna have a hard time progressing your systems and you'll you'll be stuck at the Flintstones level.
• 56:35
  SPEAKER_01: That's what I've seen.
• 56:37
  SPEAKER_02: No, I I I agree.
• 56:38
  SPEAKER_02: And to me, that's a crazy thought because I I think back to you know law school or when I first started practicing.
• 56:44
  SPEAKER_02: If I had to look at a blank Microsoft Word document and just start filling it out, just start, you know, like that that to me was overwhelming, right?
• 56:54
  SPEAKER_02: Just to say, okay, I'm gonna write a 15-page document from a blank white sheet, no template, right?
• 57:01
  SPEAKER_02: And and and now we're so far from that.
• 57:03
  SPEAKER_02: You know, we obviously that you got to the point where you could pull templates easily, easily off of you know, court drive, pacer, whatever.
• 57:10
  SPEAKER_02: You can get whatever pleading that you want.
• 57:11
  SPEAKER_02: So you can always start somewhere.
• 57:13
  SPEAKER_02: And and now you could get to the point where you have AI get you to 60, 70, 80, 90 percent.
• 57:20
  SPEAKER_02: Because even even the task of filling out the header, the heading correctly every time, that's just it's just it was frustrating to have to type it over and over again for every pleading.
• 57:29
  SPEAKER_02: So where we are now, if you if you can get me to 70, 80 percent, that's a great start.
• 57:36
  SPEAKER_02: And then especially if it's something I'm doing over and over again, I'm just gonna refine that quickly.
• 57:41
  SPEAKER_02: And quickly, we're gonna get to that 95 to 100% mark.
• 57:44
  SPEAKER_01: Chad Van Horn, this has been a great episode.
• 57:47
  SPEAKER_01: I very much appreciated your coming by and and sharing your experience.
• 57:52
  SPEAKER_01: Uh, it sounds like you have an exciting firm, and I look forward to seeing what more you do.
• 57:58
  SPEAKER_01: AI in consumer bankruptcy law.
• 58:01
  SPEAKER_02: It was so much fun being being on.
• 58:02
  SPEAKER_02: I hope to uh be back on again soon as as uh the AI world's ever changing.
• 58:09
  SPEAKER_01: Thank you.
• 58:09
  SPEAKER_01: All right, we'll see you all next time on AI Tools for Practicing Lawyers.
• 58:14
  SPEAKER_00: That's it for today's episode of AI Tools for Practicing Lawyers.
• 58:19
  SPEAKER_00: Thanks for listening.
• 58:20
  SPEAKER_00: We'll see you next time.