1 00:00:02,636 --> 00:00:04,499 Alex Su, how are you this morning? 2 00:00:04,808 --> 00:00:05,536 I'm doing great. 3 00:00:05,536 --> 00:00:07,011 Thanks for having me on the show. 4 00:00:07,011 --> 00:00:09,393 Yeah, I really appreciate you joining, man. 5 00:00:09,393 --> 00:00:14,216 I read some of your stuff and you've got great videos on LinkedIn. 6 00:00:14,216 --> 00:00:18,548 I've been a fan, so it's awesome to have you on. 7 00:00:18,890 --> 00:00:24,497 for Thanks for following along and yeah, I post a lot of crazy stuff, occasionally serious stuff. 8 00:00:24,497 --> 00:00:26,988 So love to talk about it more. 9 00:00:26,988 --> 00:00:28,319 Yeah, well good stuff. 10 00:00:28,319 --> 00:00:32,694 So you started off as an attorney yourself, right? 11 00:00:32,694 --> 00:00:36,959 I think at Sullivan and Cromwell, you were in eDiscovery for a while. 12 00:00:36,959 --> 00:00:40,232 You're now a CRO at Latitude Legal. 13 00:00:40,232 --> 00:00:42,675 Why don't you give us just expand on that a little bit. 14 00:00:42,675 --> 00:00:45,928 Tell us who you are, what you do and where you do it. 15 00:00:46,652 --> 00:00:47,262 Absolutely. 16 00:00:47,262 --> 00:00:50,864 So I am the chief revenue officer of Latitude Legal. 17 00:00:50,864 --> 00:00:56,086 We provide contract attorneys to law firms and legal departments. 18 00:00:56,446 --> 00:01:00,208 I've spent my entire career focused on legal innovation. 19 00:01:00,208 --> 00:01:05,570 I practiced law for six years, clerked for a federal judge, worked at Sullivan Cromwell. 20 00:01:05,590 --> 00:01:08,451 And then about six years out, I decided to make a change. 21 00:01:08,451 --> 00:01:12,867 I found myself in California and saw that the tech startup scene was really bustling. 22 00:01:12,867 --> 00:01:14,590 so 23 00:01:14,590 --> 00:01:23,077 You know, for a lot of reasons, I took a pivot, became a salesperson for selling e-discovery software, then moved to the contract management space, doing sales, then 24 00:01:23,077 --> 00:01:24,117 marketing. 25 00:01:24,158 --> 00:01:30,673 And so earlier this year, I found myself at Latitude, now leading sales and marketing as the chief revenue officer. 26 00:01:30,673 --> 00:01:38,550 Uh, and it's been super interesting because, you know, through my 14 or so years of experience, I've seen the practice of law evolve. 27 00:01:38,550 --> 00:01:43,914 I've seen software services, alternative legal services providers, all sorts of things. 28 00:01:44,075 --> 00:01:47,568 um, make their way into our profession. 29 00:01:47,568 --> 00:01:51,272 And so, um, yeah, it's been, it's been great to see. 30 00:01:51,272 --> 00:01:55,386 And so I write about some of these topics in my newsletter off the record. 31 00:01:55,386 --> 00:02:03,664 Um, you know, very often also writing about AI, which I think is how, you know, you and I got connected and, it's some of the stuff that we're going to talk about today. 32 00:02:03,682 --> 00:02:04,502 Yeah. 33 00:02:04,502 --> 00:02:04,832 Yeah. 34 00:02:04,832 --> 00:02:09,854 Your, uh, your off the record newsletter was actually what grabbed my attention. 35 00:02:09,854 --> 00:02:19,806 You had a, you had an article, God, I guess it was maybe a month ago about advice for, for legal tech startups or legal AI startups. 36 00:02:19,806 --> 00:02:22,567 And I thought there was some really good nuggets in there. 37 00:02:22,567 --> 00:02:32,610 And I think, I think we're fairly aligned in our perspectives on how difficult it is to disrupt the legal services market. 38 00:02:32,610 --> 00:02:33,370 Um, 39 00:02:33,710 --> 00:02:41,270 You know, we, we started off as a company called Acrowire back in 2008, just fell into the legal rabbit hole. 40 00:02:41,690 --> 00:02:51,650 And you know, when cloud really started to take off in the enterprise, it seemed like an absolute no brainer. 41 00:02:51,650 --> 00:02:54,570 Law firms were, were, were going to head that direction. 42 00:02:54,570 --> 00:03:03,974 I'm going back to like 2010, 2011, and I grossly overestimated how quickly that 43 00:03:04,206 --> 00:03:09,826 migration would take and wasted a lot of time and energy. 44 00:03:09,826 --> 00:03:23,686 I did 22, 24 road shows in 2014 out there really highlighting the advantages of Microsoft 365 and SharePoint Online. 45 00:03:23,686 --> 00:03:32,224 And it would be a solid eight, nine plus years before big law would transition in earnest. 46 00:03:32,270 --> 00:03:33,831 to Microsoft 365. 47 00:03:33,831 --> 00:03:35,053 It's finally happening now. 48 00:03:35,053 --> 00:03:36,854 AmLaw is moving there now. 49 00:03:36,854 --> 00:03:51,578 There's only a handful of firms that are still on-prem in that suite, but man, it's, know, current state thinking and, you know, current models are pretty well entrenched. 50 00:03:51,578 --> 00:03:56,923 It's not easy to move in a new direction in legal tech, it seems. 51 00:03:56,923 --> 00:03:58,654 Is that your perspective as well? 52 00:03:59,146 --> 00:04:01,166 Yeah, no, it's not easy at all. 53 00:04:01,166 --> 00:04:03,566 My own journey began in 2016. 54 00:04:03,566 --> 00:04:10,446 So a few years after that, we were selling e-discovery software that was cloud-based to law firms. 55 00:04:10,446 --> 00:04:13,776 And the firms would say, they would have all sorts of objections. 56 00:04:13,776 --> 00:04:16,106 And I remember it was what, 2016. 57 00:04:16,106 --> 00:04:21,646 Five years before that, I had started using Dropbox for file sharing. 58 00:04:21,646 --> 00:04:24,076 And of course, firms are different, right? 59 00:04:24,076 --> 00:04:26,666 There's highly sensitive documents they've got to deal with. 60 00:04:26,666 --> 00:04:28,694 So it's not the same thing, but... 61 00:04:28,694 --> 00:04:31,296 to your point, there's a bit of a lag. 62 00:04:31,296 --> 00:04:40,182 And so I think it's extremely topical these days because everyone keeps talking about AI, the transformation it's having on other industries. 63 00:04:40,442 --> 00:04:48,898 so naturally the conversation has shifted now to, lawyers do things in a very antiquated way. 64 00:04:48,898 --> 00:04:51,660 Maybe AI can completely transform that. 65 00:04:51,660 --> 00:04:55,443 And that's why you have articles coming out that say, will AI replace lawyers? 66 00:04:55,443 --> 00:04:57,714 Like you see them every few months. 67 00:04:57,736 --> 00:05:05,990 So part of what I was trying to write the article, the reason for writing the article is to share kind of with the broader world that legal is different. 68 00:05:06,171 --> 00:05:18,017 And specifically, our ecosystem is built very differently than other industries and specifically firms, law firms, especially the largest ones, occupy a very unique space 69 00:05:18,117 --> 00:05:26,181 that has a lot of nuance that makes it both challenging to disrupt, but also creates 70 00:05:26,684 --> 00:05:27,905 unexpected opportunities. 71 00:05:27,905 --> 00:05:39,904 And so I was trying to share a little bit about that because I just started seeing a lot of tweets, honestly, a lot of posts on social media about how AI is going to just replace 72 00:05:39,904 --> 00:05:40,335 law firms. 73 00:05:40,335 --> 00:05:42,616 And I thought that was really misinformed. 74 00:05:42,680 --> 00:05:44,691 Yeah, it is. 75 00:05:45,052 --> 00:06:00,425 you see, especially lately, Harvey is an example where you've got a lot of investors like Google Ventures led their series C round that really don't have a lot of experience in 76 00:06:00,425 --> 00:06:04,629 legal. 77 00:06:04,629 --> 00:06:10,518 I think there's probably, especially the early indicators with with Gen.ai, what 78 00:06:10,518 --> 00:06:19,843 its capabilities were the Goldman report, um, you know, the, the, that, you know, three, five scored in the 65 percentile on the bar. 79 00:06:19,843 --> 00:06:25,766 And initially it was reported that four Oh scored in the 90th percentile. 80 00:06:25,766 --> 00:06:39,374 That's been, that's been, I guess, adjusted downward quite a bit, but in the beginning, there was a lot of reason for outsiders to the initial data points really kind of pointed. 81 00:06:39,630 --> 00:06:43,843 trajectory that was extremely steep on the transformation curve. 82 00:06:43,843 --> 00:06:51,478 And the reality is, in my opinion, things have been grossly overestimated in the short term. 83 00:06:51,478 --> 00:06:58,773 AI is absolutely going to be transformative in the legal industry, but it's going to take a little while. 84 00:06:58,773 --> 00:07:09,120 I'm not sure what the perspective is with investors who aren't used to disrupting inside of legal if their timeline 85 00:07:09,292 --> 00:07:11,225 how well that is aligned with reality. 86 00:07:11,225 --> 00:07:12,691 Do you have any sense of that? 87 00:07:13,224 --> 00:07:15,595 Yeah, it's totally not in line with reality. 88 00:07:15,595 --> 00:07:21,978 I think, you know, part of me has a lot of empathy for why the investors are so interested in legal. 89 00:07:21,978 --> 00:07:34,313 You look at a market where you see firms charging thousands of dollars an hour, handling matters and doing workflows that are really inefficient, not really leveraging software. 90 00:07:34,313 --> 00:07:39,045 So naturally investors are going to think, well, we can disrupt that with tech. 91 00:07:40,626 --> 00:07:42,687 I think that does happen in many industries. 92 00:07:42,687 --> 00:07:51,293 And I think it's been slower in legal because there's a fundamental misunderstanding of why clients pay so much money to firms. 93 00:07:51,294 --> 00:08:03,141 is tempting to say, Hey, if you're paying an outside firm $2,000 an hour, that somehow, um, a streamlined AI or tech solution can do it at a fraction of the price. 94 00:08:03,182 --> 00:08:08,245 And so, yeah, all the clients will kind of vote with their feet and go to the AI solution. 95 00:08:08,245 --> 00:08:09,460 But the truth is. 96 00:08:09,460 --> 00:08:13,833 firms offer a lot more than just the brief or memo that you see. 97 00:08:14,674 --> 00:08:20,308 What you're really paying for as a client is a bucket of things that are really valuable. 98 00:08:20,308 --> 00:08:27,583 So for example, I've written about the trust and long-term branding that firms care about. 99 00:08:27,583 --> 00:08:37,322 when you work with a firm, you know that the firm has spent decades, if not centuries, building up a brand of client satisfaction, keeping their clients happy, doing good work. 100 00:08:37,322 --> 00:08:43,682 And so you're also trusting that the firm is going to be reliable in what it produces. 101 00:08:44,162 --> 00:08:54,082 AI, for example, if you're replacing that firm with a startup, maybe that's a series A startup that's just been around for a year or two. 102 00:08:54,362 --> 00:09:02,202 so maybe the work that it does is 80 % of the way there, which it could very well be, but that last 20 % matters a lot. 103 00:09:02,202 --> 00:09:06,862 And so it's really hard to tell from a client's perspective if they're getting 104 00:09:07,018 --> 00:09:12,458 really the high quality work from a startup that they would from a law firm. 105 00:09:12,458 --> 00:09:18,858 And it turns out that that 20 % or whatever, 10%, whatever you want to call it, that little bit is so important. 106 00:09:21,278 --> 00:09:27,478 It's because legal is not in the business of something like sales or marketing, where if you roughly get it right, it's OK. 107 00:09:27,478 --> 00:09:32,358 If you get it wrong just a little bit in legal, it has huge consequences. 108 00:09:32,358 --> 00:09:34,416 So I think that's fundamentally why 109 00:09:34,416 --> 00:09:40,315 legal work is different than say work done by other functions or other industries. 110 00:09:40,482 --> 00:09:51,377 Yeah, and you had a really good point about why high hourly rates aren't nearly the problem for clients that they would seem to be. 111 00:09:51,458 --> 00:09:53,900 Let's drill into that a little bit. 112 00:09:55,944 --> 00:09:57,896 expand on that thought. 113 00:09:58,346 --> 00:10:06,465 Yeah, so when I first started off as an associate at Sullivan and Cromwell, I heard stories of how much we were charging clients and how much we were being billed out on an 114 00:10:06,465 --> 00:10:07,056 hourly basis. 115 00:10:07,056 --> 00:10:09,436 And I was like this like fresh associate. 116 00:10:09,436 --> 00:10:10,836 You know, I didn't really have that much expertise. 117 00:10:10,836 --> 00:10:12,666 I knew that I just worked hard, right? 118 00:10:12,666 --> 00:10:18,366 And I'd like to think I was smart, but I, it just, it just confused me why these clients would pay so much. 119 00:10:18,366 --> 00:10:27,486 And so it took me a while, but I kind of zoomed out and I realized, okay, well, if you add up my hours and the hours of other members of my team and all the way up to the partners, 120 00:10:27,486 --> 00:10:33,089 and you kind of come up with a number every month, let's call it like 5 million a month. 121 00:10:33,089 --> 00:10:42,474 That's a pretty significant legal bill on an absolute scale, but compared to the deals and cases that we were working, these were billion dollar matters. 122 00:10:42,474 --> 00:10:55,732 so yes, the hourly rate was high, but when you kind of compared the total legal fees generated and compared it to the matter itself, it was such a small percentage that it 123 00:10:55,732 --> 00:10:57,082 made me realize that 124 00:10:57,110 --> 00:10:59,661 Um, you know, there's more than meets the eye. 125 00:10:59,661 --> 00:11:02,272 You can't just look at the hourly rate in a vacuum. 126 00:11:02,272 --> 00:11:11,226 Yes, that would be an insane hourly rate to charge for an individual who had say, you know, a trust in a state's matter, or I don't know, like a real estate transaction, like 127 00:11:11,226 --> 00:11:21,480 you're buying a house, but on the corporate side, um, the numbers get so big that, that legal fees, um, are often just a drop in the bucket. 128 00:11:21,480 --> 00:11:23,621 the re really hit home for me. 129 00:11:23,621 --> 00:11:25,142 I didn't really realize that until much later. 130 00:11:25,142 --> 00:11:25,982 I didn't, 131 00:11:26,106 --> 00:11:31,007 The lesson didn't hit home for me until much later when I realized that hourly rates kept on going up. 132 00:11:31,127 --> 00:11:33,448 If they were so high, why would they continue going up? 133 00:11:33,448 --> 00:11:35,249 Why wouldn't clients leave? 134 00:11:35,249 --> 00:11:45,521 And it made me also understand that there are a lot of matters that law firms handle, especially the AMLaw firms, the largest firms in the world handle, that they're getting a 135 00:11:45,521 --> 00:11:54,558 great deal from the law firm because the clients are getting a great deal because if they're willing to pay so much and they're not willing to leave, 136 00:11:54,558 --> 00:11:56,369 that it must be that they're getting something out of it. 137 00:11:56,369 --> 00:12:01,803 I think that was hard for me to understand and grasp because I was using my own frame of reference for the numbers. 138 00:12:01,803 --> 00:12:05,776 And I think a lot of people fall into this trap, especially those in the tech space. 139 00:12:05,776 --> 00:12:09,849 Like they think $1,000 an hour for a junior associate is so much. 140 00:12:09,849 --> 00:12:13,452 when you zoom out, really, the clients are the ones that are voting with their feet. 141 00:12:13,452 --> 00:12:16,434 And so they are continuing to pay these rates and the rates keep going up. 142 00:12:16,434 --> 00:12:22,308 So there must be something there that we're not seeing on a surface level. 143 00:12:22,796 --> 00:12:23,016 Yeah. 144 00:12:23,016 --> 00:12:36,599 And you also talked a little bit about how the, you know, decades old or century plus old reputation of the law firms inhibit disruption, right? 145 00:12:36,599 --> 00:12:38,421 Like talk, talk a little bit about that. 146 00:12:38,421 --> 00:12:44,276 You kind of touched on it high level, but how specifically does that come into play? 147 00:12:44,862 --> 00:12:49,603 Yeah, so I think a good example to compare it to is a startup sales team. 148 00:12:49,703 --> 00:12:56,365 For a startup, you need revenue or else you're not gonna succeed. 149 00:12:56,365 --> 00:13:05,768 And so there's a huge premium placed on experimentation, trying new technology, doing things more efficiently, because for sales, you just need to be like 50 % right. 150 00:13:05,968 --> 00:13:10,329 Whether it's email automation that's sent out that may have typos in it, doesn't really matter. 151 00:13:10,329 --> 00:13:13,126 As long as the emails are going out, you're gonna generate sales. 152 00:13:13,126 --> 00:13:15,707 And so it's a perfect fit. 153 00:13:15,808 --> 00:13:18,690 Law firms on the other hand, can't do the same thing. 154 00:13:18,931 --> 00:13:28,519 They've been around for decades and centuries, as I mentioned, and they often get their business because they are trusted advisors to the business. 155 00:13:28,519 --> 00:13:34,624 They have longstanding strong reputations individually in the legal community. 156 00:13:34,624 --> 00:13:42,850 And so people know that if I use this lawyer or this firm, I can be sure that I'm going to get a quality work product out of them. 157 00:13:43,710 --> 00:13:51,552 that type of business does not lend itself so well to technology that gets it 50 % right or 80 % right. 158 00:13:51,552 --> 00:14:04,816 I mentioned this earlier, but the bar for firms is much higher when it comes to the reliability of the technology than it is for most other businesses. 159 00:14:05,076 --> 00:14:13,758 And if you look at kind of how technology has made its way into firms, you can see, right, very often functionality and features 160 00:14:13,780 --> 00:14:18,412 play a secondary role to things like data security or reliability. 161 00:14:18,813 --> 00:14:24,656 The firms want to make sure that what they're using in terms of software or technology is reliable. 162 00:14:24,656 --> 00:14:34,951 And that's because they have this duty to their clients to make sure that they're not passing on or relying on technology that's not reliable, that leads to bad advice, that 163 00:14:34,951 --> 00:14:39,184 would blow up their reputation that they've taken decades and centuries to build. 164 00:14:39,184 --> 00:14:40,745 that's what makes law firms very different. 165 00:14:40,745 --> 00:14:43,586 And there's a few other ways that law firms 166 00:14:44,270 --> 00:14:46,703 are different and where they sit in the entire ecosystem. 167 00:14:46,703 --> 00:14:50,287 But I would say that is why the bar is so high for firms. 168 00:14:50,287 --> 00:15:01,310 And that's why software companies often find themselves paying more attention to making sure they have data security and reliability versus the latest and greatest features. 169 00:15:01,708 --> 00:15:01,978 Yeah. 170 00:15:01,978 --> 00:15:08,521 And you know, another way I think, reputation law firm reputation, it affects the dynamics of the matter itself. 171 00:15:08,521 --> 00:15:19,566 If you know, a one side comes to the table with, you know, do we cheat them and how down the street and the other side is Kirkland and Ellis, you know, you've got that, that 172 00:15:19,566 --> 00:15:25,589 affects the dynamics before any legal, um, actions take place. 173 00:15:25,589 --> 00:15:28,930 Just the weight of the reputation of the firm. 174 00:15:29,240 --> 00:15:41,960 can ultimately impact the outcome of the case because the other side understands the capabilities, the reputation, the success rate, the relationships, the expertise, all of 175 00:15:41,960 --> 00:15:46,163 those things that a law firm brand brings to the table. 176 00:15:46,163 --> 00:15:49,826 And that is very difficult to replicate as well. 177 00:15:49,826 --> 00:15:50,622 Agreed? 178 00:15:50,622 --> 00:15:55,185 yeah, there's this like really famous saying like nobody ever got fired for using IBM. 179 00:15:55,605 --> 00:15:59,417 I think that, you know, it says a lot of things. 180 00:15:59,417 --> 00:16:03,589 It's saying that if you rely on a brand name vendor, you're not going to go wrong. 181 00:16:03,830 --> 00:16:13,735 That can be overstated and overdone in many parts of the world, but in legal, you know, you can replace IBM with the name of, you know, some of these most well-known firms like 182 00:16:13,735 --> 00:16:14,746 Cravath as an example. 183 00:16:14,746 --> 00:16:18,818 I think I've heard nobody ever got fired for going with Cravath. 184 00:16:18,818 --> 00:16:19,678 Now, 185 00:16:20,020 --> 00:16:30,605 part of the reason is because as you say, hiring a Cravath or a Kirkland and Ellis or whoever, that does send a message to the opposing counsel that you're, you you mean 186 00:16:30,605 --> 00:16:36,317 business, you're ready to spend to win the case, but it also sends a message to your internal stakeholders. 187 00:16:36,317 --> 00:16:47,374 So for example, if you're the chief legal officer of a publicly traded company and you, you know, retain Cravath or Kirkland or whoever, you know, that creates this level of 188 00:16:47,374 --> 00:16:52,055 It sends a message to your board because the board is likely paying attention to this litigation or matter. 189 00:16:52,676 --> 00:17:06,040 And the flip side of that is that if you do rely on some other firm with a lighter reputation or a firm that no one's really heard of, that can impact the perception of your 190 00:17:06,040 --> 00:17:08,960 competence as a CLO, as a chief legal officer. 191 00:17:08,960 --> 00:17:12,021 so legal departments are very savvy about this. 192 00:17:12,021 --> 00:17:14,652 They know that for the most important matters, 193 00:17:14,866 --> 00:17:21,530 which firm they choose sends a message to their stakeholders, especially their board and to their leadership, like their CEO. 194 00:17:22,090 --> 00:17:28,554 that's why it's not as simple as let's go with the cheaper option. 195 00:17:28,914 --> 00:17:35,678 Again, the clients are getting something, a lot of things out of spending that premium for more expensive firms. 196 00:17:35,810 --> 00:17:37,591 Yeah, absolutely. 197 00:17:37,591 --> 00:17:50,174 Well, switching gears a little bit, talking about, uh, you and I had a great conversation about AI and its transformative effect, its impact on, on jobs in the industry. 198 00:17:50,174 --> 00:18:03,497 And, know, I'm a firm believer there's, there's a fair number of analogs that we can look at to gauge what happens when game changing technology enters the marketplace and drives 199 00:18:03,497 --> 00:18:05,774 efficiency significantly higher. 200 00:18:05,774 --> 00:18:10,074 A good example is spreadsheets in accounting. 201 00:18:10,074 --> 00:18:17,874 So when Lotus one, two, three came out sometime in the early eighties, I think it was 1983, 1984, I went back and looked at a little research. 202 00:18:17,874 --> 00:18:24,334 There were about 850,000 professionals in the accounting industry at that time. 203 00:18:24,334 --> 00:18:28,734 I, and I happen to remember, this is when I was really getting into computers. 204 00:18:28,734 --> 00:18:32,034 I was what 11, 12 years old at the time. 205 00:18:32,034 --> 00:18:34,772 And, um, I remember 206 00:18:34,958 --> 00:18:40,858 My mom was in the restaurant business and she had an accountant who was completely had no, he didn't have a computer in his office. 207 00:18:40,858 --> 00:18:50,158 And I was in the back of the restaurant working on, uh, that my mom owned and he was asking me questions and it didn't occur to me until much, much later why he was, he was 208 00:18:50,158 --> 00:18:50,418 worried. 209 00:18:50,418 --> 00:18:51,818 He was concerned. 210 00:18:51,818 --> 00:19:01,658 Um, and what has happened in that 20 or 30 years since the number of professionals in the industry has almost doubled. 211 00:19:01,658 --> 00:19:02,238 Right? 212 00:19:02,238 --> 00:19:03,270 So they're, 213 00:19:03,350 --> 00:19:07,072 there's been a huge net gain of jobs, not a net loss. 214 00:19:07,392 --> 00:19:10,892 There's so much new legislation coming out constantly. 215 00:19:10,892 --> 00:19:13,595 I think there will be no legal jobs lost. 216 00:19:13,595 --> 00:19:21,680 I think there will be a ton of impact and redistribution of roles as a result of this technology. 217 00:19:21,680 --> 00:19:26,742 But I don't know, what is your take on how AI is going to impact the job market itself? 218 00:19:27,206 --> 00:19:29,869 I think it's going to create more demand for legal services. 219 00:19:29,869 --> 00:19:37,255 We're starting to see that with litigation, also regulation coming out around AI. 220 00:19:37,255 --> 00:19:39,247 Every company is going to soon become an AI company. 221 00:19:39,247 --> 00:19:45,813 And so that's going to create a lot of impact on what type of legal advice is needed. 222 00:19:45,813 --> 00:19:53,909 And I generally believe something similar to what you believe, which is this technology creates new roles. 223 00:19:54,378 --> 00:20:03,923 commentators and most of us, when we look at technology, we think, this is going to replace the way many of these people who have these jobs right now. 224 00:20:03,923 --> 00:20:06,524 So inevitably it's going to lead to job loss. 225 00:20:07,666 --> 00:20:14,789 For legal in particular, I do think that we underestimate how much complexity technology adds. 226 00:20:15,630 --> 00:20:18,891 And that complexity needs to be solved by someone. 227 00:20:19,032 --> 00:20:20,593 New problems are being created every day. 228 00:20:20,593 --> 00:20:23,384 Like I never could have imagined that what impact 229 00:20:23,454 --> 00:20:26,176 AI would have on copyright and IP. 230 00:20:26,176 --> 00:20:30,779 There's going to be a lot more demand for that practice area. 231 00:20:30,779 --> 00:20:39,684 So there's going to be a lot of things like that where we don't expect to know where increased demand of services come from. 232 00:20:39,684 --> 00:20:49,640 I can't imagine a world where AI replaces jobs and then doesn't create new ones through the increasing complexity of laws and regulation and lawsuits. 233 00:20:49,848 --> 00:20:50,188 Yeah. 234 00:20:50,188 --> 00:20:54,620 And how much unmet need is there in the legal world? 235 00:20:54,620 --> 00:21:08,546 mean, you and I talked about startups and you know, uh, we are, uh, we have a very small investment from a venture fund, so we're largely bootstrapped and you know, we don't have 236 00:21:08,706 --> 00:21:15,149 tens of millions of dollars in our war chest that we can deploy, um, inefficiently. 237 00:21:15,149 --> 00:21:18,750 have to spend our dollars very carefully and we do. 238 00:21:18,958 --> 00:21:27,898 And as a result, you know, we, we negotiate, uh, we've been doing this long enough that, um, we negotiate a lot of red lines on our own. 239 00:21:27,898 --> 00:21:39,598 Um, and you know, maybe we'll have a lawyer look at the final, the final draft before we execute the document, but we don't, we don't send our lawyers in with our clients, lawyers 240 00:21:39,598 --> 00:21:41,038 and let them have at it. 241 00:21:41,038 --> 00:21:42,478 No, we manage a lot of that. 242 00:21:42,478 --> 00:21:49,128 Ideally, I'd love to have a legal counsel manage the whole process, but it's just, it's, it's out of reach. 243 00:21:49,312 --> 00:21:50,873 financially for us to do that. 244 00:21:50,873 --> 00:22:03,806 So, you know, there's been, there's been firms out there like Wilson, Sincini with Neuron, think Cooley go, there's like some platforms out there where they've tried to, um, 245 00:22:03,806 --> 00:22:16,590 automate, uh, some of this, you know, like even I can't remember which, which firm it is, you know, they, they will draft your offer letters, which, know, again, it'd be really 246 00:22:16,590 --> 00:22:17,742 tough to have a 247 00:22:17,742 --> 00:22:19,797 person sit there and do that for free. 248 00:22:19,797 --> 00:22:21,692 you know, I don't know. 249 00:22:21,692 --> 00:22:26,252 How do you see automation increasing access to legal services? 250 00:22:27,294 --> 00:22:29,535 I think more firms will become more in tech enabled. 251 00:22:29,535 --> 00:22:33,997 And what that means is that their cost structure for delivering legal services will change. 252 00:22:34,298 --> 00:22:44,424 You're seeing that with Neuron as an example, like they are leveraging the know-how and expertise of their lawyers to automate the creation of certain types of documents. 253 00:22:44,424 --> 00:22:47,846 Before you would have to have a whole team of lawyers maybe drafting something. 254 00:22:47,846 --> 00:22:49,817 Now it can be replicated at a low cost. 255 00:22:49,817 --> 00:22:53,228 And so that means that they can pass on some of those savings to clients. 256 00:22:53,329 --> 00:22:55,791 So I do think that while... 257 00:22:56,222 --> 00:23:07,605 technology and AI is not going to replace lawyers, it will become infused in the practice such that the firms that lean on innovation or technology will have an advantage. 258 00:23:07,605 --> 00:23:11,966 And that does have broader impact on who gets access to legal services. 259 00:23:11,966 --> 00:23:23,150 But the key takeaway though is that I do think that the firms that have leaned into innovation, whether it's experimenting, trialing software, building their own software, 260 00:23:23,150 --> 00:23:25,310 just really focusing on innovation. 261 00:23:25,372 --> 00:23:34,586 I do think that they'll have an advantage because what then happens is they can offer more to their clients or new clients at a lower price point. 262 00:23:35,186 --> 00:23:48,712 And as I mentioned, it is true that many of these firms can continue to charge high rates, but if they can deliver that at a lower cost and either keep their billable hours or rates 263 00:23:48,712 --> 00:23:52,733 similar or higher, that just means more profits. 264 00:23:52,733 --> 00:23:55,192 so I do think that... 265 00:23:55,192 --> 00:23:57,934 tech enabled is the future. 266 00:23:58,296 --> 00:23:59,166 Yeah. 267 00:23:59,186 --> 00:24:05,230 So what about the AI's current capabilities versus the expectations in the marketplace? 268 00:24:05,230 --> 00:24:07,170 It feels like there's a disconnect. 269 00:24:08,432 --> 00:24:18,437 in my mind, you know, I see, for example, I see firms rolling out technologies like co-pilot, Microsoft co-pilot wholesale, like enterprise wide. 270 00:24:18,437 --> 00:24:25,141 Um, it's happened in a few places and you know, my take on co-pilot is it's not very good. 271 00:24:25,141 --> 00:24:27,950 Um, in fact it is. 272 00:24:27,950 --> 00:24:32,910 very much inferior to chat GPT and Claude. 273 00:24:32,910 --> 00:24:40,610 Even though, you know, Microsoft is a huge investor in open AI and they clearly have access to those, to the foundational model. 274 00:24:40,610 --> 00:24:45,150 I'm not sure exactly what they're doing to make it worse, but it is, it is worse. 275 00:24:45,150 --> 00:24:57,222 As an example, I, you know, I play around with it a lot and I test it, you know, in, in outlook, I went into, pulled up the co-pilot side panel. 276 00:24:57,576 --> 00:24:57,786 it. 277 00:24:57,786 --> 00:25:01,289 How many, how many calendar entries do I have? 278 00:25:01,289 --> 00:25:03,990 I'll block time off for thought leadership. 279 00:25:03,990 --> 00:25:09,554 How many calendar entries do I have on my calendar this week for thought leadership? 280 00:25:09,554 --> 00:25:12,056 Came back and said, you have nine. 281 00:25:12,056 --> 00:25:13,906 I actually had 12. 282 00:25:14,478 --> 00:25:15,498 So it was incorrect. 283 00:25:15,498 --> 00:25:18,981 And I know AI is not great at counting, but hear me out on this. 284 00:25:18,981 --> 00:25:20,822 So then I went back in and I reprompted. 285 00:25:20,822 --> 00:25:23,013 I was like, why did you say I only had nine? 286 00:25:23,013 --> 00:25:25,285 I actually have 15, even though I only had 12. 287 00:25:25,285 --> 00:25:26,075 He goes, you're right. 288 00:25:26,075 --> 00:25:27,546 You do have 15. 289 00:25:28,302 --> 00:25:32,925 So then to further test this, I was like, all right, AI is not great at counting. 290 00:25:32,925 --> 00:25:37,747 Let me, let me see if this is a copilot problem or there's a fundamental LLM problem. 291 00:25:37,747 --> 00:25:44,851 So I took a screenshot and I went to chat GPT and Claude and I uploaded the screenshot with the same prompt. 292 00:25:44,851 --> 00:25:46,331 They both got it right. 293 00:25:46,432 --> 00:25:54,286 So it's like, um, you know, I feel like the marketing and I know Microsoft's going to get copilot right at some, at some point, right? 294 00:25:54,286 --> 00:25:58,388 They're, too invested in this, but right now it's just, it's just not good in my opinion. 295 00:25:58,646 --> 00:26:09,272 You know how how big of a gap do you think there is between the capabilities that are being marketed by people like Microsoft who are trusted in the space and there's this 296 00:26:09,272 --> 00:26:18,517 whole agentic thing which that's we can talk about that later that they're pitching with Copilot Studio but I don't know do you feel like they're in line or do you feel like me 297 00:26:18,517 --> 00:26:20,498 that there's a bit of a disconnect. 298 00:26:21,736 --> 00:26:28,562 I think we're all trying to figure out what AI is capable of and it's hard not to get caught up in the possibility. 299 00:26:29,050 --> 00:26:31,913 I've not had that experience with Copilot. 300 00:26:31,913 --> 00:26:34,634 I have only really played around with ChatGPT. 301 00:26:34,695 --> 00:26:40,029 And so my personal experience really is probably more limited than yours, Ted. 302 00:26:40,029 --> 00:26:53,720 But what I have noticed is that if you can do some really cool stuff with something like ChatGPT, then your mind starts thinking that it could be done for a lot of other broader 303 00:26:53,720 --> 00:26:56,532 use cases that may not be a good fit. 304 00:26:56,532 --> 00:26:58,824 The classic example in legal is 305 00:26:58,824 --> 00:26:59,465 legal research. 306 00:26:59,465 --> 00:27:07,280 You remember like when it first came out, when chat GPT first came out, everyone started using it for legal research and then started hallucinating cases and lawyers got in 307 00:27:07,280 --> 00:27:08,221 trouble. 308 00:27:08,602 --> 00:27:15,645 I think that this tech enthusiasm leads us down the wrong path. 309 00:27:15,887 --> 00:27:25,404 What typically happens, and I'm just analogizing to other non-AI related innovations, is new technology comes out, people get excited about it. 310 00:27:25,404 --> 00:27:32,636 And the initial applications are in fast moving industries or sectors like sales and marketing, maybe customer service. 311 00:27:33,156 --> 00:27:35,974 And then it takes a while before it really makes its way to legal. 312 00:27:35,974 --> 00:27:37,898 I mean, you cloud as an example. 313 00:27:37,898 --> 00:27:45,200 Like we were using Dropbox and Box and things like that, like cloud-based document storage for a long time before firms did. 314 00:27:45,200 --> 00:27:50,302 And probably for good reason, because you kind of want things to be tested out and validated. 315 00:27:51,438 --> 00:28:01,847 We got to a point eventually where I think in 2024 right now, late 2024, adopting the cloud for firms seems not that controversial, but it took a long time. 316 00:28:01,847 --> 00:28:03,578 And so I think that's going to happen with AI too. 317 00:28:03,578 --> 00:28:07,552 Like right now, people are really stoked about it, really excited about it. 318 00:28:07,552 --> 00:28:11,505 And so it's inevitably going to have heightened expectations. 319 00:28:11,505 --> 00:28:19,926 You see the headlines of AI replacing lawyers and tweets that say AI is going to replace law firms, which again is how this all started. 320 00:28:19,926 --> 00:28:29,422 Um, but yeah, I think, I think you're right that it's, you know, we've got to keep our expectations realistic and, let this, this play out, um, before diving headfirst into it. 321 00:28:29,720 --> 00:28:30,151 Yeah. 322 00:28:30,151 --> 00:28:40,999 So along those lines, how should law firms be thinking about ROI when it comes to these GENAI platforms? 323 00:28:40,999 --> 00:28:42,730 Do you have any insight on that? 324 00:28:43,900 --> 00:28:50,093 I have some, I'll probably speak in generalities, but from what I've seen, it's got to tie into with your practice. 325 00:28:50,093 --> 00:29:00,967 Like it can be tempting to see the latest and shiniest new AI tool that could do so much, but if it doesn't fit in with your practice, I think you're setting yourself up for 326 00:29:00,967 --> 00:29:01,538 failure. 327 00:29:01,538 --> 00:29:05,909 And what I mean by that is, know, some firms do rely on outside vendors. 328 00:29:05,909 --> 00:29:08,541 They will bring on new technologies for AI. 329 00:29:08,541 --> 00:29:09,901 Totally makes sense. 330 00:29:09,941 --> 00:29:12,287 Others will create their own AI. 331 00:29:12,287 --> 00:29:16,369 that's tied to their own specific practice area strengths. 332 00:29:16,369 --> 00:29:17,789 That'll work too. 333 00:29:18,149 --> 00:29:29,824 So there's not one clear way to do it, but I do think that part of what's driving this rush to AI from the firm perspective, my sense is that many clients, and when I say 334 00:29:29,824 --> 00:29:39,017 clients, I really mean corporate clients, the legal departments that serve as firm clients are themselves under pressure to use AI. 335 00:29:39,838 --> 00:29:46,700 Boards are telling CEOs leverage AI, CEOs are telling their chief legal officers and GCs use AI. 336 00:29:46,920 --> 00:29:53,762 And so the legal department is also probably very interested in which firms are leveraging AI to generate efficiencies. 337 00:29:53,762 --> 00:29:55,802 Now we don't know if there's truly efficiencies. 338 00:29:55,802 --> 00:30:05,055 I think that the jury is still out on that, but there's no question that there is a broad mandate from the top down that's trickling its way out to law from boards and CEOs to the 339 00:30:05,055 --> 00:30:08,286 outside firms that serve them to use AI. 340 00:30:08,510 --> 00:30:21,667 And what I've seen really, you know, some firms be really smart about it is that instead of just, you know, buying the latest and greatest tool from a startup, which, know, often 341 00:30:21,667 --> 00:30:23,979 can make sense again, like depending on what you need. 342 00:30:23,979 --> 00:30:31,283 But I've seen some firms develop their own AI and, you know, build their own models that are tied to their own proprietary data. 343 00:30:31,343 --> 00:30:37,222 I shouldn't say proprietary data, but like their, their own documents, because it can be hard for a firm to take 344 00:30:37,222 --> 00:30:48,665 know, client matters and, kind of like, um, you know, put it onto some third party startups training algorithm, but if they do it themselves, that that's different. 345 00:30:48,705 --> 00:31:04,189 so, you know, I've personally seen, uh, some firms really leverage the, um, their own data and using AI to kind of create their own process that creates value in ROI. 346 00:31:04,189 --> 00:31:06,792 Um, it's not as exciting and sexy as like, you know, 347 00:31:06,792 --> 00:31:14,148 all the headlines, but that is something I've seen firms really benefit from AI. 348 00:31:14,744 --> 00:31:15,084 Yeah. 349 00:31:15,084 --> 00:31:25,497 So along those lines, you and I spoke about scenarios where there are alternative approaches like to staffing models. 350 00:31:25,497 --> 00:31:32,839 You know, law firms have to get their, their, their, their house in order really to leverage AI effectively. 351 00:31:32,839 --> 00:31:39,881 You know, most, if you're AmLaw 50 or the floor of the AmLaw 50 is around a thousand attorneys, you have a thousand attorneys or more. 352 00:31:39,881 --> 00:31:44,322 You have probably about a minimum of a hundred million documents. 353 00:31:44,566 --> 00:31:49,609 in your DMS and that could be all sorts of content. 354 00:31:49,609 --> 00:31:50,890 could be emails. 355 00:31:50,890 --> 00:31:53,851 could be unused drafts. 356 00:31:53,851 --> 00:32:01,214 It could be red lines, know, 17 versions back from the final product. 357 00:32:01,636 --> 00:32:06,118 And it's a big, it's a heavy lift, right? 358 00:32:06,158 --> 00:32:08,069 You can't manually curate. 359 00:32:08,069 --> 00:32:13,722 You can't build an army of people big enough to manually curate the entire thing. 360 00:32:14,060 --> 00:32:15,521 Like wholesale, right? 361 00:32:15,521 --> 00:32:27,218 You have to kind of strategically attack the problem, but you know, it, sounded like you had seen some, and, and, and I think maybe you guys, that's part of what you do is, um, 362 00:32:27,378 --> 00:32:30,891 around staffing for those types of engagements. 363 00:32:30,891 --> 00:32:33,502 Like tell me, tell me a little bit about that. 364 00:32:33,864 --> 00:32:34,174 Yeah. 365 00:32:34,174 --> 00:32:45,180 So, you we talked about how AI is such a critical part of a firm's competitive advantage or could be, but there's certainly a lot of other technology related initiatives that, 366 00:32:45,180 --> 00:32:49,342 knowledge management or chief innovation officers are well aware of. 367 00:32:49,503 --> 00:32:50,824 Now here's the problem. 368 00:32:50,824 --> 00:32:58,948 Because of the staffing model of firms, you have associates and counsel who are essentially being billed out at very high rates. 369 00:32:58,948 --> 00:33:02,356 And so the opportunity cost of using your own people 370 00:33:02,356 --> 00:33:06,468 to work on these internal strategic projects is very high. 371 00:33:06,568 --> 00:33:18,875 How it plays out is essentially, let's say that your firm's decided to launch a new AI initiative, build its own proprietary AI related to a specific practice area. 372 00:33:19,075 --> 00:33:24,378 And you basically need associates from that group to train the model. 373 00:33:24,718 --> 00:33:28,080 So KM knows this, the chief innovation officer knows this. 374 00:33:28,080 --> 00:33:30,782 And so they ask, hey, who can help train this model? 375 00:33:30,782 --> 00:33:31,870 Problem is, 376 00:33:31,870 --> 00:33:39,015 these associates are so busy, they're not gonna put the time and energy into training that AI. 377 00:33:39,015 --> 00:33:42,678 They're gonna instead work on some matter that the partner assigns them. 378 00:33:42,718 --> 00:33:48,131 And so this is, think, one big challenge that firms are running into that Latitude helps with. 379 00:33:48,131 --> 00:33:58,670 So we've worked with AMLaw firms that are struggling to staff their strategic initiatives around KM, around documents and AI. 380 00:33:59,026 --> 00:34:06,351 And so we provide this associate level, so same caliber as the firm's associates on a contract basis. 381 00:34:06,371 --> 00:34:11,595 And those folks are able to get the work done without that same opportunity cost. 382 00:34:11,595 --> 00:34:23,884 And because we provide these contract attorneys at a fraction of the rate that these existing full-time associates bill out at, it tends to be a financially attractive 383 00:34:23,884 --> 00:34:24,645 proposition. 384 00:34:24,645 --> 00:34:27,016 So I think that... 385 00:34:27,016 --> 00:34:36,900 because of this drive to AI for firms to build their own AI, it's created opportunities for companies like us, like Latitude, where we can provide high-end, high-caliber 386 00:34:36,900 --> 00:34:40,632 attorneys to work on these projects so they don't kind of fall by the wayside. 387 00:34:40,632 --> 00:34:49,555 I think that's the big worry is that, know, KM, knowledge management officers and folks, they know that some of these projects are so important, but it's just a question of how 388 00:34:49,555 --> 00:34:55,668 are they going to get that subject matter expertise to train the technology to work on those projects? 389 00:34:55,668 --> 00:34:56,756 And I think that's where we... 390 00:34:56,756 --> 00:34:57,667 Fill the gap. 391 00:34:58,166 --> 00:35:03,598 And then why is there such a strong interest in building proprietary AI systems? 392 00:35:03,598 --> 00:35:05,868 I mean, I know a lot of firms are doing it. 393 00:35:05,888 --> 00:35:22,593 know Gunderson, I had Naveen and Joe Green on from Gunderson a little over a year ago when ChatGD came out and it was novel at the time, but many firms have done it since then. 394 00:35:23,733 --> 00:35:26,324 Why is there such a strong push in that direction? 395 00:35:26,622 --> 00:35:27,643 I think it's twofold. 396 00:35:27,643 --> 00:35:30,604 First, it's to show clients that they're serious about AI. 397 00:35:30,604 --> 00:35:35,887 And I mentioned earlier, that mandate comes all the way straight from the top of the client's organization. 398 00:35:35,887 --> 00:35:38,729 showing your clients that you're focused on AI helps. 399 00:35:38,949 --> 00:35:45,633 That's probably especially true for a firm like Gunderson or Wilson-Sonsini, these firms that serve the technology industry. 400 00:35:46,514 --> 00:35:48,854 So that's at one level. 401 00:35:49,555 --> 00:35:53,657 The second reason has to do with the cost of delivery. 402 00:35:54,902 --> 00:36:03,726 we don't know if any of these chatbots or proprietary AI or any of these models will actually end up being more efficient. 403 00:36:03,726 --> 00:36:08,207 There's certainly a lot of risks of imperfection, hallucinations, things like that. 404 00:36:08,328 --> 00:36:21,603 But if AI can fulfill its promise, that could enable a firm to essentially provide the same level of service, generate the same amount of revenue and billable hours at a much 405 00:36:21,603 --> 00:36:22,424 lower cost structure. 406 00:36:22,424 --> 00:36:23,732 And again, that goes to the 407 00:36:23,732 --> 00:36:24,533 firms profits. 408 00:36:24,533 --> 00:36:34,399 And so I think the possibility that technology can create higher levels of profit, I think that is a huge opportunity. 409 00:36:34,399 --> 00:36:46,537 And so those two reasons both work together to kind of drive firms towards tech adoption in ways that they haven't before, honestly, with the cloud, the internet, other 410 00:36:46,537 --> 00:36:47,678 innovations. 411 00:36:47,982 --> 00:36:48,662 Yeah. 412 00:36:48,662 --> 00:36:54,606 You know, so we used to be a custom dev shop at, uh, our previous company was called Acrowire. 413 00:36:54,606 --> 00:37:02,470 We kind of rebranded and relaunched this InfoDash, a product company in about three years ago. 414 00:37:02,470 --> 00:37:04,791 Um, and our timing was great. 415 00:37:04,791 --> 00:37:07,132 It's we've had a fabulous run. 416 00:37:07,132 --> 00:37:17,646 Things are, we're having a lot of fun over here, but for many years we would develop solutions for clients and they would 417 00:37:17,646 --> 00:37:21,726 really have challenges during the handoff. 418 00:37:22,006 --> 00:37:30,066 sometimes it was, you know, KM and innovation would work with us to build a solution that ultimately IT had to support. 419 00:37:30,066 --> 00:37:38,626 And maybe IT had one ear in on the project, but didn't fully grasp what was coming their way when it was time to transition. 420 00:37:38,626 --> 00:37:44,686 You know, and much of what I'm talking about here, it was during the on-prem days, right? 421 00:37:44,686 --> 00:37:47,718 So they had to stand up infrastructure sometimes that they weren't 422 00:37:47,906 --> 00:37:54,228 really familiar with or didn't have the capabilities to manage effectively. 423 00:37:54,448 --> 00:38:07,601 And, you know, I find it interesting that so many firms are doing what you're describing because I've seen a lot of struggles, especially in the mid middle market, right? 424 00:38:07,601 --> 00:38:11,623 Like, so let's say, you know, Amla 50 to a hundred, right? 425 00:38:11,623 --> 00:38:17,114 Where they've got the, they've got the resources to write the checks, but they don't have spared dev. 426 00:38:17,766 --> 00:38:21,398 bandwidth sitting around waiting to for things to do right. 427 00:38:21,398 --> 00:38:28,330 If they do have any any dev resources, they're fully deployed and it's probably a fairly small team. 428 00:38:28,411 --> 00:38:43,417 So it seems like it would be a challenge for firms to go down this path of building, even if it's just a wrapper around, you know, an API for a model. 429 00:38:43,697 --> 00:38:46,338 Have you seen firms have success with this approach? 430 00:38:46,804 --> 00:38:47,855 I think the jury is still out. 431 00:38:47,855 --> 00:38:51,448 haven't seen, you again, it's still early. 432 00:38:51,448 --> 00:38:56,482 And so I think we'll start to see the results in the next year or two. 433 00:38:56,602 --> 00:39:04,909 But generally speaking, I think part of it has to do with what the firm's top initiatives or objectives are. 434 00:39:04,909 --> 00:39:08,903 There are some firms where this is, they're going to dedicate more resources because it's critical to them. 435 00:39:08,903 --> 00:39:11,825 Like you mentioned Gunderson and, you know, I mentioned Wilson-Sansini. 436 00:39:11,825 --> 00:39:13,076 They are serving the tech industry. 437 00:39:13,076 --> 00:39:15,688 And so you're going to see a lot more organizational. 438 00:39:16,110 --> 00:39:19,452 momentum and support for something like this. 439 00:39:19,573 --> 00:39:29,700 You could also imagine another firm that may be focused on another industry, more traditional industries, um, that this may not be an oppressing concern. 440 00:39:29,700 --> 00:39:38,066 may say, you know, I know that for example, there's one firm out there that's already come on record to say we are not going to use any sort of generative AI. 441 00:39:38,066 --> 00:39:41,270 And so, you know, those are two extremes that I just provided. 442 00:39:41,270 --> 00:39:42,850 Um, some are in the middle. 443 00:39:42,850 --> 00:39:43,850 And so. 444 00:39:45,098 --> 00:39:50,178 Not every firm has the same appetite to push forward tech adoption and dedicate the right resources to it. 445 00:39:50,178 --> 00:39:58,678 So think that's going to be a big challenge, but that's also why there's reward for those who can really be ahead of the curve and lean into it and then generate operational 446 00:39:58,678 --> 00:40:01,418 efficiencies, goodwill with their clients. 447 00:40:02,198 --> 00:40:06,158 But I think we're going to have to see how things play out in the next 12 to 24 months. 448 00:40:06,478 --> 00:40:07,158 I the firm. 449 00:40:07,158 --> 00:40:13,202 I think I know the firm you're talking about that they got a lot of heat over that statement. 450 00:40:13,202 --> 00:40:16,524 And it really, they're not a tech forward firm. 451 00:40:16,524 --> 00:40:22,417 In fact, I think, I don't even think they're members of Ilta, but which is a little bit of an outlier. 452 00:40:22,417 --> 00:40:29,551 There's only about three or four Amlaw firms that I know of that aren't Ilta members. 453 00:40:29,551 --> 00:40:35,104 yeah, so I just discovered that we're deploying a new CRM and my sales team. 454 00:40:35,138 --> 00:40:40,001 was like, Hey, can we use the Ilta roster as kind of our contact list? 455 00:40:40,001 --> 00:40:42,723 And they were like, Hey, where's the, where's, are these three firms? 456 00:40:42,723 --> 00:40:44,934 And I went and looked and I'll be darned. 457 00:40:44,934 --> 00:40:46,165 They weren't on there. 458 00:40:46,185 --> 00:40:48,146 So yeah, they caught a lot of heat over that. 459 00:40:48,146 --> 00:40:53,409 It was, I don't think that was the best PR move. 460 00:40:53,409 --> 00:40:56,592 Um, I don't know. 461 00:40:56,592 --> 00:41:00,874 What are your thoughts on somebody putting a stake in the ground and going, yeah, we're not, we're not using gen AI. 462 00:41:02,746 --> 00:41:06,239 I tend to think that firms are pretty smart about all this stuff and sophisticated. 463 00:41:06,239 --> 00:41:10,252 And so there must've been some reason why they decided to put their stake in the ground. 464 00:41:10,252 --> 00:41:21,371 And one thing that I've in my relatively short career, right, in the legal industry have realized is that when it comes to firms, they take a lot of cues from their clients or 465 00:41:21,371 --> 00:41:22,642 their industry. 466 00:41:22,702 --> 00:41:32,700 So for example, you know, we talked about the tech firms that are the firms that support technology companies that are very much tech forward. 467 00:41:33,391 --> 00:41:45,168 They also tend to have pretty relaxed remote or hybrid work schedules where I seem to realize that the firms that rely heavily or have served the commercial real estate 468 00:41:45,168 --> 00:41:49,200 industry, they tend to have more strict return to office requirements. 469 00:41:49,200 --> 00:41:55,924 And so I do think firms take a cue from their clients as they should in focusing all of these things. 470 00:41:55,924 --> 00:41:59,516 And so my guess is that these firms that planted a stake in the ground, 471 00:41:59,730 --> 00:42:00,762 They know what they're doing. 472 00:42:00,762 --> 00:42:02,094 They're not doing it by accident. 473 00:42:02,094 --> 00:42:04,878 They are very, very thoughtful about all this stuff. 474 00:42:04,878 --> 00:42:08,322 yeah, I guess, you know, we'll see if that ends up being the right decision for them. 475 00:42:08,332 --> 00:42:09,222 Yeah. 476 00:42:09,363 --> 00:42:09,753 Yeah. 477 00:42:09,753 --> 00:42:11,643 No, that's, that's, that's a good point. 478 00:42:11,643 --> 00:42:25,790 Well, what other change, like you mentioned, um, the opportunity costs challenge with firms trying to manage this training and sorting of data and all that sort of stuff using 479 00:42:25,790 --> 00:42:27,751 kind of in-house legal talent for that. 480 00:42:27,751 --> 00:42:30,742 Um, obviously the opportunity cost is a big one. 481 00:42:30,742 --> 00:42:35,554 Are there other challenges that they might not be thinking about in that approach? 482 00:42:38,127 --> 00:42:49,295 You know, I think that, and this is a broader comment, historically, legal industry has operated in a very simple way. 483 00:42:49,295 --> 00:42:56,720 It's all full-time jobs, and it's mostly at law firms or in-house legal departments, maybe government too, or nonprofit. 484 00:42:56,720 --> 00:42:58,651 But generally, it's all full-time roles. 485 00:42:58,651 --> 00:43:04,905 And I think what we're seeing, given the trends around the pandemic, 486 00:43:05,928 --> 00:43:13,014 the emergence of Zoom, the adoption of cloud tools means that you don't need to be in the same office as someone else. 487 00:43:13,014 --> 00:43:19,150 And what that's done is that it's led to creative approaches to sourcing talent. 488 00:43:19,150 --> 00:43:24,104 Now that sourcing that talent, mean, you we work with a lot of in-house departments just to fill their roles. 489 00:43:24,104 --> 00:43:27,017 If they can't get a full-time hire, they leverage a contractor. 490 00:43:27,017 --> 00:43:29,239 That's been kind of what they've done for many years. 491 00:43:29,239 --> 00:43:33,462 But firms are now starting to see that there is an equally 492 00:43:33,726 --> 00:43:44,736 high caliber talent bench that exists in smaller cities and low cost of living areas where you have people with big law credentials who may not want to work a traditional, you know, 493 00:43:44,736 --> 00:43:46,717 law firm job in a big city. 494 00:43:46,717 --> 00:43:55,425 What that means for the firm is that there's a lot of work that needs to be done and you can rely on this separate talent pool to get it done. 495 00:43:55,425 --> 00:44:01,958 So you can, you know, what we've seen is like firms are leveraging contractors who have big law credentials and experience. 496 00:44:01,958 --> 00:44:10,741 often 15, 20 years out of law school at a rate that's surprisingly affordable or fits into the model of the firm. 497 00:44:10,822 --> 00:44:23,687 so yes, they can be there to help with AI projects and KM projects, but we've also seen it be helpful for firms to deploy these contractors on practice areas that have sporadic 498 00:44:23,687 --> 00:44:31,010 demand, like ups and downs, deal flow, like maybe M &A deals that spike up sometimes and then kind of go away at other times. 499 00:44:31,538 --> 00:44:38,942 Innovation does include AI and technology, it's also, we're seeing an impact on the human capital side of it, the talent management side of it. 500 00:44:38,942 --> 00:44:42,604 So that's kind of something that's cool that I've seen just being part of Latitude. 501 00:44:42,604 --> 00:44:46,346 And I expect to see more of that in the coming months and years. 502 00:44:46,540 --> 00:44:46,790 Yeah. 503 00:44:46,790 --> 00:44:50,551 I mean, that's a model that the tech world has leveraged forever. 504 00:44:50,551 --> 00:44:58,233 I I started offshoring projects in 2006 on ODesk, which is now Upwork. 505 00:44:58,734 --> 00:45:02,434 But that was the marketplace for international tech talent. 506 00:45:02,434 --> 00:45:07,556 And I was an early adopter and had a lot of success. 507 00:45:07,656 --> 00:45:11,137 We have leveraged a lot of different geographies. 508 00:45:11,137 --> 00:45:13,448 We did a lot in the Ukraine for many years. 509 00:45:13,448 --> 00:45:14,828 I had a team of 510 00:45:15,456 --> 00:45:29,922 eight or nine folks over there in 2014 when the whole Crimea situation happened and we had one client who supporting most of that who got nervous and we had to pull and we went to 511 00:45:29,922 --> 00:45:34,810 other areas and now we have folks in South Africa. 512 00:45:34,810 --> 00:45:37,876 I have one in Bolivia. 513 00:45:37,876 --> 00:45:39,916 I've got someone in the UK. 514 00:45:41,717 --> 00:45:44,058 I've got somebody in Canada. 515 00:45:44,396 --> 00:45:47,928 So it's a model that's proven and it works. 516 00:45:47,928 --> 00:45:51,240 And I don't know how we would get by without it. 517 00:45:51,240 --> 00:45:57,474 So it sounds like you're providing a very important service for these firms. 518 00:45:57,918 --> 00:45:58,288 Yeah. 519 00:45:58,288 --> 00:46:04,241 And I think, remember we talked about how the legal world is, you know, a couple of steps behind the rest of the business world. 520 00:46:04,241 --> 00:46:06,642 You know, we can call it five years, we can call it 15 years. 521 00:46:06,642 --> 00:46:11,344 If you've been using it since 2006, then you're probably 20 years ahead of some of our clients. 522 00:46:11,344 --> 00:46:21,948 But tying back to our original point, think trying something new for firms is challenging because of that whole brand reputation they've got to protect. 523 00:46:21,948 --> 00:46:23,969 They really value the relationship with clients. 524 00:46:23,969 --> 00:46:26,930 And so that's why they've been a little bit slower on it. 525 00:46:26,930 --> 00:46:37,117 And that's part of the reason why here at Latitude, what we do is we try to hire people who come from the industry, who have been big law partners or GCs who have contacts in the 526 00:46:37,117 --> 00:46:38,238 community. 527 00:46:38,858 --> 00:46:45,483 That enables us to build that trust for them to essentially try something new and different. 528 00:46:45,483 --> 00:46:48,755 And I know that that's a really hard thing to do. 529 00:46:48,755 --> 00:46:55,666 And that's the job I have, that's the job you have of convincing firms to do something differently that actually has a lot of benefits. 530 00:46:55,666 --> 00:47:09,347 It often is a hard slog to get there, but once you build that trust with them, I think it can lead to some of the most valuable and meaningful partnerships in the entire business 531 00:47:09,347 --> 00:47:10,557 world. 532 00:47:10,678 --> 00:47:17,003 I think that's something that I've come to realize over my time in innovation. 533 00:47:17,003 --> 00:47:21,986 You really need to build that trust before you've earned the right to say, here's a way to do it differently. 534 00:47:22,606 --> 00:47:23,956 Yeah, that's really good stuff. 535 00:47:23,956 --> 00:47:25,886 Well, this is, this has been a lot of fun. 536 00:47:25,886 --> 00:47:32,386 I, I appreciate you coming in and spending a few minutes and, I know you're, you're, busy. 537 00:47:32,386 --> 00:47:35,926 The CRO role is, is not, not an easy one. 538 00:47:35,926 --> 00:47:41,566 And you know, we talked about before we kicked off here, this is planning season for 2025. 539 00:47:41,566 --> 00:47:47,566 So we just had our offsite this week and I know you, you guys are, are working on your plans too. 540 00:47:47,566 --> 00:47:52,332 How, how do you, how do folks find out more about the, the, 541 00:47:52,332 --> 00:47:57,214 the writing that you do and, you your, your newsletter and whatnot. 542 00:47:57,214 --> 00:48:00,978 Yeah, my newsletter is at alexofftherecord.com. 543 00:48:00,978 --> 00:48:03,721 You can also find me on LinkedIn and there's links to it there. 544 00:48:04,202 --> 00:48:07,115 yeah, I'm very active on social, especially LinkedIn. 545 00:48:07,115 --> 00:48:12,331 So if you haven't checked out my profile or seen the stuff I've written, please do. 546 00:48:12,331 --> 00:48:15,194 And yeah, I'm always open to connecting. 547 00:48:15,328 --> 00:48:15,731 Awesome. 548 00:48:15,731 --> 00:48:16,155 Good stuff. 549 00:48:16,155 --> 00:48:18,611 And we'll include some links in the show notes as well. 550 00:48:19,400 --> 00:48:19,993 Thank you, Ted. 551 00:48:19,993 --> 00:48:22,184 And hey, thanks for having me on the show. 552 00:48:22,184 --> 00:48:23,756 Really enjoyed this conversation. 553 00:48:23,756 --> 00:48:24,857 Yeah, absolutely. 554 00:48:24,857 --> 00:48:31,542 I appreciate you spending the time and I'm sure I'll see you at one of these legal conferences. 555 00:48:31,542 --> 00:48:35,935 I go to about a dozen, so we'll have to connect in person while we're there. 556 00:48:37,137 --> 00:48:37,857 All right. 557 00:48:37,857 --> 00:48:39,077 Have a good one. 558 00:48:39,899 --> 00:48:40,269 All right. 559 00:48:40,269 --> 00:48:41,299 Thanks, Alex. 00:00:04,499 Alex Su, how are you this morning? 2 00:00:04,808 --> 00:00:05,536 I'm doing great. 3 00:00:05,536 --> 00:00:07,011 Thanks for having me on the show. 4 00:00:07,011 --> 00:00:09,393 Yeah, I really appreciate you joining, man. 5 00:00:09,393 --> 00:00:14,216 I read some of your stuff and you've got great videos on LinkedIn. 6 00:00:14,216 --> 00:00:18,548 I've been a fan, so it's awesome to have you on. 7 00:00:18,890 --> 00:00:24,497 for Thanks for following along and yeah, I post a lot of crazy stuff, occasionally serious stuff. 8 00:00:24,497 --> 00:00:26,988 So love to talk about it more. 9 00:00:26,988 --> 00:00:28,319 Yeah, well good stuff. 10 00:00:28,319 --> 00:00:32,694 So you started off as an attorney yourself, right? 11 00:00:32,694 --> 00:00:36,959 I think at Sullivan and Cromwell, you were in eDiscovery for a while. 12 00:00:36,959 --> 00:00:40,232 You're now a CRO at Latitude Legal. 13 00:00:40,232 --> 00:00:42,675 Why don't you give us just expand on that a little bit. 14 00:00:42,675 --> 00:00:45,928 Tell us who you are, what you do and where you do it. 15 00:00:46,652 --> 00:00:47,262 Absolutely. 16 00:00:47,262 --> 00:00:50,864 So I am the chief revenue officer of Latitude Legal. 17 00:00:50,864 --> 00:00:56,086 We provide contract attorneys to law firms and legal departments. 18 00:00:56,446 --> 00:01:00,208 I've spent my entire career focused on legal innovation. 19 00:01:00,208 --> 00:01:05,570 I practiced law for six years, clerked for a federal judge, worked at Sullivan Cromwell. 20 00:01:05,590 --> 00:01:08,451 And then about six years out, I decided to make a change. 21 00:01:08,451 --> 00:01:12,867 I found myself in California and saw that the tech startup scene was really bustling. 22 00:01:12,867 --> 00:01:14,590 so 23 00:01:14,590 --> 00:01:23,077 You know, for a lot of reasons, I took a pivot, became a salesperson for selling e-discovery software, then moved to the contract management space, doing sales, then 24 00:01:23,077 --> 00:01:24,117 marketing. 25 00:01:24,158 --> 00:01:30,673 And so earlier this year, I found myself at Latitude, now leading sales and marketing as the chief revenue officer. 26 00:01:30,673 --> 00:01:38,550 Uh, and it's been super interesting because, you know, through my 14 or so years of experience, I've seen the practice of law evolve. 27 00:01:38,550 --> 00:01:43,914 I've seen software services, alternative legal services providers, all sorts of things. 28 00:01:44,075 --> 00:01:47,568 um, make their way into our profession. 29 00:01:47,568 --> 00:01:51,272 And so, um, yeah, it's been, it's been great to see. 30 00:01:51,272 --> 00:01:55,386 And so I write about some of these topics in my newsletter off the record. 31 00:01:55,386 --> 00:02:03,664 Um, you know, very often also writing about AI, which I think is how, you know, you and I got connected and, it's some of the stuff that we're going to talk about today. 32 00:02:03,682 --> 00:02:04,502 Yeah. 33 00:02:04,502 --> 00:02:04,832 Yeah. 34 00:02:04,832 --> 00:02:09,854 Your, uh, your off the record newsletter was actually what grabbed my attention. 35 00:02:09,854 --> 00:02:19,806 You had a, you had an article, God, I guess it was maybe a month ago about advice for, for legal tech startups or legal AI startups. 36 00:02:19,806 --> 00:02:22,567 And I thought there was some really good nuggets in there. 37 00:02:22,567 --> 00:02:32,610 And I think, I think we're fairly aligned in our perspectives on how difficult it is to disrupt the legal services market. 38 00:02:32,610 --> 00:02:33,370 Um, 39 00:02:33,710 --> 00:02:41,270 You know, we, we started off as a company called Acrowire back in 2008, just fell into the legal rabbit hole. 40 00:02:41,690 --> 00:02:51,650 And you know, when cloud really started to take off in the enterprise, it seemed like an absolute no brainer. 41 00:02:51,650 --> 00:02:54,570 Law firms were, were, were going to head that direction. 42 00:02:54,570 --> 00:03:03,974 I'm going back to like 2010, 2011, and I grossly overestimated how quickly that 43 00:03:04,206 --> 00:03:09,826 migration would take and wasted a lot of time and energy. 44 00:03:09,826 --> 00:03:23,686 I did 22, 24 road shows in 2014 out there really highlighting the advantages of Microsoft 365 and SharePoint Online. 45 00:03:23,686 --> 00:03:32,224 And it would be a solid eight, nine plus years before big law would transition in earnest. 46 00:03:32,270 --> 00:03:33,831 to Microsoft 365. 47 00:03:33,831 --> 00:03:35,053 It's finally happening now. 48 00:03:35,053 --> 00:03:36,854 AmLaw is moving there now. 49 00:03:36,854 --> 00:03:51,578 There's only a handful of firms that are still on-prem in that suite, but man, it's, know, current state thinking and, you know, current models are pretty well entrenched. 50 00:03:51,578 --> 00:03:56,923 It's not easy to move in a new direction in legal tech, it seems. 51 00:03:56,923 --> 00:03:58,654 Is that your perspective as well? 52 00:03:59,146 --> 00:04:01,166 Yeah, no, it's not easy at all. 53 00:04:01,166 --> 00:04:03,566 My own journey began in 2016. 54 00:04:03,566 --> 00:04:10,446 So a few years after that, we were selling e-discovery software that was cloud-based to law firms. 55 00:04:10,446 --> 00:04:13,776 And the firms would say, they would have all sorts of objections. 56 00:04:13,776 --> 00:04:16,106 And I remember it was what, 2016. 57 00:04:16,106 --> 00:04:21,646 Five years before that, I had started using Dropbox for file sharing. 58 00:04:21,646 --> 00:04:24,076 And of course, firms are different, right? 59 00:04:24,076 --> 00:04:26,666 There's highly sensitive documents they've got to deal with. 60 00:04:26,666 --> 00:04:28,694 So it's not the same thing, but... 61 00:04:28,694 --> 00:04:31,296 to your point, there's a bit of a lag. 62 00:04:31,296 --> 00:04:40,182 And so I think it's extremely topical these days because everyone keeps talking about AI, the transformation it's having on other industries. 63 00:04:40,442 --> 00:04:48,898 so naturally the conversation has shifted now to, lawyers do things in a very antiquated way. 64 00:04:48,898 --> 00:04:51,660 Maybe AI can completely transform that. 65 00:04:51,660 --> 00:04:55,443 And that's why you have articles coming out that say, will AI replace lawyers? 66 00:04:55,443 --> 00:04:57,714 Like you see them every few months. 67 00:04:57,736 --> 00:05:05,990 So part of what I was trying to write the article, the reason for writing the article is to share kind of with the broader world that legal is different. 68 00:05:06,171 --> 00:05:18,017 And specifically, our ecosystem is built very differently than other industries and specifically firms, law firms, especially the largest ones, occupy a very unique space 69 00:05:18,117 --> 00:05:26,181 that has a lot of nuance that makes it both challenging to disrupt, but also creates 70 00:05:26,684 --> 00:05:27,905 unexpected opportunities. 71 00:05:27,905 --> 00:05:39,904 And so I was trying to share a little bit about that because I just started seeing a lot of tweets, honestly, a lot of posts on social media about how AI is going to just replace 72 00:05:39,904 --> 00:05:40,335 law firms. 73 00:05:40,335 --> 00:05:42,616 And I thought that was really misinformed. 74 00:05:42,680 --> 00:05:44,691 Yeah, it is. 75 00:05:45,052 --> 00:06:00,425 you see, especially lately, Harvey is an example where you've got a lot of investors like Google Ventures led their series C round that really don't have a lot of experience in 76 00:06:00,425 --> 00:06:04,629 legal. 77 00:06:04,629 --> 00:06:10,518 I think there's probably, especially the early indicators with with Gen.ai, what 78 00:06:10,518 --> 00:06:19,843 its capabilities were the Goldman report, um, you know, the, the, that, you know, three, five scored in the 65 percentile on the bar. 79 00:06:19,843 --> 00:06:25,766 And initially it was reported that four Oh scored in the 90th percentile. 80 00:06:25,766 --> 00:06:39,374 That's been, that's been, I guess, adjusted downward quite a bit, but in the beginning, there was a lot of reason for outsiders to the initial data points really kind of pointed. 81 00:06:39,630 --> 00:06:43,843 trajectory that was extremely steep on the transformation curve. 82 00:06:43,843 --> 00:06:51,478 And the reality is, in my opinion, things have been grossly overestimated in the short term. 83 00:06:51,478 --> 00:06:58,773 AI is absolutely going to be transformative in the legal industry, but it's going to take a little while. 84 00:06:58,773 --> 00:07:09,120 I'm not sure what the perspective is with investors who aren't used to disrupting inside of legal if their timeline 85 00:07:09,292 --> 00:07:11,225 how well that is aligned with reality. 86 00:07:11,225 --> 00:07:12,691 Do you have any sense of that? 87 00:07:13,224 --> 00:07:15,595 Yeah, it's totally not in line with reality. 88 00:07:15,595 --> 00:07:21,978 I think, you know, part of me has a lot of empathy for why the investors are so interested in legal. 89 00:07:21,978 --> 00:07:34,313 You look at a market where you see firms charging thousands of dollars an hour, handling matters and doing workflows that are really inefficient, not really leveraging software. 90 00:07:34,313 --> 00:07:39,045 So naturally investors are going to think, well, we can disrupt that with tech. 91 00:07:40,626 --> 00:07:42,687 I think that does happen in many industries. 92 00:07:42,687 --> 00:07:51,293 And I think it's been slower in legal because there's a fundamental misunderstanding of why clients pay so much money to firms. 93 00:07:51,294 --> 00:08:03,141 is tempting to say, Hey, if you're paying an outside firm $2,000 an hour, that somehow, um, a streamlined AI or tech solution can do it at a fraction of the price. 94 00:08:03,182 --> 00:08:08,245 And so, yeah, all the clients will kind of vote with their feet and go to the AI solution. 95 00:08:08,245 --> 00:08:09,460 But the truth is. 96 00:08:09,460 --> 00:08:13,833 firms offer a lot more than just the brief or memo that you see. 97 00:08:14,674 --> 00:08:20,308 What you're really paying for as a client is a bucket of things that are really valuable. 98 00:08:20,308 --> 00:08:27,583 So for example, I've written about the trust and long-term branding that firms care about. 99 00:08:27,583 --> 00:08:37,322 when you work with a firm, you know that the firm has spent decades, if not centuries, building up a brand of client satisfaction, keeping their clients happy, doing good work. 100 00:08:37,322 --> 00:08:43,682 And so you're also trusting that the firm is going to be reliable in what it produces. 101 00:08:44,162 --> 00:08:54,082 AI, for example, if you're replacing that firm with a startup, maybe that's a series A startup that's just been around for a year or two. 102 00:08:54,362 --> 00:09:02,202 so maybe the work that it does is 80 % of the way there, which it could very well be, but that last 20 % matters a lot. 103 00:09:02,202 --> 00:09:06,862 And so it's really hard to tell from a client's perspective if they're getting 104 00:09:07,018 --> 00:09:12,458 really the high quality work from a startup that they would from a law firm. 105 00:09:12,458 --> 00:09:18,858 And it turns out that that 20 % or whatever, 10%, whatever you want to call it, that little bit is so important. 106 00:09:21,278 --> 00:09:27,478 It's because legal is not in the business of something like sales or marketing, where if you roughly get it right, it's OK. 107 00:09:27,478 --> 00:09:32,358 If you get it wrong just a little bit in legal, it has huge consequences. 108 00:09:32,358 --> 00:09:34,416 So I think that's fundamentally why 109 00:09:34,416 --> 00:09:40,315 legal work is different than say work done by other functions or other industries. 110 00:09:40,482 --> 00:09:51,377 Yeah, and you had a really good point about why high hourly rates aren't nearly the problem for clients that they would seem to be. 111 00:09:51,458 --> 00:09:53,900 Let's drill into that a little bit. 112 00:09:55,944 --> 00:09:57,896 expand on that thought. 113 00:09:58,346 --> 00:10:06,465 Yeah, so when I first started off as an associate at Sullivan and Cromwell, I heard stories of how much we were charging clients and how much we were being billed out on an 114 00:10:06,465 --> 00:10:07,056 hourly basis. 115 00:10:07,056 --> 00:10:09,436 And I was like this like fresh associate. 116 00:10:09,436 --> 00:10:10,836 You know, I didn't really have that much expertise. 117 00:10:10,836 --> 00:10:12,666 I knew that I just worked hard, right? 118 00:10:12,666 --> 00:10:18,366 And I'd like to think I was smart, but I, it just, it just confused me why these clients would pay so much. 119 00:10:18,366 --> 00:10:27,486 And so it took me a while, but I kind of zoomed out and I realized, okay, well, if you add up my hours and the hours of other members of my team and all the way up to the partners, 120 00:10:27,486 --> 00:10:33,089 and you kind of come up with a number every month, let's call it like 5 million a month. 121 00:10:33,089 --> 00:10:42,474 That's a pretty significant legal bill on an absolute scale, but compared to the deals and cases that we were working, these were billion dollar matters. 122 00:10:42,474 --> 00:10:55,732 so yes, the hourly rate was high, but when you kind of compared the total legal fees generated and compared it to the matter itself, it was such a small percentage that it 123 00:10:55,732 --> 00:10:57,082 made me realize that 124 00:10:57,110 --> 00:10:59,661 Um, you know, there's more than meets the eye. 125 00:10:59,661 --> 00:11:02,272 You can't just look at the hourly rate in a vacuum. 126 00:11:02,272 --> 00:11:11,226 Yes, that would be an insane hourly rate to charge for an individual who had say, you know, a trust in a state's matter, or I don't know, like a real estate transaction, like 127 00:11:11,226 --> 00:11:21,480 you're buying a house, but on the corporate side, um, the numbers get so big that, that legal fees, um, are often just a drop in the bucket. 128 00:11:21,480 --> 00:11:23,621 the re really hit home for me. 129 00:11:23,621 --> 00:11:25,142 I didn't really realize that until much later. 130 00:11:25,142 --> 00:11:25,982 I didn't, 131 00:11:26,106 --> 00:11:31,007 The lesson didn't hit home for me until much later when I realized that hourly rates kept on going up. 132 00:11:31,127 --> 00:11:33,448 If they were so high, why would they continue going up? 133 00:11:33,448 --> 00:11:35,249 Why wouldn't clients leave? 134 00:11:35,249 --> 00:11:45,521 And it made me also understand that there are a lot of matters that law firms handle, especially the AMLaw firms, the largest firms in the world handle, that they're getting a 135 00:11:45,521 --> 00:11:54,558 great deal from the law firm because the clients are getting a great deal because if they're willing to pay so much and they're not willing to leave, 136 00:11:54,558 --> 00:11:56,369 that it must be that they're getting something out of it. 137 00:11:56,369 --> 00:12:01,803 I think that was hard for me to understand and grasp because I was using my own frame of reference for the numbers. 138 00:12:01,803 --> 00:12:05,776 And I think a lot of people fall into this trap, especially those in the tech space. 139 00:12:05,776 --> 00:12:09,849 Like they think $1,000 an hour for a junior associate is so much. 140 00:12:09,849 --> 00:12:13,452 when you zoom out, really, the clients are the ones that are voting with their feet. 141 00:12:13,452 --> 00:12:16,434 And so they are continuing to pay these rates and the rates keep going up. 142 00:12:16,434 --> 00:12:22,308 So there must be something there that we're not seeing on a surface level. 143 00:12:22,796 --> 00:12:23,016 Yeah. 144 00:12:23,016 --> 00:12:36,599 And you also talked a little bit about how the, you know, decades old or century plus old reputation of the law firms inhibit disruption, right? 145 00:12:36,599 --> 00:12:38,421 Like talk, talk a little bit about that. 146 00:12:38,421 --> 00:12:44,276 You kind of touched on it high level, but how specifically does that come into play? 147 00:12:44,862 --> 00:12:49,603 Yeah, so I think a good example to compare it to is a startup sales team. 148 00:12:49,703 --> 00:12:56,365 For a startup, you need revenue or else you're not gonna succeed. 149 00:12:56,365 --> 00:13:05,768 And so there's a huge premium placed on experimentation, trying new technology, doing things more efficiently, because for sales, you just need to be like 50 % right. 150 00:13:05,968 --> 00:13:10,329 Whether it's email automation that's sent out that may have typos in it, doesn't really matter. 151 00:13:10,329 --> 00:13:13,126 As long as the emails are going out, you're gonna generate sales. 152 00:13:13,126 --> 00:13:15,707 And so it's a perfect fit. 153 00:13:15,808 --> 00:13:18,690 Law firms on the other hand, can't do the same thing. 154 00:13:18,931 --> 00:13:28,519 They've been around for decades and centuries, as I mentioned, and they often get their business because they are trusted advisors to the business. 155 00:13:28,519 --> 00:13:34,624 They have longstanding strong reputations individually in the legal community. 156 00:13:34,624 --> 00:13:42,850 And so people know that if I use this lawyer or this firm, I can be sure that I'm going to get a quality work product out of them. 157 00:13:43,710 --> 00:13:51,552 that type of business does not lend itself so well to technology that gets it 50 % right or 80 % right. 158 00:13:51,552 --> 00:14:04,816 I mentioned this earlier, but the bar for firms is much higher when it comes to the reliability of the technology than it is for most other businesses. 159 00:14:05,076 --> 00:14:13,758 And if you look at kind of how technology has made its way into firms, you can see, right, very often functionality and features 160 00:14:13,780 --> 00:14:18,412 play a secondary role to things like data security or reliability. 161 00:14:18,813 --> 00:14:24,656 The firms want to make sure that what they're using in terms of software or technology is reliable. 162 00:14:24,656 --> 00:14:34,951 And that's because they have this duty to their clients to make sure that they're not passing on or relying on technology that's not reliable, that leads to bad advice, that 163 00:14:34,951 --> 00:14:39,184 would blow up their reputation that they've taken decades and centuries to build. 164 00:14:39,184 --> 00:14:40,745 that's what makes law firms very different. 165 00:14:40,745 --> 00:14:43,586 And there's a few other ways that law firms 166 00:14:44,270 --> 00:14:46,703 are different and where they sit in the entire ecosystem. 167 00:14:46,703 --> 00:14:50,287 But I would say that is why the bar is so high for firms. 168 00:14:50,287 --> 00:15:01,310 And that's why software companies often find themselves paying more attention to making sure they have data security and reliability versus the latest and greatest features. 169 00:15:01,708 --> 00:15:01,978 Yeah. 170 00:15:01,978 --> 00:15:08,521 And you know, another way I think, reputation law firm reputation, it affects the dynamics of the matter itself. 171 00:15:08,521 --> 00:15:19,566 If you know, a one side comes to the table with, you know, do we cheat them and how down the street and the other side is Kirkland and Ellis, you know, you've got that, that 172 00:15:19,566 --> 00:15:25,589 affects the dynamics before any legal, um, actions take place. 173 00:15:25,589 --> 00:15:28,930 Just the weight of the reputation of the firm. 174 00:15:29,240 --> 00:15:41,960 can ultimately impact the outcome of the case because the other side understands the capabilities, the reputation, the success rate, the relationships, the expertise, all of 175 00:15:41,960 --> 00:15:46,163 those things that a law firm brand brings to the table. 176 00:15:46,163 --> 00:15:49,826 And that is very difficult to replicate as well. 177 00:15:49,826 --> 00:15:50,622 Agreed? 178 00:15:50,622 --> 00:15:55,185 yeah, there's this like really famous saying like nobody ever got fired for using IBM. 179 00:15:55,605 --> 00:15:59,417 I think that, you know, it says a lot of things. 180 00:15:59,417 --> 00:16:03,589 It's saying that if you rely on a brand name vendor, you're not going to go wrong. 181 00:16:03,830 --> 00:16:13,735 That can be overstated and overdone in many parts of the world, but in legal, you know, you can replace IBM with the name of, you know, some of these most well-known firms like 182 00:16:13,735 --> 00:16:14,746 Cravath as an example. 183 00:16:14,746 --> 00:16:18,818 I think I've heard nobody ever got fired for going with Cravath. 184 00:16:18,818 --> 00:16:19,678 Now, 185 00:16:20,020 --> 00:16:30,605 part of the reason is because as you say, hiring a Cravath or a Kirkland and Ellis or whoever, that does send a message to the opposing counsel that you're, you you mean 186 00:16:30,605 --> 00:16:36,317 business, you're ready to spend to win the case, but it also sends a message to your internal stakeholders. 187 00:16:36,317 --> 00:16:47,374 So for example, if you're the chief legal officer of a publicly traded company and you, you know, retain Cravath or Kirkland or whoever, you know, that creates this level of 188 00:16:47,374 --> 00:16:52,055 It sends a message to your board because the board is likely paying attention to this litigation or matter. 189 00:16:52,676 --> 00:17:06,040 And the flip side of that is that if you do rely on some other firm with a lighter reputation or a firm that no one's really heard of, that can impact the perception of your 190 00:17:06,040 --> 00:17:08,960 competence as a CLO, as a chief legal officer. 191 00:17:08,960 --> 00:17:12,021 so legal departments are very savvy about this. 192 00:17:12,021 --> 00:17:14,652 They know that for the most important matters, 193 00:17:14,866 --> 00:17:21,530 which firm they choose sends a message to their stakeholders, especially their board and to their leadership, like their CEO. 194 00:17:22,090 --> 00:17:28,554 that's why it's not as simple as let's go with the cheaper option. 195 00:17:28,914 --> 00:17:35,678 Again, the clients are getting something, a lot of things out of spending that premium for more expensive firms. 196 00:17:35,810 --> 00:17:37,591 Yeah, absolutely. 197 00:17:37,591 --> 00:17:50,174 Well, switching gears a little bit, talking about, uh, you and I had a great conversation about AI and its transformative effect, its impact on, on jobs in the industry. 198 00:17:50,174 --> 00:18:03,497 And, know, I'm a firm believer there's, there's a fair number of analogs that we can look at to gauge what happens when game changing technology enters the marketplace and drives 199 00:18:03,497 --> 00:18:05,774 efficiency significantly higher. 200 00:18:05,774 --> 00:18:10,074 A good example is spreadsheets in accounting. 201 00:18:10,074 --> 00:18:17,874 So when Lotus one, two, three came out sometime in the early eighties, I think it was 1983, 1984, I went back and looked at a little research. 202 00:18:17,874 --> 00:18:24,334 There were about 850,000 professionals in the accounting industry at that time. 203 00:18:24,334 --> 00:18:28,734 I, and I happen to remember, this is when I was really getting into computers. 204 00:18:28,734 --> 00:18:32,034 I was what 11, 12 years old at the time. 205 00:18:32,034 --> 00:18:34,772 And, um, I remember 206 00:18:34,958 --> 00:18:40,858 My mom was in the restaurant business and she had an accountant who was completely had no, he didn't have a computer in his office. 207 00:18:40,858 --> 00:18:50,158 And I was in the back of the restaurant working on, uh, that my mom owned and he was asking me questions and it didn't occur to me until much, much later why he was, he was 208 00:18:50,158 --> 00:18:50,418 worried. 209 00:18:50,418 --> 00:18:51,818 He was concerned. 210 00:18:51,818 --> 00:19:01,658 Um, and what has happened in that 20 or 30 years since the number of professionals in the industry has almost doubled. 211 00:19:01,658 --> 00:19:02,238 Right? 212 00:19:02,238 --> 00:19:03,270 So they're, 213 00:19:03,350 --> 00:19:07,072 there's been a huge net gain of jobs, not a net loss. 214 00:19:07,392 --> 00:19:10,892 There's so much new legislation coming out constantly. 215 00:19:10,892 --> 00:19:13,595 I think there will be no legal jobs lost. 216 00:19:13,595 --> 00:19:21,680 I think there will be a ton of impact and redistribution of roles as a result of this technology. 217 00:19:21,680 --> 00:19:26,742 But I don't know, what is your take on how AI is going to impact the job market itself? 218 00:19:27,206 --> 00:19:29,869 I think it's going to create more demand for legal services. 219 00:19:29,869 --> 00:19:37,255 We're starting to see that with litigation, also regulation coming out around AI. 220 00:19:37,255 --> 00:19:39,247 Every company is going to soon become an AI company. 221 00:19:39,247 --> 00:19:45,813 And so that's going to create a lot of impact on what type of legal advice is needed. 222 00:19:45,813 --> 00:19:53,909 And I generally believe something similar to what you believe, which is this technology creates new roles. 223 00:19:54,378 --> 00:20:03,923 commentators and most of us, when we look at technology, we think, this is going to replace the way many of these people who have these jobs right now. 224 00:20:03,923 --> 00:20:06,524 So inevitably it's going to lead to job loss. 225 00:20:07,666 --> 00:20:14,789 For legal in particular, I do think that we underestimate how much complexity technology adds. 226 00:20:15,630 --> 00:20:18,891 And that complexity needs to be solved by someone. 227 00:20:19,032 --> 00:20:20,593 New problems are being created every day. 228 00:20:20,593 --> 00:20:23,384 Like I never could have imagined that what impact 229 00:20:23,454 --> 00:20:26,176 AI would have on copyright and IP. 230 00:20:26,176 --> 00:20:30,779 There's going to be a lot more demand for that practice area. 231 00:20:30,779 --> 00:20:39,684 So there's going to be a lot of things like that where we don't expect to know where increased demand of services come from. 232 00:20:39,684 --> 00:20:49,640 I can't imagine a world where AI replaces jobs and then doesn't create new ones through the increasing complexity of laws and regulation and lawsuits. 233 00:20:49,848 --> 00:20:50,188 Yeah. 234 00:20:50,188 --> 00:20:54,620 And how much unmet need is there in the legal world? 235 00:20:54,620 --> 00:21:08,546 mean, you and I talked about startups and you know, uh, we are, uh, we have a very small investment from a venture fund, so we're largely bootstrapped and you know, we don't have 236 00:21:08,706 --> 00:21:15,149 tens of millions of dollars in our war chest that we can deploy, um, inefficiently. 237 00:21:15,149 --> 00:21:18,750 have to spend our dollars very carefully and we do. 238 00:21:18,958 --> 00:21:27,898 And as a result, you know, we, we negotiate, uh, we've been doing this long enough that, um, we negotiate a lot of red lines on our own. 239 00:21:27,898 --> 00:21:39,598 Um, and you know, maybe we'll have a lawyer look at the final, the final draft before we execute the document, but we don't, we don't send our lawyers in with our clients, lawyers 240 00:21:39,598 --> 00:21:41,038 and let them have at it. 241 00:21:41,038 --> 00:21:42,478 No, we manage a lot of that. 242 00:21:42,478 --> 00:21:49,128 Ideally, I'd love to have a legal counsel manage the whole process, but it's just, it's, it's out of reach. 243 00:21:49,312 --> 00:21:50,873 financially for us to do that. 244 00:21:50,873 --> 00:22:03,806 So, you know, there's been, there's been firms out there like Wilson, Sincini with Neuron, think Cooley go, there's like some platforms out there where they've tried to, um, 245 00:22:03,806 --> 00:22:16,590 automate, uh, some of this, you know, like even I can't remember which, which firm it is, you know, they, they will draft your offer letters, which, know, again, it'd be really 246 00:22:16,590 --> 00:22:17,742 tough to have a 247 00:22:17,742 --> 00:22:19,797 person sit there and do that for free. 248 00:22:19,797 --> 00:22:21,692 you know, I don't know. 249 00:22:21,692 --> 00:22:26,252 How do you see automation increasing access to legal services? 250 00:22:27,294 --> 00:22:29,535 I think more firms will become more in tech enabled. 251 00:22:29,535 --> 00:22:33,997 And what that means is that their cost structure for delivering legal services will change. 252 00:22:34,298 --> 00:22:44,424 You're seeing that with Neuron as an example, like they are leveraging the know-how and expertise of their lawyers to automate the creation of certain types of documents. 253 00:22:44,424 --> 00:22:47,846 Before you would have to have a whole team of lawyers maybe drafting something. 254 00:22:47,846 --> 00:22:49,817 Now it can be replicated at a low cost. 255 00:22:49,817 --> 00:22:53,228 And so that means that they can pass on some of those savings to clients. 256 00:22:53,329 --> 00:22:55,791 So I do think that while... 257 00:22:56,222 --> 00:23:07,605 technology and AI is not going to replace lawyers, it will become infused in the practice such that the firms that lean on innovation or technology will have an advantage. 258 00:23:07,605 --> 00:23:11,966 And that does have broader impact on who gets access to legal services. 259 00:23:11,966 --> 00:23:23,150 But the key takeaway though is that I do think that the firms that have leaned into innovation, whether it's experimenting, trialing software, building their own software, 260 00:23:23,150 --> 00:23:25,310 just really focusing on innovation. 261 00:23:25,372 --> 00:23:34,586 I do think that they'll have an advantage because what then happens is they can offer more to their clients or new clients at a lower price point. 262 00:23:35,186 --> 00:23:48,712 And as I mentioned, it is true that many of these firms can continue to charge high rates, but if they can deliver that at a lower cost and either keep their billable hours or rates 263 00:23:48,712 --> 00:23:52,733 similar or higher, that just means more profits. 264 00:23:52,733 --> 00:23:55,192 so I do think that... 265 00:23:55,192 --> 00:23:57,934 tech enabled is the future. 266 00:23:58,296 --> 00:23:59,166 Yeah. 267 00:23:59,186 --> 00:24:05,230 So what about the AI's current capabilities versus the expectations in the marketplace? 268 00:24:05,230 --> 00:24:07,170 It feels like there's a disconnect. 269 00:24:08,432 --> 00:24:18,437 in my mind, you know, I see, for example, I see firms rolling out technologies like co-pilot, Microsoft co-pilot wholesale, like enterprise wide. 270 00:24:18,437 --> 00:24:25,141 Um, it's happened in a few places and you know, my take on co-pilot is it's not very good. 271 00:24:25,141 --> 00:24:27,950 Um, in fact it is. 272 00:24:27,950 --> 00:24:32,910 very much inferior to chat GPT and Claude. 273 00:24:32,910 --> 00:24:40,610 Even though, you know, Microsoft is a huge investor in open AI and they clearly have access to those, to the foundational model. 274 00:24:40,610 --> 00:24:45,150 I'm not sure exactly what they're doing to make it worse, but it is, it is worse. 275 00:24:45,150 --> 00:24:57,222 As an example, I, you know, I play around with it a lot and I test it, you know, in, in outlook, I went into, pulled up the co-pilot side panel. 276 00:24:57,576 --> 00:24:57,786 it. 277 00:24:57,786 --> 00:25:01,289 How many, how many calendar entries do I have? 278 00:25:01,289 --> 00:25:03,990 I'll block time off for thought leadership. 279 00:25:03,990 --> 00:25:09,554 How many calendar entries do I have on my calendar this week for thought leadership? 280 00:25:09,554 --> 00:25:12,056 Came back and said, you have nine. 281 00:25:12,056 --> 00:25:13,906 I actually had 12. 282 00:25:14,478 --> 00:25:15,498 So it was incorrect. 283 00:25:15,498 --> 00:25:18,981 And I know AI is not great at counting, but hear me out on this. 284 00:25:18,981 --> 00:25:20,822 So then I went back in and I reprompted. 285 00:25:20,822 --> 00:25:23,013 I was like, why did you say I only had nine? 286 00:25:23,013 --> 00:25:25,285 I actually have 15, even though I only had 12. 287 00:25:25,285 --> 00:25:26,075 He goes, you're right. 288 00:25:26,075 --> 00:25:27,546 You do have 15. 289 00:25:28,302 --> 00:25:32,925 So then to further test this, I was like, all right, AI is not great at counting. 290 00:25:32,925 --> 00:25:37,747 Let me, let me see if this is a copilot problem or there's a fundamental LLM problem. 291 00:25:37,747 --> 00:25:44,851 So I took a screenshot and I went to chat GPT and Claude and I uploaded the screenshot with the same prompt. 292 00:25:44,851 --> 00:25:46,331 They both got it right. 293 00:25:46,432 --> 00:25:54,286 So it's like, um, you know, I feel like the marketing and I know Microsoft's going to get copilot right at some, at some point, right? 294 00:25:54,286 --> 00:25:58,388 They're, too invested in this, but right now it's just, it's just not good in my opinion. 295 00:25:58,646 --> 00:26:09,272 You know how how big of a gap do you think there is between the capabilities that are being marketed by people like Microsoft who are trusted in the space and there's this 296 00:26:09,272 --> 00:26:18,517 whole agentic thing which that's we can talk about that later that they're pitching with Copilot Studio but I don't know do you feel like they're in line or do you feel like me 297 00:26:18,517 --> 00:26:20,498 that there's a bit of a disconnect. 298 00:26:21,736 --> 00:26:28,562 I think we're all trying to figure out what AI is capable of and it's hard not to get caught up in the possibility. 299 00:26:29,050 --> 00:26:31,913 I've not had that experience with Copilot. 300 00:26:31,913 --> 00:26:34,634 I have only really played around with ChatGPT. 301 00:26:34,695 --> 00:26:40,029 And so my personal experience really is probably more limited than yours, Ted. 302 00:26:40,029 --> 00:26:53,720 But what I have noticed is that if you can do some really cool stuff with something like ChatGPT, then your mind starts thinking that it could be done for a lot of other broader 303 00:26:53,720 --> 00:26:56,532 use cases that may not be a good fit. 304 00:26:56,532 --> 00:26:58,824 The classic example in legal is 305 00:26:58,824 --> 00:26:59,465 legal research. 306 00:26:59,465 --> 00:27:07,280 You remember like when it first came out, when chat GPT first came out, everyone started using it for legal research and then started hallucinating cases and lawyers got in 307 00:27:07,280 --> 00:27:08,221 trouble. 308 00:27:08,602 --> 00:27:15,645 I think that this tech enthusiasm leads us down the wrong path. 309 00:27:15,887 --> 00:27:25,404 What typically happens, and I'm just analogizing to other non-AI related innovations, is new technology comes out, people get excited about it. 310 00:27:25,404 --> 00:27:32,636 And the initial applications are in fast moving industries or sectors like sales and marketing, maybe customer service. 311 00:27:33,156 --> 00:27:35,974 And then it takes a while before it really makes its way to legal. 312 00:27:35,974 --> 00:27:37,898 I mean, you cloud as an example. 313 00:27:37,898 --> 00:27:45,200 Like we were using Dropbox and Box and things like that, like cloud-based document storage for a long time before firms did. 314 00:27:45,200 --> 00:27:50,302 And probably for good reason, because you kind of want things to be tested out and validated. 315 00:27:51,438 --> 00:28:01,847 We got to a point eventually where I think in 2024 right now, late 2024, adopting the cloud for firms seems not that controversial, but it took a long time. 316 00:28:01,847 --> 00:28:03,578 And so I think that's going to happen with AI too. 317 00:28:03,578 --> 00:28:07,552 Like right now, people are really stoked about it, really excited about it. 318 00:28:07,552 --> 00:28:11,505 And so it's inevitably going to have heightened expectations. 319 00:28:11,505 --> 00:28:19,926 You see the headlines of AI replacing lawyers and tweets that say AI is going to replace law firms, which again is how this all started. 320 00:28:19,926 --> 00:28:29,422 Um, but yeah, I think, I think you're right that it's, you know, we've got to keep our expectations realistic and, let this, this play out, um, before diving headfirst into it. 321 00:28:29,720 --> 00:28:30,151 Yeah. 322 00:28:30,151 --> 00:28:40,999 So along those lines, how should law firms be thinking about ROI when it comes to these GENAI platforms? 323 00:28:40,999 --> 00:28:42,730 Do you have any insight on that? 324 00:28:43,900 --> 00:28:50,093 I have some, I'll probably speak in generalities, but from what I've seen, it's got to tie into with your practice. 325 00:28:50,093 --> 00:29:00,967 Like it can be tempting to see the latest and shiniest new AI tool that could do so much, but if it doesn't fit in with your practice, I think you're setting yourself up for 326 00:29:00,967 --> 00:29:01,538 failure. 327 00:29:01,538 --> 00:29:05,909 And what I mean by that is, know, some firms do rely on outside vendors. 328 00:29:05,909 --> 00:29:08,541 They will bring on new technologies for AI. 329 00:29:08,541 --> 00:29:09,901 Totally makes sense. 330 00:29:09,941 --> 00:29:12,287 Others will create their own AI. 331 00:29:12,287 --> 00:29:16,369 that's tied to their own specific practice area strengths. 332 00:29:16,369 --> 00:29:17,789 That'll work too. 333 00:29:18,149 --> 00:29:29,824 So there's not one clear way to do it, but I do think that part of what's driving this rush to AI from the firm perspective, my sense is that many clients, and when I say 334 00:29:29,824 --> 00:29:39,017 clients, I really mean corporate clients, the legal departments that serve as firm clients are themselves under pressure to use AI. 335 00:29:39,838 --> 00:29:46,700 Boards are telling CEOs leverage AI, CEOs are telling their chief legal officers and GCs use AI. 336 00:29:46,920 --> 00:29:53,762 And so the legal department is also probably very interested in which firms are leveraging AI to generate efficiencies. 337 00:29:53,762 --> 00:29:55,802 Now we don't know if there's truly efficiencies. 338 00:29:55,802 --> 00:30:05,055 I think that the jury is still out on that, but there's no question that there is a broad mandate from the top down that's trickling its way out to law from boards and CEOs to the 339 00:30:05,055 --> 00:30:08,286 outside firms that serve them to use AI. 340 00:30:08,510 --> 00:30:21,667 And what I've seen really, you know, some firms be really smart about it is that instead of just, you know, buying the latest and greatest tool from a startup, which, know, often 341 00:30:21,667 --> 00:30:23,979 can make sense again, like depending on what you need. 342 00:30:23,979 --> 00:30:31,283 But I've seen some firms develop their own AI and, you know, build their own models that are tied to their own proprietary data. 343 00:30:31,343 --> 00:30:37,222 I shouldn't say proprietary data, but like their, their own documents, because it can be hard for a firm to take 344 00:30:37,222 --> 00:30:48,665 know, client matters and, kind of like, um, you know, put it onto some third party startups training algorithm, but if they do it themselves, that that's different. 345 00:30:48,705 --> 00:31:04,189 so, you know, I've personally seen, uh, some firms really leverage the, um, their own data and using AI to kind of create their own process that creates value in ROI. 346 00:31:04,189 --> 00:31:06,792 Um, it's not as exciting and sexy as like, you know, 347 00:31:06,792 --> 00:31:14,148 all the headlines, but that is something I've seen firms really benefit from AI. 348 00:31:14,744 --> 00:31:15,084 Yeah. 349 00:31:15,084 --> 00:31:25,497 So along those lines, you and I spoke about scenarios where there are alternative approaches like to staffing models. 350 00:31:25,497 --> 00:31:32,839 You know, law firms have to get their, their, their, their house in order really to leverage AI effectively. 351 00:31:32,839 --> 00:31:39,881 You know, most, if you're AmLaw 50 or the floor of the AmLaw 50 is around a thousand attorneys, you have a thousand attorneys or more. 352 00:31:39,881 --> 00:31:44,322 You have probably about a minimum of a hundred million documents. 353 00:31:44,566 --> 00:31:49,609 in your DMS and that could be all sorts of content. 354 00:31:49,609 --> 00:31:50,890 could be emails. 355 00:31:50,890 --> 00:31:53,851 could be unused drafts. 356 00:31:53,851 --> 00:32:01,214 It could be red lines, know, 17 versions back from the final product. 357 00:32:01,636 --> 00:32:06,118 And it's a big, it's a heavy lift, right? 358 00:32:06,158 --> 00:32:08,069 You can't manually curate. 359 00:32:08,069 --> 00:32:13,722 You can't build an army of people big enough to manually curate the entire thing. 360 00:32:14,060 --> 00:32:15,521 Like wholesale, right? 361 00:32:15,521 --> 00:32:27,218 You have to kind of strategically attack the problem, but you know, it, sounded like you had seen some, and, and, and I think maybe you guys, that's part of what you do is, um, 362 00:32:27,378 --> 00:32:30,891 around staffing for those types of engagements. 363 00:32:30,891 --> 00:32:33,502 Like tell me, tell me a little bit about that. 364 00:32:33,864 --> 00:32:34,174 Yeah. 365 00:32:34,174 --> 00:32:45,180 So, you we talked about how AI is such a critical part of a firm's competitive advantage or could be, but there's certainly a lot of other technology related initiatives that, 366 00:32:45,180 --> 00:32:49,342 knowledge management or chief innovation officers are well aware of. 367 00:32:49,503 --> 00:32:50,824 Now here's the problem. 368 00:32:50,824 --> 00:32:58,948 Because of the staffing model of firms, you have associates and counsel who are essentially being billed out at very high rates. 369 00:32:58,948 --> 00:33:02,356 And so the opportunity cost of using your own people 370 00:33:02,356 --> 00:33:06,468 to work on these internal strategic projects is very high. 371 00:33:06,568 --> 00:33:18,875 How it plays out is essentially, let's say that your firm's decided to launch a new AI initiative, build its own proprietary AI related to a specific practice area. 372 00:33:19,075 --> 00:33:24,378 And you basically need associates from that group to train the model. 373 00:33:24,718 --> 00:33:28,080 So KM knows this, the chief innovation officer knows this. 374 00:33:28,080 --> 00:33:30,782 And so they ask, hey, who can help train this model? 375 00:33:30,782 --> 00:33:31,870 Problem is, 376 00:33:31,870 --> 00:33:39,015 these associates are so busy, they're not gonna put the time and energy into training that AI. 377 00:33:39,015 --> 00:33:42,678 They're gonna instead work on some matter that the partner assigns them. 378 00:33:42,718 --> 00:33:48,131 And so this is, think, one big challenge that firms are running into that Latitude helps with. 379 00:33:48,131 --> 00:33:58,670 So we've worked with AMLaw firms that are struggling to staff their strategic initiatives around KM, around documents and AI. 380 00:33:59,026 --> 00:34:06,351 And so we provide this associate level, so same caliber as the firm's associates on a contract basis. 381 00:34:06,371 --> 00:34:11,595 And those folks are able to get the work done without that same opportunity cost. 382 00:34:11,595 --> 00:34:23,884 And because we provide these contract attorneys at a fraction of the rate that these existing full-time associates bill out at, it tends to be a financially attractive 383 00:34:23,884 --> 00:34:24,645 proposition. 384 00:34:24,645 --> 00:34:27,016 So I think that... 385 00:34:27,016 --> 00:34:36,900 because of this drive to AI for firms to build their own AI, it's created opportunities for companies like us, like Latitude, where we can provide high-end, high-caliber 386 00:34:36,900 --> 00:34:40,632 attorneys to work on these projects so they don't kind of fall by the wayside. 387 00:34:40,632 --> 00:34:49,555 I think that's the big worry is that, know, KM, knowledge management officers and folks, they know that some of these projects are so important, but it's just a question of how 388 00:34:49,555 --> 00:34:55,668 are they going to get that subject matter expertise to train the technology to work on those projects? 389 00:34:55,668 --> 00:34:56,756 And I think that's where we... 390 00:34:56,756 --> 00:34:57,667 Fill the gap. 391 00:34:58,166 --> 00:35:03,598 And then why is there such a strong interest in building proprietary AI systems? 392 00:35:03,598 --> 00:35:05,868 I mean, I know a lot of firms are doing it. 393 00:35:05,888 --> 00:35:22,593 know Gunderson, I had Naveen and Joe Green on from Gunderson a little over a year ago when ChatGD came out and it was novel at the time, but many firms have done it since then. 394 00:35:23,733 --> 00:35:26,324 Why is there such a strong push in that direction? 395 00:35:26,622 --> 00:35:27,643 I think it's twofold. 396 00:35:27,643 --> 00:35:30,604 First, it's to show clients that they're serious about AI. 397 00:35:30,604 --> 00:35:35,887 And I mentioned earlier, that mandate comes all the way straight from the top of the client's organization. 398 00:35:35,887 --> 00:35:38,729 showing your clients that you're focused on AI helps. 399 00:35:38,949 --> 00:35:45,633 That's probably especially true for a firm like Gunderson or Wilson-Sonsini, these firms that serve the technology industry. 400 00:35:46,514 --> 00:35:48,854 So that's at one level. 401 00:35:49,555 --> 00:35:53,657 The second reason has to do with the cost of delivery. 402 00:35:54,902 --> 00:36:03,726 we don't know if any of these chatbots or proprietary AI or any of these models will actually end up being more efficient. 403 00:36:03,726 --> 00:36:08,207 There's certainly a lot of risks of imperfection, hallucinations, things like that. 404 00:36:08,328 --> 00:36:21,603 But if AI can fulfill its promise, that could enable a firm to essentially provide the same level of service, generate the same amount of revenue and billable hours at a much 405 00:36:21,603 --> 00:36:22,424 lower cost structure. 406 00:36:22,424 --> 00:36:23,732 And again, that goes to the 407 00:36:23,732 --> 00:36:24,533 firms profits. 408 00:36:24,533 --> 00:36:34,399 And so I think the possibility that technology can create higher levels of profit, I think that is a huge opportunity. 409 00:36:34,399 --> 00:36:46,537 And so those two reasons both work together to kind of drive firms towards tech adoption in ways that they haven't before, honestly, with the cloud, the internet, other 410 00:36:46,537 --> 00:36:47,678 innovations. 411 00:36:47,982 --> 00:36:48,662 Yeah. 412 00:36:48,662 --> 00:36:54,606 You know, so we used to be a custom dev shop at, uh, our previous company was called Acrowire. 413 00:36:54,606 --> 00:37:02,470 We kind of rebranded and relaunched this InfoDash, a product company in about three years ago. 414 00:37:02,470 --> 00:37:04,791 Um, and our timing was great. 415 00:37:04,791 --> 00:37:07,132 It's we've had a fabulous run. 416 00:37:07,132 --> 00:37:17,646 Things are, we're having a lot of fun over here, but for many years we would develop solutions for clients and they would 417 00:37:17,646 --> 00:37:21,726 really have challenges during the handoff. 418 00:37:22,006 --> 00:37:30,066 sometimes it was, you know, KM and innovation would work with us to build a solution that ultimately IT had to support. 419 00:37:30,066 --> 00:37:38,626 And maybe IT had one ear in on the project, but didn't fully grasp what was coming their way when it was time to transition. 420 00:37:38,626 --> 00:37:44,686 You know, and much of what I'm talking about here, it was during the on-prem days, right? 421 00:37:44,686 --> 00:37:47,718 So they had to stand up infrastructure sometimes that they weren't 422 00:37:47,906 --> 00:37:54,228 really familiar with or didn't have the capabilities to manage effectively. 423 00:37:54,448 --> 00:38:07,601 And, you know, I find it interesting that so many firms are doing what you're describing because I've seen a lot of struggles, especially in the mid middle market, right? 424 00:38:07,601 --> 00:38:11,623 Like, so let's say, you know, Amla 50 to a hundred, right? 425 00:38:11,623 --> 00:38:17,114 Where they've got the, they've got the resources to write the checks, but they don't have spared dev. 426 00:38:17,766 --> 00:38:21,398 bandwidth sitting around waiting to for things to do right. 427 00:38:21,398 --> 00:38:28,330 If they do have any any dev resources, they're fully deployed and it's probably a fairly small team. 428 00:38:28,411 --> 00:38:43,417 So it seems like it would be a challenge for firms to go down this path of building, even if it's just a wrapper around, you know, an API for a model. 429 00:38:43,697 --> 00:38:46,338 Have you seen firms have success with this approach? 430 00:38:46,804 --> 00:38:47,855 I think the jury is still out. 431 00:38:47,855 --> 00:38:51,448 haven't seen, you again, it's still early. 432 00:38:51,448 --> 00:38:56,482 And so I think we'll start to see the results in the next year or two. 433 00:38:56,602 --> 00:39:04,909 But generally speaking, I think part of it has to do with what the firm's top initiatives or objectives are. 434 00:39:04,909 --> 00:39:08,903 There are some firms where this is, they're going to dedicate more resources because it's critical to them. 435 00:39:08,903 --> 00:39:11,825 Like you mentioned Gunderson and, you know, I mentioned Wilson-Sansini. 436 00:39:11,825 --> 00:39:13,076 They are serving the tech industry. 437 00:39:13,076 --> 00:39:15,688 And so you're going to see a lot more organizational. 438 00:39:16,110 --> 00:39:19,452 momentum and support for something like this. 439 00:39:19,573 --> 00:39:29,700 You could also imagine another firm that may be focused on another industry, more traditional industries, um, that this may not be an oppressing concern. 440 00:39:29,700 --> 00:39:38,066 may say, you know, I know that for example, there's one firm out there that's already come on record to say we are not going to use any sort of generative AI. 441 00:39:38,066 --> 00:39:41,270 And so, you know, those are two extremes that I just provided. 442 00:39:41,270 --> 00:39:42,850 Um, some are in the middle. 443 00:39:42,850 --> 00:39:43,850 And so. 444 00:39:45,098 --> 00:39:50,178 Not every firm has the same appetite to push forward tech adoption and dedicate the right resources to it. 445 00:39:50,178 --> 00:39:58,678 So think that's going to be a big challenge, but that's also why there's reward for those who can really be ahead of the curve and lean into it and then generate operational 446 00:39:58,678 --> 00:40:01,418 efficiencies, goodwill with their clients. 447 00:40:02,198 --> 00:40:06,158 But I think we're going to have to see how things play out in the next 12 to 24 months. 448 00:40:06,478 --> 00:40:07,158 I the firm. 449 00:40:07,158 --> 00:40:13,202 I think I know the firm you're talking about that they got a lot of heat over that statement. 450 00:40:13,202 --> 00:40:16,524 And it really, they're not a tech forward firm. 451 00:40:16,524 --> 00:40:22,417 In fact, I think, I don't even think they're members of Ilta, but which is a little bit of an outlier. 452 00:40:22,417 --> 00:40:29,551 There's only about three or four Amlaw firms that I know of that aren't Ilta members. 453 00:40:29,551 --> 00:40:35,104 yeah, so I just discovered that we're deploying a new CRM and my sales team. 454 00:40:35,138 --> 00:40:40,001 was like, Hey, can we use the Ilta roster as kind of our contact list? 455 00:40:40,001 --> 00:40:42,723 And they were like, Hey, where's the, where's, are these three firms? 456 00:40:42,723 --> 00:40:44,934 And I went and looked and I'll be darned. 457 00:40:44,934 --> 00:40:46,165 They weren't on there. 458 00:40:46,185 --> 00:40:48,146 So yeah, they caught a lot of heat over that. 459 00:40:48,146 --> 00:40:53,409 It was, I don't think that was the best PR move. 460 00:40:53,409 --> 00:40:56,592 Um, I don't know. 461 00:40:56,592 --> 00:41:00,874 What are your thoughts on somebody putting a stake in the ground and going, yeah, we're not, we're not using gen AI. 462 00:41:02,746 --> 00:41:06,239 I tend to think that firms are pretty smart about all this stuff and sophisticated. 463 00:41:06,239 --> 00:41:10,252 And so there must've been some reason why they decided to put their stake in the ground. 464 00:41:10,252 --> 00:41:21,371 And one thing that I've in my relatively short career, right, in the legal industry have realized is that when it comes to firms, they take a lot of cues from their clients or 465 00:41:21,371 --> 00:41:22,642 their industry. 466 00:41:22,702 --> 00:41:32,700 So for example, you know, we talked about the tech firms that are the firms that support technology companies that are very much tech forward. 467 00:41:33,391 --> 00:41:45,168 They also tend to have pretty relaxed remote or hybrid work schedules where I seem to realize that the firms that rely heavily or have served the commercial real estate 468 00:41:45,168 --> 00:41:49,200 industry, they tend to have more strict return to office requirements. 469 00:41:49,200 --> 00:41:55,924 And so I do think firms take a cue from their clients as they should in focusing all of these things. 470 00:41:55,924 --> 00:41:59,516 And so my guess is that these firms that planted a stake in the ground, 471 00:41:59,730 --> 00:42:00,762 They know what they're doing. 472 00:42:00,762 --> 00:42:02,094 They're not doing it by accident. 473 00:42:02,094 --> 00:42:04,878 They are very, very thoughtful about all this stuff. 474 00:42:04,878 --> 00:42:08,322 yeah, I guess, you know, we'll see if that ends up being the right decision for them. 475 00:42:08,332 --> 00:42:09,222 Yeah. 476 00:42:09,363 --> 00:42:09,753 Yeah. 477 00:42:09,753 --> 00:42:11,643 No, that's, that's, that's a good point. 478 00:42:11,643 --> 00:42:25,790 Well, what other change, like you mentioned, um, the opportunity costs challenge with firms trying to manage this training and sorting of data and all that sort of stuff using 479 00:42:25,790 --> 00:42:27,751 kind of in-house legal talent for that. 480 00:42:27,751 --> 00:42:30,742 Um, obviously the opportunity cost is a big one. 481 00:42:30,742 --> 00:42:35,554 Are there other challenges that they might not be thinking about in that approach? 482 00:42:38,127 --> 00:42:49,295 You know, I think that, and this is a broader comment, historically, legal industry has operated in a very simple way. 483 00:42:49,295 --> 00:42:56,720 It's all full-time jobs, and it's mostly at law firms or in-house legal departments, maybe government too, or nonprofit. 484 00:42:56,720 --> 00:42:58,651 But generally, it's all full-time roles. 485 00:42:58,651 --> 00:43:04,905 And I think what we're seeing, given the trends around the pandemic, 486 00:43:05,928 --> 00:43:13,014 the emergence of Zoom, the adoption of cloud tools means that you don't need to be in the same office as someone else. 487 00:43:13,014 --> 00:43:19,150 And what that's done is that it's led to creative approaches to sourcing talent. 488 00:43:19,150 --> 00:43:24,104 Now that sourcing that talent, mean, you we work with a lot of in-house departments just to fill their roles. 489 00:43:24,104 --> 00:43:27,017 If they can't get a full-time hire, they leverage a contractor. 490 00:43:27,017 --> 00:43:29,239 That's been kind of what they've done for many years. 491 00:43:29,239 --> 00:43:33,462 But firms are now starting to see that there is an equally 492 00:43:33,726 --> 00:43:44,736 high caliber talent bench that exists in smaller cities and low cost of living areas where you have people with big law credentials who may not want to work a traditional, you know, 493 00:43:44,736 --> 00:43:46,717 law firm job in a big city. 494 00:43:46,717 --> 00:43:55,425 What that means for the firm is that there's a lot of work that needs to be done and you can rely on this separate talent pool to get it done. 495 00:43:55,425 --> 00:44:01,958 So you can, you know, what we've seen is like firms are leveraging contractors who have big law credentials and experience. 496 00:44:01,958 --> 00:44:10,741 often 15, 20 years out of law school at a rate that's surprisingly affordable or fits into the model of the firm. 497 00:44:10,822 --> 00:44:23,687 so yes, they can be there to help with AI projects and KM projects, but we've also seen it be helpful for firms to deploy these contractors on practice areas that have sporadic 498 00:44:23,687 --> 00:44:31,010 demand, like ups and downs, deal flow, like maybe M &A deals that spike up sometimes and then kind of go away at other times. 499 00:44:31,538 --> 00:44:38,942 Innovation does include AI and technology, it's also, we're seeing an impact on the human capital side of it, the talent management side of it. 500 00:44:38,942 --> 00:44:42,604 So that's kind of something that's cool that I've seen just being part of Latitude. 501 00:44:42,604 --> 00:44:46,346 And I expect to see more of that in the coming months and years. 502 00:44:46,540 --> 00:44:46,790 Yeah. 503 00:44:46,790 --> 00:44:50,551 I mean, that's a model that the tech world has leveraged forever. 504 00:44:50,551 --> 00:44:58,233 I I started offshoring projects in 2006 on ODesk, which is now Upwork. 505 00:44:58,734 --> 00:45:02,434 But that was the marketplace for international tech talent. 506 00:45:02,434 --> 00:45:07,556 And I was an early adopter and had a lot of success. 507 00:45:07,656 --> 00:45:11,137 We have leveraged a lot of different geographies. 508 00:45:11,137 --> 00:45:13,448 We did a lot in the Ukraine for many years. 509 00:45:13,448 --> 00:45:14,828 I had a team of 510 00:45:15,456 --> 00:45:29,922 eight or nine folks over there in 2014 when the whole Crimea situation happened and we had one client who supporting most of that who got nervous and we had to pull and we went to 511 00:45:29,922 --> 00:45:34,810 other areas and now we have folks in South Africa. 512 00:45:34,810 --> 00:45:37,876 I have one in Bolivia. 513 00:45:37,876 --> 00:45:39,916 I've got someone in the UK. 514 00:45:41,717 --> 00:45:44,058 I've got somebody in Canada. 515 00:45:44,396 --> 00:45:47,928 So it's a model that's proven and it works. 516 00:45:47,928 --> 00:45:51,240 And I don't know how we would get by without it. 517 00:45:51,240 --> 00:45:57,474 So it sounds like you're providing a very important service for these firms. 518 00:45:57,918 --> 00:45:58,288 Yeah. 519 00:45:58,288 --> 00:46:04,241 And I think, remember we talked about how the legal world is, you know, a couple of steps behind the rest of the business world. 520 00:46:04,241 --> 00:46:06,642 You know, we can call it five years, we can call it 15 years. 521 00:46:06,642 --> 00:46:11,344 If you've been using it since 2006, then you're probably 20 years ahead of some of our clients. 522 00:46:11,344 --> 00:46:21,948 But tying back to our original point, think trying something new for firms is challenging because of that whole brand reputation they've got to protect. 523 00:46:21,948 --> 00:46:23,969 They really value the relationship with clients. 524 00:46:23,969 --> 00:46:26,930 And so that's why they've been a little bit slower on it. 525 00:46:26,930 --> 00:46:37,117 And that's part of the reason why here at Latitude, what we do is we try to hire people who come from the industry, who have been big law partners or GCs who have contacts in the 526 00:46:37,117 --> 00:46:38,238 community. 527 00:46:38,858 --> 00:46:45,483 That enables us to build that trust for them to essentially try something new and different. 528 00:46:45,483 --> 00:46:48,755 And I know that that's a really hard thing to do. 529 00:46:48,755 --> 00:46:55,666 And that's the job I have, that's the job you have of convincing firms to do something differently that actually has a lot of benefits. 530 00:46:55,666 --> 00:47:09,347 It often is a hard slog to get there, but once you build that trust with them, I think it can lead to some of the most valuable and meaningful partnerships in the entire business 531 00:47:09,347 --> 00:47:10,557 world. 532 00:47:10,678 --> 00:47:17,003 I think that's something that I've come to realize over my time in innovation. 533 00:47:17,003 --> 00:47:21,986 You really need to build that trust before you've earned the right to say, here's a way to do it differently. 534 00:47:22,606 --> 00:47:23,956 Yeah, that's really good stuff. 535 00:47:23,956 --> 00:47:25,886 Well, this is, this has been a lot of fun. 536 00:47:25,886 --> 00:47:32,386 I, I appreciate you coming in and spending a few minutes and, I know you're, you're, busy. 537 00:47:32,386 --> 00:47:35,926 The CRO role is, is not, not an easy one. 538 00:47:35,926 --> 00:47:41,566 And you know, we talked about before we kicked off here, this is planning season for 2025. 539 00:47:41,566 --> 00:47:47,566 So we just had our offsite this week and I know you, you guys are, are working on your plans too. 540 00:47:47,566 --> 00:47:52,332 How, how do you, how do folks find out more about the, the, 541 00:47:52,332 --> 00:47:57,214 the writing that you do and, you your, your newsletter and whatnot. 542 00:47:57,214 --> 00:48:00,978 Yeah, my newsletter is at alexofftherecord.com. 543 00:48:00,978 --> 00:48:03,721 You can also find me on LinkedIn and there's links to it there. 544 00:48:04,202 --> 00:48:07,115 yeah, I'm very active on social, especially LinkedIn. 545 00:48:07,115 --> 00:48:12,331 So if you haven't checked out my profile or seen the stuff I've written, please do. 546 00:48:12,331 --> 00:48:15,194 And yeah, I'm always open to connecting. 547 00:48:15,328 --> 00:48:15,731 Awesome. 548 00:48:15,731 --> 00:48:16,155 Good stuff. 549 00:48:16,155 --> 00:48:18,611 And we'll include some links in the show notes as well. 550 00:48:19,400 --> 00:48:19,993 Thank you, Ted. 551 00:48:19,993 --> 00:48:22,184 And hey, thanks for having me on the show. 552 00:48:22,184 --> 00:48:23,756 Really enjoyed this conversation. 553 00:48:23,756 --> 00:48:24,857 Yeah, absolutely. 554 00:48:24,857 --> 00:48:31,542 I appreciate you spending the time and I'm sure I'll see you at one of these legal conferences. 555 00:48:31,542 --> 00:48:35,935 I go to about a dozen, so we'll have to connect in person while we're there. 556 00:48:37,137 --> 00:48:37,857 All right. 557 00:48:37,857 --> 00:48:39,077 Have a good one. 558 00:48:39,899 --> 00:48:40,269 All right. 559 00:48:40,269 --> 00:48:41,299 Thanks, Alex. -->