Why Legal AI Pioneer Pablo Arredondo Is Joining Clio to Build Technology for the Courts

0:02Welcome to Lawnext. This is Bob Ambrogi and today we have a bit of a spontaneous breaking news edition of Lawnext. The breaking news, Clio has just hired Pablo Arredondo as senior vice president to lead Clio's expansion into the judiciary and the courts. Pablo is well known as a former co-founder of Case Text, where he developed a number of innovative products, among them Cara, the very first brief analyzer, and then, of course, Co-Counsel, one of the very first generative AI legal assistants. Co-Counsel was so innovative that it caused Hopson Reuters to acquire Case Text for a whopping$650 million.

0:45dollars. Pablo went along with that deal, working at Thomson Reuters for a couple of years. If his name sounds familiar to listeners of this podcast, it's because just last February, I had a chance to visit Pablo at his home in Tiburon, just north of San Francisco in California, and talk about his work at Case Text and about Thomson Reuters. Well, ahead of today's announcement, I had a chance to sit down with Pablo again for a fairly impromptu conversation over the weekend. I happened to be in Nashville, Tennessee for ILTACON, and this was something that came up spur of the moment, so I grabbed the opportunity to sit down and talk to him.

1:25We talk about what this new role means for him, for Cleo, and for the courts. It's a really interesting conversation, so stay tuned. But first, please take this moment to hear from the sponsors who so generously support this podcast.

1:42This episode is brought to you by Briefpoint, the discovery governance and automation platform used by more than a thousand law firms. Briefpoint just launched Discovery Playbooks. Here's the idea. You respond to Discovery like you always do. Save it to your playbook in one click. And the next time a matching request shows up on any case, Briefpoint reapplies your objections and response strategy automatically. No data entry, no setup. Watch the 60-second demo at BriefPoint.ai. Capture lost revenue by accurately tracking time anytime, anyplace. Bill for Time helps your firm accurately track time with intuitive timers, automate your billing with customizable templates, and process payments with built-in online payment processing.

2:33Plus, with award-winning customer support, you'll be up and running in no time. Take your practice to the next level at Bill4Time.com. That's Bill, the number four, Time.com. Now on to today's conversation. Pablo, welcome back to Law Next. Great to be back, Bob. I interviewed you not all that long ago, back in February, you were on the show. And I had the great honor and pleasure of visiting you at your home in Tiburon, just north of San Francisco, right? And sitting down and having a wonderful conversation about your career and what you've been up to since leaving Thomson Reuters. And you are back because you have more news relative to all of that.

3:26So why don't you tell us the news rather than let me know? Right. I think that was six months ago. So that's, I think, two eons in current AI time dilation, right? Well, I'm excited to say that I'm going to be putting my shoulder towards working on technology for the judiciary. And I'm going to be doing it as part of Clio. I'm so excited. I've been having conversations with Jack Newton, the great founder of Clio, these last few months, and really thinking about where I want to focus. And I think Jack shared my vision, and Clio is so well-situated to be a place to build in this space. And so, yeah, that's the news.

4:10Yeah, it sounds really exciting. I mean, I know that when we did talk in February, you were talking about your possible, you were still kind of weighing what you wanted to be, kind of talking about your interest in working with the judiciary and you're already doing that in some ways. What exactly does this mean now? What's the assignment going to be with Cleo? Well, you know, it's still very early days. And I think, you know, I want to get the right people together to work on it with me. And, but, you know, rather than the specific assignment, we might look at like, what's the problem space, if you will, right?

4:44And so, you know, today we're Saturday all over the world, but let's just start with the United States. people are getting into car accidents or having land disputes or filing for family matters. All of the many, many things that law has to go and try to interdict with are happening. And on Monday, somewhere on the order of 3 billion tokens are going to hit court dockets, are going to be sent into the court dockets. My thinking about this is sort of what are the best things that we can do? In some ways, it's funny because 10 years ago, we were talking about CARA, right? where an attorney could upload a brief.

5:22And what are all the useful things that we could do when you upload a brief? And so in a way, this is sort of like, okay, well, now let's look across the whole genome of everything that's being uploaded, everything that's hitting the docket. And what are all the useful things that we can do? And what are the things that we can do in a way that not cause harm, whether intended or unintended? Because I think that with the judiciary, you have to focus on that. And if you'll forgive me, as is my want, I have a biological metaphor here, which is if you think of a nuclear envelope, right? In every one of your cells, you have your critical nucleus, but it's surrounded by this layer that's basically there to help, you know, referee what comes in and out, process what comes in and out, help, you know, sort of aid in that process.

6:03And so, you know, to me, that's really what I think the immediate thing to think about is how do you best handle this just absolute stream that's coming in? And a stream, I say stream, maybe turrent is a better word. and a torrent that is, thanks to the power of AI, getting bigger, right? I think there was a study out of Princeton showing a 15 % increase in pro-safe filings. As with every study, go check it yourself on this. But I believe at least double-digit increase in pro-safe filings. And then among people that do have representation, in increasing filings as well. So we're just seeing more and more that's flooding into the courts.

6:44And I think there's no questioning I can help. The question is, how best can it help? And how do we design systems for that rather awesome responsibility? Yeah. In the last time you were on the show, we kind of talked about your career and you obviously had been a co-founder at Case Text. You helped develop co-counsel, Kara, as you talked about other products there, then went to Thomson Reuters. A lot of that time was focused on developing technology or developing AI products, focused on legal practitioners. How did your interest kind of shift toward, I'm sure it hasn't shifted away entirely from the legal practitioner side, but how did you start looking at the judiciary side of all this?

7:28Right. Well, there's definitely been obviously some overlap, but it's sort of interesting. So I think the actual, the first time the judiciary interacted with language models, not necessarily large language models yet, because I think if we look at BERT, right, which was that first transformer-based neural net, one of the very first ones, we at K-Stacks had created something called Parallel Search. And to me, it was very seismic. I think actually you have a blog post about four ILTAs ago, you know, talking about how you could use these language models to transcend keyword search. And I remember noticing organically, oh, my goodness, courts are adopting this, right?

8:07Like we were just seeing them using it. And that's something, you know, we weren't marketing there at all, et cetera. So in some ways, there have been like intro signals, right, that the work we were doing really with a focus on litigators, as you say, was spilling over, if you will, into courts. And the courts were finding the same functionality there. But increasingly, I think part of it, to be honest, is just intellectually the challenge with courts. It's got all of the wonderful meaty challenge of design and product challenges that you get building for lawyers. and then a lot of societal challenge, a lot of sort of policy questions, right?

8:44And a lot of questions like that. And that is something that in my old age, I guess, if you will, I'm starting to see, you know, I think with the new advent. You've got a long way to go. Well, I think with this new wave of AI, it's getting so much more powerful. I think increasingly that stuff is coming to the fore. Yeah. And the last few months, I had the actual privilege of working at Cornell's Legal Information Institute on a project where we basically, Harris Tillowicz and I, the great Harris Tillowicz of Scattered ARPS fame, we were taking the code of conduct for judges, the U.S. code of conduct for judges.

9:23and then saying, can we use LLMs to help facilitate what you might call comparative law, annotating that code with historical wisdom, if you will, across the continents and the generations. So, you know, everything from rabbinic discourse, right, which was one of our main focuses, and from Islamic and Sharia tradition, right, from Roman Catholic tradition. And you can't spend too much time doing that without becoming just like profoundly interested in just how important the role is. And it's almost sort of cliche to say, like, of course, judges are very important in the legal system. But the sophistication of the debate around the judiciary and just how far back that goes and how deeply that's been is something I didn't fully appreciate, I think, until this Cornell project.

10:13And so, you know, please go ahead. Well, I was going to say, what does that project tell you

10:22about judicial conduct, I guess. Right. Well, that's a lot. And Harris and I are going to be releasing publicly the sort of output of this. We're very privileged to be working with experts in the field who are going to help us sort of write like a formal intro for it. But I mean, it's a different thing. It's a new one. So some of it is things that we already think about, but maybe there are deeper ways to think about it. So like an example would be verbal bribery. So in rabbinical discourse, if somebody had flattered a judge, that could be considered something that should disqualify the judge from being able to rule on it, right?

10:58Not money per se, I think the kind of things we're used to thinking about, but just the sort of the fact that very kind words or sort of social standing there can be used to do that, right? So I think often it's just a matter of maybe more robust examination or more strict scrutiny over the various things that can interfere with it. And then there are things, I mean, and again, you want to keep grounded in this world, right? But there's a quote, I think, from Rashi where he says, somebody who judges for even an hour, it is as though they have become partners with the creator, right? You say, oh, boy, all right, well, that's worthy of a user interface.

11:38You really want to get that UI right if somebody's doing that. So I'm probably doing a poor job explaining many of the many things. Some of it also, remember that study that showed that if judges, depending on what time of the day they were doing the sentencing, if they were hungry or sleepy or sort of circadian rhythm type stuff? Well, that was something in Sharia. like centuries ago, they were talking about how like you shouldn't go in there if you haven't eaten the least X or Y. And, you know, a lot of things, you know, again, litigants should be equal in the eyes. You shouldn't have one wearing rich clothes and one wearing poor clothes, right?

12:16And things like that. And so then you say, okay, well, what does that mean in terms of if the filing is one filing is very polished, beautiful font and spacing and everything like that, and one is more messy and sprawled and things like that. So for like sort of each of these, you can almost think about like, well, what would this mean for how we design the technology that underpins it, right? Now, not all of them are one-to-one, but to me, I think that that's the right place to sort of start, is that the judiciary long precedes AI, long precedes computer science, long precedes information It goes way, way back.

12:52And a lot of those tensions are still here, are still just as real. And so how we design things needs to sort of not ignore those or sort of wish those away, but rather embrace them head on. Yeah. Does that also inform, I mean, this may sound like a stupid thing to say to a lot of lawyers listening to this right now, but I mean, why is it that you can't just pick up, say, Clio Manage and drop it over in the judiciary and say, get to work? And why do they need unique technology? So this is a whole thing with me, Bob, right? It's like, why do I need to tailor this dress just for Beyonce? Why don't I just make a dress that everybody can wear, right?

13:31Yeah. It's not that you can't get somewhere with that, right? But it would be, I mean, why would it not be better if you think about the ways that you can tailor it? And part of it is removing things, right? You don't want the judge to have to navigate through something that helps with, you know, depot prep, much less mergers and acquisitions, my God, right? Sort of like staring it in the face, right? So I think that there's fiduciary grade you hear talked about. Well, I think judiciary grade is its own system. And it's not that, I think lawyers just as much as judges care about security and care about accuracy.

14:08It's not sort of different in that regard. But I think creating the best environment and the best functionality, it's just a different set of things you're going after when you're doing it for judges. Yeah. I think we might have talked about this when I was out there in February, but I have this sort of Forrest Gump memory of being at a conference with you at Harvard like three years ago or two years ago about AI. And there are a bunch of judges there from the Massachusetts judiciary who were all saying, basically had their jaws dropping as they were listening to people talk about generative AI.

14:45And sort of at the end of the day, they're like, how do we get ourselves some of this stuff? It was basically, you know, to paraphrase. And I kind of recall you jumping up out of your seat and running over to talk to them. Was that the genesis of some of what we're doing right now? The life of a first-year associate at Kirkland or other big firms is like, I wonder if I'll ever get to see a judge, because these are high-stake litigations. They're going to put the more senior folks there. And then, yeah, they invited me to lunch at the conference. And I was like, would you like to join us? And I was like, would I?

15:22And I think that you sort of have these two parallel things happening. The hallucinations are flooding in. Right. And that's creating one narrative about just sort of like, oh, it's the bane of these. Right. But I think that much I think a much deeper thing that's happening is judges are starting to think, how can I use this technology when we are? There are many courts that I think, frankly, are overloaded. I think there is more matters coming in than they have the resources to fully work with, at least if they want to try to do it efficiently. Right. And then very quickly, you know, quickly is not the right word, but efficiently, let's say.

15:57At the same time, there's just, it's a huge challenge to administer justice. It's like, it's the, in some way sort of the final boss of law is how do you then come to the right decision once everything's put in front of you? So I think that what we're going to see increasingly is that courts are adopting AI in various forms. And the question then is, how do we do it right? And that isn't just how do we make sure it works as intended, but also how do we make sure that when it's working as intended, we're not causing problems that we haven't thought about, right, that are sort of these unforeseen issues.

16:35Yeah.

16:46I'm speaking with Pablo Arredondo about his new role as Senior Vice President of Judiciary, where he's going to lead Clio's expansion into the courts. Pablo was kind enough to sit down with me over the weekend ahead of this announcement coming out. We've got a lot more to talk to. We're going to get back to it in just a moment. But first, please take a moment to hear about the sponsors who make this podcast possible.

17:13Law firms lose thousands of billable hours each year because the work they do never gets captured. Ajax uses AI to turn activity across screens, email calendars, and documents into ready-to-bill entries tailored to your firm's billing rules and outside counsel guidelines. Entries sync directly to your billing system, helping firms capture over 12 % more billable time. 97 % of firms that pilot Ajax choose to keep it. Learn more at joinajax.com. Streamline your time tracking and simplify firm operations with Bill4Time, the most trusted name in legal time tracking and billing software. With Bill for Time's intuitive time tracker, standardized invoicing, customizable templates, and built-in online payment processing, you could be getting paid up to 70 % faster.

18:10Visit BillForTime.com today to learn how to take your practice to the next level. That's Bill, the number four, Time.com.

18:22Welcome back to Law Next. I am talking to Pablo Arredondo, formerly the co-founder of Case Text, who helped develop its co-counsel, generative AI, legal assistance that was later acquired by Thomson Reuters. He has now just taken on a new role as senior vice president of judiciary at Clio. And ahead of the announcement of that today, he and I had a chance to sit down over the weekend and talk about that new role. Let's get back to that conversation.

18:56We're talking before this is being announced, and I know you don't start till officially start until August 31st, although this is all started in your mind long ago, I think. But the press release says you're going to lead Clio's strategy behind its move into the courts. So, I mean, is what we're talking about here just kind of trying to build a practice management system for judges? Or are we really talking about looking at, not we, are you really talking about looking at all the ways AI can impact and improve the administration of justice through the judiciary? Definitely the latter. Definitely the latter.

19:34I think that, you know, there are existing practice management systems that are set up. I think that the question is just there's now this whole new slew of functionality that's enabled with LLMs. Now, to be clear, sometimes there's useful things to do that don't need LLMs that just nobody seems to have gotten around to. So I'm not precluding those by any means. But I think that it's much more thinking holistically sort of zero to one on doing this. it's funny so um there's a great vc firm that after you uh sell a company they bring you down to be like come on go start another one right it's this beautiful silicon valley uh tradition i guess and they talk about how a mistake that you can make is some founders decide to go completely out of their wheelhouse like there's something so different that it's like completely you know nothing like what they know about right and then the other mistake is to sort of go try to just do the same thing again, right?

20:30And kind of do that. And I'd like to think that with the judiciary here, I'm sort of listening to their guidance on that a bit, which is to say that there is definitely an overlap with everything I've been working on for the last 17 years and 25, I guess, if you count law school. But I mean, you know, law school on, but there's definitely unique things like pro se filings, right? I never would think, I used to pride myself on how no non-lawyers allowed to touch any stuff that I built because I don't think that they're, you know, can handle it. As soon as you take the court as who you're working for, that means that's no longer a luxury you have, right?

21:08And so then the question is, how do you handle pro se filings? And now what are the way to do it? So that's one example, I think. Yeah. Does that mean that it then begins to extend into self-represented litigants as a sort of a target audience of the work you'll be doing, as well as the judges and clerks and other people who make up the judicial system? Well, I would say this, that I think that the judge, the court has that coming in and has that process. And there's already some sort of guided filing systems you've seen, some sort of rudimentary ones that are supposed to sort of help that process.

21:42I would consider that surface to absolutely be one that you have to think about. And then to the extent you are, in a sense, yes, the user becomes the per se filer. Yeah. You introduced me to Shlomo Clapper, alerted hand, who's kind of built out a technology platform for judges, an AI-based technology platform for judges. Are you now competing with him, working with him? What do you see is happening here? You know, speaking of flattery causing bias, right? Shlomo has said some kind things about me. I'm a huge Shlomo Clapper fan. He was on this podcast a while ago, too. You're urging. I interviewed him, yeah.

22:20Oh, right, right, right. Yeah. No, I'm a huge Shlomo Clapper fan. But I'll tell you what I told him, which is I don't necessarily think of the judiciary as a startup play per se, right? You can't really growth hack it. There are a lot of reasons not to mention a small thing called having access to the case law and it's at the level you need it to have. Are Shlomo and I focused right now in some very similar space? Absolutely. You know me, Bob. I transcend your competition. No, no, we're brother in arms. I'd say he's a brother in arms. And I've been very impressed with what I've seen him do. Yeah.

22:56But so it sounds like what you're saying is that in order to tackle this issue, you need a much, much greater set of resources and capabilities. And, you know, you talked about case law. Obviously, Clio now has not just a lot of money, but a whole lot of case law and a whole lot of other data at its disposal and probably a whole lot more developers and other stuff. So is that, I mean, how does that, how do you see that carrying over into what you want to do? It's so great. So like Cleo Docket, right? I mean, I remember Cleo Docket back in the day, right, was Michael Sanders' Docket Alarm. Docket Alarm.

23:40And I used to joke with him. I'm like, you know, Michael, you're one of the few people that I sometimes think about just committing a crime and taking all of your data from you somehow. Like, you know, just somehow through like, you know, brute force or threat of violence. Just that's how jealous I am of having that. and, ah, well now here it is, right? So I think the docket data is just organically so what you would want, right? Is to have the ability to go and experiment with and test on the real world flow. And this, you know, this is a time where you see efforts at synthetic data, right? And I think I was talking to Neil Guha about this to some extent, right?

24:19About, you know, what you can do with synthetic data, which is nice is find big problems, right? Like if it doesn't work with the synthetic data, the chances of it working with the real world data. And then by the way, these are my views without Neil's, you know, like talk to Neil directly on anything on this. I'm not trying to put him on record of anything like that. But the thing is though, is the real world can also be very noisy and messy, right? And that's only when you go see that real flow come in, do you realize just how varied and how many crazy type weird mistakes and then just sort of all sorts of peculiarities you have.

24:52So while I applaud any and all efforts to create synthetic data, I also hope that, you know, I think there's a big place for putting your shoulder towards working with the real stuff. Yeah. So explain that. What does having access to all that data enable you to do in terms of working with the judiciary? Well, I mean, this is, you know, we talked about the nuclear envelope, right? And sort of everything that comes in, those 3 billion tokens a day or whatever you want to call it, right? Well, this is that. This is that very fabric, right? And so, you know, there's some things I'll keep for a little bit till later on this, you know, not to be cagey, but I think any legal informaticist worth their salt would say, well, you know, there's many, many things you could start to do with it.

25:36And then, of course, to be able to test against real data, right, and say, okay, does this really work across what is actually happening in the real world? Yeah. But I mean, as part of that, just the ability to analyze how things flow through the courts, what kinds of things are flowing at a greater level, even just things like how many matters really do involve self-represented litigants versus represented parties. I mean, just all of that kind of stuff. exactly right i mean exactly it's it's the it's the very i mean right like it's it's sort of like you work in uh traffic control and say well what does it mean to be able to have all the ways data or something like that it's like oh my god this is the very essence of what you're doing no it's not the end of things of course right because then you have to go into the case law and do a bunch of things there right and then ultimately you have to think about the actual process by which the judge drafts right which i think is a later stage thing but clio is just this fantastic combination of a culture that I've loved and long admired and some of the fundamental content that you'd want to work with the stuff, to work on the stuff.

26:39The path question with regard to AI and lawyers has come to be the extent to which AI could replace lawyers. What about the judiciary side of that? To what extent can AI begin to perform more of the functions that have required humans open to now. All right. Well, so first I want to refer everyone to a much brighter mind than mine on this. Terence Tao, the great mathematician, has a relatively new lecture out called Mathematics in the Age of AI. And there are obviously some key differences between law and mathematics. And so I'm not saying that everything that applies, but the way that he's so crisp, not surprisingly for somebody who's a learning mathematician, right?

27:26He's so crisp in how he formulates what the problem, you know, kind of how to think about the problem. And essentially, the thing to think about is that it's not all replaced or does nothing. It's completely useless because, right, it's what we're dealing with is sort of the spectrum in between. How many of the things can be automated, but with how much oversight, right? And then how much time and things like that and at what cost. So, but short answer is no, I don't think in the being replaced with AI is anything I would recommend. You know, there's alternative dispute resolution type questions, right?

28:01And if the party is satisfied. But again, I'm just starting to sink my teeth into it a bit, right? But I think that the psychology of the litigants also matters, right? If you don't feel hurt, is that even if the outcome is the same, it's very interesting to follow on that. But short answer is absolutely not. I don't see this fully replacing judges. Yeah.

28:51in some cases. Then there's the clerks who have to sort of keep the wheels turning and keep everything running and who, you know, I know over the last decade in particular, and as you say, even more so now since the advent of AI, are just, again, overwhelmed. It's gotten worse because of self-represented litigants, not just because of the quantity of cases, but because the litigants don't understand how the system works and they have a lot of questions and all of that. And then there's the litigants themselves, whether lawyers or the self-represented litigants. It seems to me the self-represented litigants are sort of the bigger part of the problem, if you will, in that they are the ones who don't understand the system, may not know how to file something, or once they do figure it out, now they're all using JGPT to file pleadings and file many more pleadings.

29:43So, I mean, you can tell me if that's a way over a simplification. No, no, no. Those are the core, but I want to make it, I'll add one too, which is future litigants and the society that is governed by a body of judicial opinions, right? Which is a little bit more nebulous, but also a big constituent in this, right? That you sort of, these things that are created. again early days I think a lot of it depends sometimes on trial court versus appellate I was a patent litigator so I was used to stately federal court with its relatively like space diary and all these things and I'm quickly learning that state courts can be very different right in terms of what's what the process is what's available how much time you know we've judges that say what clerk I don't have a clerk I wish I had a clerk or like I share one with So I think that the judiciary is not a monolith in terms of what is available to it.

30:42I mean, sort of what their process is. And so I would even say that you would hyper-tailor, not just I want tailored for the judiciary, but I think that what would be built for an appellate court might differ from a trial court, which might differ from even the magistrate judge who's tasked with discovery. Yeah. I quoted Richard Susskind a million times, his quote out of one of his books, or maybe all of his books, about the fact that one of the problems with modernizing judiciary is that you can't change the tires on a moving vehicle and the judiciary never stops moving and it gets really hard to innovate or kind of work on redeveloping systems.

31:24do you think that's changing is that something that you see as an obstacle or a speed bump that you're going to have to face going forward working with them yeah i mean it's it's definitely real i don't think it's changing i think another problem i remember way back in you know maybe 12 years ago do you remember um the mitre folks had some project with the a right and the ao i think you know with the administrative office and i remember like running to them being like great we're going to solve everything like here here are my 10 ideas i'm sure you got 10 let's just implement you know everything.

31:52And they quickly disabused me of the idea of just the snap your fingers and the judges all do things. And in fact, each judge is to some extent a fiefdom, right? They said, right? Each judge has their own system. I mean, look at like standing orders, right? Each judge, I saw a standing order recently, the judge said, I don't want you to cite sell a text at me, right? So, you know, everyone knows you cite to sell a text, the basic summary judgment standard, because that's just the form you do it. He decided enough. I don't want to hear about it. So, you know, each court has their own, you know, chambers.

32:25And I don't want to, that's the world we have. I think there's arguments, but that's not the, you know, the end of the world. And so, yes, I think that the fact is that there's ongoing matters is one of the challenges. The fact that courts are sometimes a lesser resource than you might think, given, you know, how important what they work on is a challenge. challenge there's a bunch of challenges bob but yeah uh that that doesn't make it that doesn't mean we don't need to do it no not at all no i'm just i'm just i'm just i'm just curious because i mean just i think another one has just been judicial attitudes around ai i mean we're seeing some judges who are being uh ahead of the pack i think i think uh you know xavier rodriguez out of texas is one of the judges that talks about the fact that he'll use ai in drafting his opinions not his decision part, but pulling together the facts and all that.

33:14And then there are others who are just, you know, chock, chock, chock at that. And you can't imagine ever going down that road. So, I mean, I don't know. I don't know if there's a question there so much as are you at all concerned that judicial attitudes around AI are something you're going to have to work around too? I think there's an opportunity. I sometimes talk about, I've been mulling a lot, and again, more on this a bit later in school, But AI for people who hate AI, right? What does that look like, right? And I actually think that the answer isn't no AI per se. I think there are things you can do that are designed in a world where there's zero trust of it and sort of zero wanting to depend on it and paranoia about cognitive decline and all of these things.

33:58So we'll see. But I don't count those folks out that are sort of saying no AI. I think what they don't want is, you know, over-reliance on AI, or putting into a chat bot's hand that which should not be put into a chat bot's hand. obviously nobody wants hallucinations by God, right? So I think you're right. There's a wide spectrum, but I don't think, I think if we're doing it well, there'll be those sort of begrudging, like, well, all right, well, that thing's okay. Fine, let that happen, but nothing, right? And then I want to build that too. Yeah. So I realized, as I say, we're chatting before the release has even gone out and you don't even start till the end of the month.

34:34August 31st, first day on the job. Where do you start? You know, you just kick into that wonderful, This is why I love going zero to one, right? And it's the, you know, this is, I'm Silicon Valley, right? You do two things. You talk to users and you build and you iterate, right? Those are the two things that basically are it. And so I hope to be doing as much of that as possible. The only thing I would add is that there are some folks that have been doing great work in this space already. You know, I can start naming them and then they'll be like, oh boy, Paul is going to come bother me. But then there'll be all the ones you forget.

35:10Yeah, yeah. David Engstrom at Stanford has been doing great stuff about how do you handle default judgments and things like that. This is a massive problem. It is far more than any one person should work on, even as someone as full of enthusiasm about it as yours truly. And so I hope to connect with those folks as well and get going, because again, these tokens are going to hit on Monday. Yeah. And will you be I thought I saw a reference in the in the draft press release here to building a team? Will you be one person or will you be building a team? Oh, you need a team. You need a team. Right.

35:47And so that is definitely part of this part of this journey. Yeah. All right. Well, on short notice here, I appreciate your hopping on a call and talking a little bit about this. Is there anything else that you really wanted to talk about that we haven't? about talking about it. Well, you asked last time when you were here, you said, what's your priority? And I said, oh, I was going to lose the weight. And I want to let you know that I did. And then I found it again. I think somewhere in Maui, I found it again. So anyway, that'll be ongoing. Not really the main priority. But in case you were wondering about that.

36:19I was on the edge of my seat wondering. No, no, nothing more to say. I'm just tremendously excited. And mostly I want to have, I hope we can have future conversations as we really get going on this and have other stuff to show you. But I'm thrilled to be doing this, and I'm very grateful to Jack for building something as awesome as Clio and for taking the chance on sort of the vision that I have on this. Yeah, well, you've got a lot of great people to tap into. I mean, Clio's pretty good, a great team on a lot of levels. I'm here at Iltacon. I'm going to actually see Jack Tuesday. I think somebody told me it's his first time coming to Iltacon.

36:58I don't know if that's right. But I'm going to see him. So I'll be interested to hear what he has to say about all of this. Excellent. Excellent. All right. Well, I'll let you get on with your Saturday and appreciate your sitting down with me. All right, Bob. Enjoy Nashville. And yeah, send my best to everyone and look forward to future convos. Congratulations.

37:21That's it for today's episode. I hope you enjoyed that conversation with Pablo. If you're a fan of our show, please leave us a review. If you have any comments, just go ahead and email me, ambroji at gmail.com or comment on LinkedIn or other social media. Lawnext is a production of Lawnext Media Incorporated. I hope you'll join us again next time for another episode of Lawnext.