1 00:00:02,850 --> 00:00:04,692 Lev, how are you this afternoon? 2 00:00:05,263 --> 00:00:06,649 I'm good, Ted. 3 00:00:06,649 --> 00:00:07,780 Yourself? 4 00:00:07,918 --> 00:00:08,959 I'm great. 5 00:00:08,959 --> 00:00:10,439 I'm doing great. 6 00:00:10,560 --> 00:00:13,521 I'm really looking forward to this conversation. 7 00:00:13,861 --> 00:00:26,869 You and I, this is very often the case on how I find guests for the podcast is I see content or I have interactions on LinkedIn that align with what I know our audience likes 8 00:00:26,869 --> 00:00:27,709 to hear about. 9 00:00:27,709 --> 00:00:30,231 And you and I were on a thread. 10 00:00:30,231 --> 00:00:36,224 think it was, was that the Andrew Wang um meme or yeah. 11 00:00:37,292 --> 00:00:45,287 where he basically said, you know, AI is already replacing third year associate level work. 12 00:00:45,287 --> 00:00:51,910 And there was a lot of debate, um good debate, but I liked a lot of the things you were saying. 13 00:00:51,910 --> 00:00:57,653 So I'm looking forward to the conversation, but before we jump in, let's get you introduced. 14 00:00:57,653 --> 00:01:05,037 So you were a former &A partner at Linklater's for 20 plus years and doing corporate law work. 15 00:01:05,037 --> 00:01:06,126 um 16 00:01:06,126 --> 00:01:08,326 You're an economist by training. 17 00:01:08,546 --> 00:01:12,606 You got a lot of deep expertise in digital transformation. 18 00:01:13,006 --> 00:01:14,406 Fill in the gaps for us. 19 00:01:14,406 --> 00:01:19,406 Tell us about what your background is and what you're up to today. 20 00:01:19,731 --> 00:01:21,451 Not much to add. 21 00:01:21,451 --> 00:01:24,250 You summed it up perfectly, really. 22 00:01:24,271 --> 00:01:26,371 I am a lawyer. 23 00:01:26,371 --> 00:01:31,391 I spent 20 years in the legal profession working at Sullivan and Cromwell. 24 00:01:31,391 --> 00:01:32,711 That's where I started out. 25 00:01:32,711 --> 00:01:35,051 Then I moved to Link Laters. 26 00:01:35,311 --> 00:01:38,931 I did cross-border work, M &A. 27 00:01:39,211 --> 00:01:43,595 Worked in London, worked in Moscow, worked in New York. 28 00:01:43,860 --> 00:01:56,640 And about a year ago, I resigned from LinkLators to become a legal tech investor, a legal tech angel investor for the time being. 29 00:01:56,860 --> 00:01:59,460 I invest in various legal tech startups. 30 00:01:59,460 --> 00:02:04,120 I advise legal tech startups, formally and informally. 31 00:02:04,120 --> 00:02:08,160 And I really like the space. 32 00:02:08,160 --> 00:02:09,460 What can I say? 33 00:02:09,520 --> 00:02:13,440 That's what brings me to all those threads about 34 00:02:13,795 --> 00:02:16,463 AI and legal and innovation. 35 00:02:17,102 --> 00:02:22,524 Well, so you've got an interesting perspective as somebody who practiced law for so many years. 36 00:02:22,524 --> 00:02:36,288 And um what is your take on AI's capabilities as it relates to the area where you spent most of your time in transactional work? 37 00:02:36,489 --> 00:02:40,050 Is it replacing? 38 00:02:40,406 --> 00:02:53,506 work done by attorneys today, if not how soon in the future will it be taking at least some of those more mundane tasks and completing the lion's share of the work. 39 00:02:54,449 --> 00:03:09,445 I genuinely think that the capabilities of technology, even today, and I think that we're in the early days, I think the capabilities are unimaginably impressive. 40 00:03:09,725 --> 00:03:11,186 Unimaginably impressive. 41 00:03:11,186 --> 00:03:20,010 ah I'm not thinking about it in terms of replacing lawyers or not replacing lawyers. 42 00:03:20,010 --> 00:03:22,351 That's a debate that we can have. 43 00:03:22,351 --> 00:03:23,663 ah 44 00:03:23,663 --> 00:03:37,447 It's more about can a system, can a tool ah do work that is comparable to a junior associate or a mid-level associate? 45 00:03:37,447 --> 00:03:42,148 And the answer to that is absolutely it can, right? 46 00:03:42,148 --> 00:03:43,448 That's the truth of it. 47 00:03:43,448 --> 00:03:53,591 The best solutions out there without any question can definitely do the same work that a junior associate can do. 48 00:03:53,907 --> 00:04:00,010 ah And many of the better ones can also do the work that a mid-level associate can do. 49 00:04:00,010 --> 00:04:07,494 um The pace of changes is what really strikes me. 50 00:04:07,494 --> 00:04:23,343 um I'd say two years ago ah when I just started looking at this, um it was difficult to imagine, difficult to fathom that uh 51 00:04:23,343 --> 00:04:25,845 we'd be where we are today, right? 52 00:04:25,845 --> 00:04:32,449 ah The early chatbots work and have hopelessly riddled with uh hallucinations. 53 00:04:32,449 --> 00:04:41,596 uh It felt like the guardrails would uh take years to make and put in place. 54 00:04:41,596 --> 00:04:46,179 ah The reality is different, right? 55 00:04:46,179 --> 00:04:50,462 Philosophically, what we do as lawyers is we work with text. 56 00:04:50,462 --> 00:04:52,663 We work with a lot of text. 57 00:04:52,849 --> 00:05:02,423 I'm thinking transactional, um Litigators work with all kinds of stuff, uh images, audio and all of that. 58 00:05:02,624 --> 00:05:17,051 But the products that I'm seeing, can do due diligence, for instance, ah much faster, much cheaper and more accurately than ah a big team of junior associates. 59 00:05:17,051 --> 00:05:18,091 So it's... 60 00:05:18,277 --> 00:05:29,260 you get a due diligence report in a matter of hours as opposed to a month of a big team with sort of numerous iterations. 61 00:05:29,742 --> 00:05:32,284 I don't know if that answers your question. 62 00:05:32,445 --> 00:05:33,903 What do you think about this? 63 00:05:33,903 --> 00:05:47,223 Yeah, so I have been using AI side by side on legal work for my business for about six months. 64 00:05:47,603 --> 00:05:51,383 And it started with a simple labor and employment matter. 65 00:05:51,383 --> 00:05:56,883 We were hiring somebody from a firm with a little bit of overlap, and they had some non-compete language. 66 00:05:56,883 --> 00:06:01,683 And we hired a L &E attorney to advise us. 67 00:06:01,702 --> 00:06:05,664 And in parallel, I asked all the questions. 68 00:06:05,664 --> 00:06:13,869 I wrote down all the questions that I asked the attorney about how we keep from running afoul of that non-compete. 69 00:06:14,129 --> 00:06:18,811 And I asked the attorney, and then I asked AI. 70 00:06:18,972 --> 00:06:24,715 And the answers were very close. 71 00:06:24,715 --> 00:06:26,296 the added benefit that I got. 72 00:06:26,296 --> 00:06:27,667 And this was my first experiment. 73 00:06:27,667 --> 00:06:30,278 I've got lots of examples, but. 74 00:06:30,610 --> 00:06:35,192 What the AI did better was roleplay. 75 00:06:35,292 --> 00:06:43,075 So I have a custom GPT that's trained on, I call it my co-CEO. 76 00:06:43,075 --> 00:06:55,801 So it has as training materials, it has our core values, it has our pitch deck, it has our quarterly all hands meeting notes and deck. 77 00:06:55,801 --> 00:06:59,062 has, we do uh periodic 78 00:06:59,062 --> 00:07:03,144 retreats with leadership where we map out the next quarter. 79 00:07:03,485 --> 00:07:06,126 It has all that information as training materials. 80 00:07:06,126 --> 00:07:07,767 So then I role played with it. 81 00:07:07,767 --> 00:07:08,908 It has our org chart. 82 00:07:08,908 --> 00:07:10,209 I mean, it has a bunch of stuff. 83 00:07:10,209 --> 00:07:12,030 It knows everything about us. 84 00:07:12,030 --> 00:07:16,152 It even has our operating agreement and corporate docs in there. 85 00:07:16,152 --> 00:07:20,595 I asked it the role playing part because it knew us so well. 86 00:07:20,595 --> 00:07:22,886 And this was the first time that we had used this attorney. 87 00:07:22,886 --> 00:07:24,417 And he did a great job, by the way. 88 00:07:24,417 --> 00:07:25,958 um 89 00:07:26,370 --> 00:07:30,772 But when we role played and said, OK, how can we construct this role? 90 00:07:30,772 --> 00:07:34,133 And again, it had the language from the non-compete. 91 00:07:34,133 --> 00:07:36,444 It did a fantastic job. 92 00:07:36,444 --> 00:07:40,196 It wrote the job description for me because it knew my business. 93 00:07:40,196 --> 00:07:41,936 It knew where we had needs. 94 00:07:41,936 --> 00:07:50,970 It even placed it on the org chart in the reporting structure that, made the most sense because I had the guy's CV that I also uploaded. 95 00:07:50,970 --> 00:07:52,721 Like, these are his capabilities. 96 00:07:52,801 --> 00:07:53,881 Because we want to be careful. 97 00:07:53,881 --> 00:07:56,202 Like, as a young company, last thing you want to get 98 00:07:56,291 --> 00:08:00,162 do is get a lawsuit against you um for something like that. 99 00:08:00,162 --> 00:08:04,613 So it was incredible with how the role playing part. 100 00:08:04,834 --> 00:08:18,758 And um I think it was a uh function of I was able to train it on our organization, whereas it would have taken a lot of time for me to train the attorney to really understand us. 101 00:08:20,326 --> 00:08:21,467 No doubt about that. 102 00:08:21,467 --> 00:08:24,789 And you're not uh the exception. 103 00:08:24,789 --> 00:08:40,828 There's a lot of clients who use uh either out-of-the-box tools or custom-made, know, customized GPTs uh to either double-check the legal work or, frankly, do the legal work 104 00:08:40,828 --> 00:08:42,018 themselves. 105 00:08:43,319 --> 00:08:49,162 And then they have uh law firms just review that and sign on. 106 00:08:49,555 --> 00:08:53,058 We're not talking about multi-billion dollar transactions. 107 00:08:53,058 --> 00:08:57,762 Fine, that's probably a few years away, right? 108 00:08:57,762 --> 00:09:15,036 But uh we are talking about uh basic uh questions that uh sometimes take people a week or two to get answers to, because attorneys are busy and especially if it's a small question, 109 00:09:15,036 --> 00:09:16,297 you may not get... 110 00:09:16,297 --> 00:09:17,418 uh 111 00:09:17,752 --> 00:09:21,924 the answer right away, right? 112 00:09:22,004 --> 00:09:25,715 You need to take a spot on the queue, right? 113 00:09:25,715 --> 00:09:31,308 And then by the time the attorney gets to your question, that's when you'll get the response. 114 00:09:31,308 --> 00:09:36,980 ah So people kind of are circumventing the queue, right? 115 00:09:36,980 --> 00:09:42,302 They circumvent, you know, they go around the queue and they get the answer. 116 00:09:42,302 --> 00:09:46,874 They ask the attorney to spend sort of a few minutes to just take a look at the 117 00:09:47,089 --> 00:09:52,521 response and bless it or curse it for that matter. 118 00:09:52,521 --> 00:10:08,517 uh the other thing is uh law firms, uh AI enabled law firms, Traditional ones uh have started using it to great effect. 119 00:10:08,517 --> 00:10:17,050 I just saw a post by uh Michael Huseby uh from the investment lawyers, the fund formation lawyers. 120 00:10:17,170 --> 00:10:29,950 who said that he routinely now checks at least two tools before sort of delving into the source material himself. 121 00:10:30,730 --> 00:10:34,670 it's not, you know, Michael is not the exception either. 122 00:10:34,670 --> 00:10:46,370 He's at the cutting edge of the profession, I would say, but it is becoming a lot more common for people to do that. 123 00:10:47,582 --> 00:10:49,523 And the responses are good. 124 00:10:49,523 --> 00:10:53,826 That's the wonderful thing and the scary thing, right? 125 00:10:53,826 --> 00:10:55,606 The responses are good. 126 00:10:55,606 --> 00:11:12,538 um The thing that really brought me to this space, one of the conversations that really brought me to this space was actually in the early days when um Chad GPC just came out, I 127 00:11:12,538 --> 00:11:13,508 had a... 128 00:11:13,818 --> 00:11:27,468 with a friend of mine, a very experienced capital markets partner, uh and I asked him whether he um had experimented with ChetGPC, and he told me, yeah, I was actually short on 129 00:11:27,468 --> 00:11:35,194 time, and I needed to prepare for this management interview, management due diligence uh interview. 130 00:11:35,194 --> 00:11:40,988 um So I asked ChetGPC if for 131 00:11:40,988 --> 00:11:42,739 for kicks, right, for the hell of it. 132 00:11:42,739 --> 00:11:55,774 um I asked that to prepare an agenda, right, a list of questions to ask without betraying any client confidences, without the name of the client just said, this is the company that 133 00:11:55,774 --> 00:11:59,985 is um about to issue some public debt. 134 00:11:59,985 --> 00:12:04,127 uh What kind of questions should I ask? 135 00:12:04,147 --> 00:12:08,168 I am an experienced attorney, you know, know the drill, right? 136 00:12:08,168 --> 00:12:10,349 The prompt is simple enough. 137 00:12:10,351 --> 00:12:14,525 And he said that out of the... 138 00:12:14,525 --> 00:12:16,967 he limited it to something like 20 questions. 139 00:12:16,967 --> 00:12:26,394 uh He said out of the 20 questions, 15 were completely on point, um things that he would have asked himself. 140 00:12:26,635 --> 00:12:32,519 Two or three were probably questions that he wouldn't ask, but they weren't horrible. 141 00:12:32,720 --> 00:12:39,265 And two or three were actually things that he didn't think about, but he should have asked. 142 00:12:39,901 --> 00:12:45,146 And that's something that he has been doing his entire life. 143 00:12:45,587 --> 00:12:51,092 If at the very early stage the technology is such that it can... 144 00:12:55,437 --> 00:13:04,450 suggests something new to very experienced lawyer, then there is really something to that technology that uh holds tremendous potential. 145 00:13:04,451 --> 00:13:11,971 And uh that was one of the reasons that kind of pushed me into this uh current role. 146 00:13:12,130 --> 00:13:12,751 Yeah. 147 00:13:12,751 --> 00:13:17,935 You know, another thing I used it on is, uh, I experiment with a lot of the different models. 148 00:13:17,935 --> 00:13:24,339 There's really kind of four in my rotation now, Grok, Claude, Gemini, and Chagy BT. 149 00:13:25,321 --> 00:13:29,684 and I used Grok for this and it did really well when Grok four came out. 150 00:13:29,684 --> 00:13:32,667 It was, uh, that's when I started paying more attention. 151 00:13:32,667 --> 00:13:41,794 I took our four nine a program, which is like around kind of shadow equity, um, kind of stock option incentives and 152 00:13:42,227 --> 00:13:54,290 that had been drawn up by a lawyer and I ran it through and asked, I used to be an auditor and I worked for Bank of America in corporate audit, doing anti-money laundering audit and 153 00:13:54,290 --> 00:13:55,650 technology audit. 154 00:13:55,650 --> 00:14:03,002 So I've got a risk management background and I coached it to go through the entire document. 155 00:14:03,002 --> 00:14:10,934 It's about 40 pages long and give me a risk priority number, like identify every risk in the 156 00:14:11,022 --> 00:14:17,587 in the program and then risk it something called RPN we use in the audit world, risk priority number. 157 00:14:17,587 --> 00:14:22,971 It's um what is the magnitude of the risk? 158 00:14:22,971 --> 00:14:24,712 Should it materialize? 159 00:14:24,712 --> 00:14:28,695 It's severity detection and likely of occurrence. 160 00:14:28,695 --> 00:14:34,900 So I had it multiply those three things and give it a numeric score and rank it in descending order. 161 00:14:34,900 --> 00:14:37,782 So the highest risk came up top. 162 00:14:37,974 --> 00:14:49,671 And it gave me detailed breakdown in each, you know, what the likelihood of severity, um what is the likelihood of occurrence and what is the likelihood of detection? 163 00:14:49,671 --> 00:14:53,003 How would you know if this risk were to materialize? 164 00:14:53,003 --> 00:14:55,785 In the risk management world, that's uh a thing. 165 00:14:55,785 --> 00:15:01,358 um And it did incredible work helping us. 166 00:15:01,358 --> 00:15:03,410 I took that output, handed it to my lawyer. 167 00:15:03,410 --> 00:15:04,410 He was like, 168 00:15:04,726 --> 00:15:12,650 Okay, yeah, this is, this is a good starting point because the top three things on here, we ought to, uh, we ought to remediate this, this and this. 169 00:15:12,650 --> 00:15:20,355 Don't worry about not really an issue given the context, but I was, this was an untrained, this was not a legal specific tool. 170 00:15:20,355 --> 00:15:27,739 wasn't using a custom GPT like in my, my co CEO that knows my business and it hit it out of the park. 171 00:15:27,739 --> 00:15:34,042 I've got dozens of examples, like just over the last six months, and I'm a small buyer of legal services. 172 00:15:34,126 --> 00:15:35,646 I don't know what we spend. 173 00:15:35,646 --> 00:15:40,226 mean, it's more than six figures a year, but it's not a huge, it's not millions of dollars. 174 00:15:40,326 --> 00:15:53,766 Um, and just in six months, I have probably, I don't know, I'm guessing eight to 10 really compelling examples of where, uh, it saved me money, time, or both. 175 00:15:55,670 --> 00:15:57,010 Not surprised at all. 176 00:15:57,010 --> 00:15:59,391 Not surprised at all. 177 00:16:00,551 --> 00:16:12,694 will say ah that foundational models, I'd be wary of using them uh for Betta Farm kinds of things. 178 00:16:12,934 --> 00:16:13,414 True. 179 00:16:13,414 --> 00:16:23,277 uh For uh basic level stuff, for stuff that is uh not Betta Farm. 180 00:16:24,497 --> 00:16:28,857 The way you are using it is extremely, extremely right. 181 00:16:28,877 --> 00:16:37,817 So you're basically asking questions, you're getting the input, and then you have it double-checked by your professionals. 182 00:16:38,037 --> 00:16:41,617 That is certainly the way to go. 183 00:16:42,757 --> 00:16:53,257 There are solutions either being developed or that have been developed that I would say take this a bit further. 184 00:16:53,385 --> 00:17:07,256 And it's no longer necessary for you to have a deep understanding of what you're looking for, what you actually want to take to your lawyer, to your outside counsel. 185 00:17:07,256 --> 00:17:11,359 ah It will be sort of automatic. 186 00:17:11,520 --> 00:17:13,802 It will ask the right questions. 187 00:17:13,802 --> 00:17:15,133 The playbooks are there. 188 00:17:15,133 --> 00:17:16,444 The workflows are there. 189 00:17:16,444 --> 00:17:22,989 uh All you need to do is provide the information to the tool. 190 00:17:23,057 --> 00:17:31,871 um Importantly enough, the better tools are kind of going the agents accrued. 191 00:17:31,871 --> 00:17:41,965 Everybody's talking about the agents accrued and they are flagging ah that something ought to be reviewed. 192 00:17:42,125 --> 00:17:42,585 Right? 193 00:17:42,585 --> 00:17:52,451 So if somebody gets a document in an email, the system will recognize that it's a legal document and it will say, do you want 194 00:17:52,451 --> 00:17:56,003 the system to kind of run some kind of an analysis. 195 00:17:56,003 --> 00:17:58,924 Do you want to set the parameters of the analysis? 196 00:17:58,924 --> 00:18:02,486 Do you want to rely on the preset parameters of the analysis? 197 00:18:03,007 --> 00:18:06,708 And those are impressive, very impressive. 198 00:18:06,708 --> 00:18:22,056 So uh high level analysis is definitely something that is already there and it's at least comparable to the average lower uh 199 00:18:22,242 --> 00:18:24,614 of the relevant seniority. 200 00:18:24,614 --> 00:18:33,831 ah The in-depth solutions are things that are still being built. 201 00:18:33,831 --> 00:18:40,317 The ecosystem is still in its nascency, It's infancy, I should say. 202 00:18:40,317 --> 00:18:42,168 It's very early days. 203 00:18:42,168 --> 00:18:51,515 um Despite all the headlines, despite all the great successes that we've seen, 204 00:18:51,696 --> 00:18:58,816 it's still not at all clear who the winners will be. 205 00:19:00,156 --> 00:19:08,876 It's still something that is very much an open race. 206 00:19:08,876 --> 00:19:09,766 Yeah. 207 00:19:09,807 --> 00:19:21,074 You know, when you and I talked last time, you brought up an interesting angle on something that I've been thinking about quite deeply. 208 00:19:21,074 --> 00:19:36,565 that is, and my listeners have heard many times me talk about the foundational fundamental characteristics of the law firm world that create headwinds towards innovation and 209 00:19:36,565 --> 00:19:37,545 transformation. 210 00:19:37,545 --> 00:19:38,886 uh 211 00:19:39,353 --> 00:19:43,686 Transformation and innovation um aren't necessarily the same thing. 212 00:19:43,686 --> 00:19:48,979 um Transformation has much greater magnitude. 213 00:19:48,979 --> 00:19:50,890 Innovation can lead to transformation. 214 00:19:50,890 --> 00:20:02,077 um But for real transformation, I feel like there is a lot of legacy, cultural barriers. 215 00:20:02,077 --> 00:20:03,616 um 216 00:20:03,616 --> 00:20:12,492 Again, fundamental business structures like the partnership model that very much prioritizes profit taking at year end. 217 00:20:12,713 --> 00:20:17,495 the concept of R &D in a law firm is a bit foreign. 218 00:20:17,556 --> 00:20:22,339 They don't tend to do traditional R &D. 219 00:20:22,339 --> 00:20:27,006 And you brought up a good point that I want to expand on a little bit. 220 00:20:27,006 --> 00:20:30,605 I guess let me also say the 221 00:20:32,150 --> 00:20:40,432 adversity towards risk that lawyers that mindset tends to have as they're trained to be risk averse. 222 00:20:40,612 --> 00:20:52,916 Um, and you brought up a good point, which I hadn't really thought about before, which was law firms make investments that are risky all the time in like opening up new offices and 223 00:20:52,916 --> 00:20:53,486 areas. 224 00:20:53,486 --> 00:20:55,557 And I thought that was a good counterpoint. 225 00:20:55,557 --> 00:21:00,918 I still think my, I still think my original point is valid. 226 00:21:01,098 --> 00:21:08,024 in terms of traditional R &D, like explain how law firms take risks. 227 00:21:08,024 --> 00:21:09,545 I mean, that was one good example. 228 00:21:09,545 --> 00:21:10,606 We can expand on that. 229 00:21:10,606 --> 00:21:15,720 There's maybe others like they do take risks, correct? 230 00:21:16,154 --> 00:21:17,104 They do, they do. 231 00:21:17,104 --> 00:21:22,066 And just to be clear, Ted, I was not disagreeing with you. 232 00:21:22,066 --> 00:21:25,427 Everything you said is valid, right? 233 00:21:25,427 --> 00:21:27,447 Lawyers are risk averse. 234 00:21:27,628 --> 00:21:30,529 The partnership model is the partnership model. 235 00:21:30,529 --> 00:21:40,953 uh Where I sort of diverged a little bit from what you and a lot of others are saying is, uh 236 00:21:40,953 --> 00:21:59,137 This notion that law firms are just uh short-termist uh profits at the end of the year is ah the ultimate goal, and there is sort of no recognition or no ability to make 237 00:21:59,137 --> 00:21:59,877 investments. 238 00:21:59,877 --> 00:22:10,820 um Law firms make investments all the time, whether it's opening up a new office, hiring a very 239 00:22:11,980 --> 00:22:16,643 impressive lateral partner or a group of lateral partners, right? 240 00:22:16,944 --> 00:22:27,151 And it's the same problem with consensus building that people are talking about when they're talking about uh technology adoption, right? 241 00:22:27,151 --> 00:22:40,610 um How do you bring on board somebody who doesn't see immediate benefits from that technology in his or her own practice, right? 242 00:22:40,760 --> 00:22:47,563 A valid point, true, some people will benefit more and some people will benefit less. 243 00:22:48,284 --> 00:23:00,571 However, it is absolutely the same thing when I'm at a global law firm and somebody says, we should open an office in Indonesia. 244 00:23:00,671 --> 00:23:09,776 If my practice is very much US centric or UK centric and my clients are not looking at Indonesia, 245 00:23:09,776 --> 00:23:15,578 to invest in that uh jurisdiction, how am I going to benefit? 246 00:23:15,578 --> 00:23:16,659 I won't. 247 00:23:17,259 --> 00:23:27,623 And yet there is the notion that somehow the firm is bigger than individual partners or individual practice groups. 248 00:23:27,623 --> 00:23:39,468 uh And it's um the same with uh lateral partners or electing the current cohort of partners, homegrown partners. 249 00:23:39,504 --> 00:23:40,284 It's the same thing. 250 00:23:40,284 --> 00:23:45,127 ah You don't necessarily know how many of them are going to be successful. 251 00:23:45,127 --> 00:23:53,610 You don't know if the investment, this is an investment, ah if that investment is going um to pay for itself. 252 00:23:54,251 --> 00:23:56,331 And yet law firms routinely do it. 253 00:23:56,331 --> 00:24:06,672 ah And the truth of the matter is uh when people imagine people on the outside, 254 00:24:06,672 --> 00:24:09,594 sort of not having worked at law firms. 255 00:24:09,594 --> 00:24:16,778 They imagined law firms to be almost like Greek democracies, right? 256 00:24:16,939 --> 00:24:26,865 Where everybody uh gets a vote and you need to convince pretty much everyone uh in order to get a decision through. 257 00:24:26,865 --> 00:24:30,967 That is just not the case for big law firms. 258 00:24:31,928 --> 00:24:34,890 Big law firms are run by management. 259 00:24:34,890 --> 00:24:36,391 um 260 00:24:37,121 --> 00:24:41,134 The management does consult with the rank and file. 261 00:24:41,134 --> 00:24:44,806 um There are multiple reporting lines. 262 00:24:44,806 --> 00:24:52,822 There's multiple ways in which the management solicits input um from the partnership. 263 00:24:52,822 --> 00:24:55,603 And they're doing a marvelous job of that. 264 00:24:55,884 --> 00:25:01,007 But ultimately, it is the management that needs to be convinced. 265 00:25:01,007 --> 00:25:06,573 And then once the management posits something to the partnership, the partnership 266 00:25:06,573 --> 00:25:11,694 tends to listen, which is not to say that it's a rubber stamp, right? 267 00:25:11,694 --> 00:25:21,527 There are situations where the partnership is going to be opposed and will say, well, actually, no, we don't like this idea. 268 00:25:21,807 --> 00:25:28,479 But by and large, it's more common for people to trust their management. 269 00:25:28,479 --> 00:25:30,269 Otherwise, they wouldn't have elected them, right? 270 00:25:30,269 --> 00:25:36,323 um So the management can and does um 271 00:25:36,323 --> 00:25:39,905 propose things that are sort of inherently risky. 272 00:25:39,905 --> 00:25:46,149 And it's, it's no different with technology adoption is what I'm saying, right? 273 00:25:46,229 --> 00:26:02,679 The management uh has been talking to the partnerships at pretty much all the big law firms in the world, telling them, guys, we have to be um at least cognizant of the risk. 274 00:26:03,212 --> 00:26:15,563 We have to be least cognizant of the risk that the poses to the profession, and we have to explore how we make best use of this technology without running afoul of our professional 275 00:26:15,563 --> 00:26:20,166 duties and without ah getting in trouble with our clients. 276 00:26:20,407 --> 00:26:24,049 And our clients demand it, we have to follow. 277 00:26:24,050 --> 00:26:28,454 That's the nature of the law firm environment. 278 00:26:28,454 --> 00:26:30,535 ah 279 00:26:31,140 --> 00:26:37,504 Hopefully, this once again, this answers your question, but let me know if you'd like me to expand on it. 280 00:26:37,504 --> 00:26:38,675 No, that makes sense. 281 00:26:38,675 --> 00:26:41,316 Let's have a little debate here. 282 00:26:41,316 --> 00:26:50,661 um I would argue that the law firm business model has limited ability to scale. 283 00:26:50,661 --> 00:26:55,464 um And I'll tell you what I'm basing that on. 284 00:26:55,464 --> 00:26:57,635 We talked a little bit about it the last time. 285 00:26:57,635 --> 00:27:07,030 The largest law firm in the world, K &E, roughly $8 billion in revenue, is one fifth the size of the smallest of the big four. 286 00:27:07,328 --> 00:27:16,443 If you add up all of the revenues of the top five law firms in the world, they have like 6 % of market share. 287 00:27:16,443 --> 00:27:23,467 If you look at the big four, they have like over 90 % of the audit work of all publicly traded companies. 288 00:27:23,467 --> 00:27:25,988 It's extremely fragmented. 289 00:27:26,088 --> 00:27:31,731 So no firm has been able to scale, even K &E, which is an outlier at 8 billion. 290 00:27:31,831 --> 00:27:37,184 The average revenue at an AMLAL 100 firm is around 1 billion. 291 00:27:37,874 --> 00:27:44,201 Why do we think the current model accommodates scale? 292 00:27:47,365 --> 00:27:58,955 There are multiple reasons why uh the legal market is more fragmented than the big four market, right, than the audit market. 293 00:27:58,955 --> 00:27:59,976 uh 294 00:28:00,888 --> 00:28:04,189 I'd say some of it has to do with barriers to entry. 295 00:28:04,229 --> 00:28:19,416 Some of it has to do with the fact that uh we are talking about different jurisdictions, some of which are less uh susceptible to be folded into your practice. 296 00:28:19,416 --> 00:28:25,658 ah And some of it is just uh the fact that lawyers 297 00:28:28,111 --> 00:28:30,733 Culture is probably a bit more important. 298 00:28:30,733 --> 00:28:33,114 I know Big Four uh fairly well. 299 00:28:33,114 --> 00:28:34,215 I've worked with them. 300 00:28:34,215 --> 00:28:38,457 They used to be clients of mine, uh great people working there. 301 00:28:38,457 --> 00:28:41,379 uh My wife actually worked for a Big Four firm. 302 00:28:41,379 --> 00:28:54,667 uh At a certain point, the Big Four became fairly, firms within the Big Four became... 303 00:28:55,415 --> 00:28:58,596 I'd say pretty much mirror images of one another. 304 00:28:59,037 --> 00:29:04,599 Culturally, they're very similar, um or so it seems from the outside. 305 00:29:04,939 --> 00:29:06,340 Law firms are different. 306 00:29:06,340 --> 00:29:08,361 uh that's the truth of it. 307 00:29:08,361 --> 00:29:17,604 um Integrating two big law firms is a major challenge. 308 00:29:17,625 --> 00:29:23,707 Integrating 10 big law firms is a task that very few 309 00:29:23,757 --> 00:29:25,849 have been able to do successfully. 310 00:29:25,849 --> 00:29:39,080 Actually, Linklater has done that um in the 90s when it merged the Alliance firms, as they were known at the time, and uh sort of expanded it seriously. 311 00:29:39,080 --> 00:29:46,086 um There have been some good examples, and there have been some bad examples, and everyone knows it. 312 00:29:46,086 --> 00:29:48,267 I'm not even going to talk about that. 313 00:29:48,268 --> 00:29:52,815 So um law firms are not... 314 00:29:52,815 --> 00:29:56,115 looking to be like the big four. 315 00:29:56,115 --> 00:30:03,915 Law firms do not have ambitions to be sort of a Fortune 500 company. 316 00:30:04,795 --> 00:30:08,615 Some of it has to do with the restriction of outside investments in law firms. 317 00:30:08,615 --> 00:30:18,155 Some of it has to do with the ethical and professional restrictions that are imposed on lawyers. 318 00:30:18,395 --> 00:30:21,314 But ultimately, the... 319 00:30:21,314 --> 00:30:29,936 Firms are not going to invest billions and billions of dollars um in legal AI. 320 00:30:29,936 --> 00:30:30,597 They won't. 321 00:30:30,597 --> 00:30:39,179 uh Solutions need to be sensible in the context of the law firm model that exists right now. 322 00:30:40,059 --> 00:30:50,742 That leaves open the question of, is it possible to build a technology company that becomes a law firm? 323 00:30:50,742 --> 00:31:01,611 right, the AI native law firms that Y Combinator has called for and that uh have sprung up, uh know, Crosby comes to mind and the others. 324 00:31:01,611 --> 00:31:05,854 uh And that's an experiment. 325 00:31:05,914 --> 00:31:11,398 That's an experiment that uh we are still waiting to see the outcome. 326 00:31:14,312 --> 00:31:30,653 If it does work, um the value of those businesses will be tremendously higher than ah that of a traditional law firm. 327 00:31:30,653 --> 00:31:33,115 They're going to be valued differently. 328 00:31:34,096 --> 00:31:36,037 The multiples will be higher. 329 00:31:36,337 --> 00:31:43,582 And in a sense, for some of the more ambitious law firms, 330 00:31:44,687 --> 00:31:52,151 probably outside of the Amlo 100, uh mid-sized law firms, smaller law firms. 331 00:31:52,291 --> 00:31:58,114 This is clearly a path that they can take, and they can compete with the big guys. 332 00:31:58,594 --> 00:32:05,618 And ah that's not an existential threat at the moment. 333 00:32:05,618 --> 00:32:09,420 I wouldn't call it that, but it is a threat down the line. 334 00:32:09,420 --> 00:32:13,542 And it is something that uh 335 00:32:13,674 --> 00:32:32,610 I would certainly consider if I were running a successful mid-market firm or a small firm of several partners with technology uh using the technological footprint, they can 336 00:32:32,610 --> 00:32:36,994 certainly compete with the big guys and that can scale quickly. 337 00:32:37,134 --> 00:32:41,878 For big firms, uh the legacy oh systems are 338 00:32:42,230 --> 00:32:55,995 difficult to replace uh and that's not a path that I think any big firm is probably going to explore, throwing out the baby with the bath water and um building completely from 339 00:32:55,995 --> 00:32:56,966 scratch. 340 00:32:57,228 --> 00:33:06,658 Yeah, I have laid out a blueprint for, I call it a big law 2.0 blueprint for transformation. 341 00:33:06,658 --> 00:33:17,974 I would, if I were in that position leading a large law firm, I think, I'm not a lawyer, but I've worked with more than half the AMLaw in roughly 20 years. 342 00:33:18,155 --> 00:33:20,277 And so I know a lot about how they work. 343 00:33:20,277 --> 00:33:27,082 I've seen a really good cross-sectional view and some of the opportunities and challenges inside these organizations. 344 00:33:27,574 --> 00:33:44,629 If I were running one of these, I would be advocating to deploy a plan like that blueprint, which is basically take a non-trivial amount of bottom line revenue and deploy 345 00:33:44,629 --> 00:33:56,481 it into an entirely new structure, not a partnership, uh one that's a C-corp led by professional management with a board and 346 00:33:56,481 --> 00:33:58,072 outside investors. 347 00:33:58,072 --> 00:34:04,718 And obviously we have, you know, the ABA model rule five four that prevents legal, non-legal ownership at the moment. 348 00:34:04,718 --> 00:34:10,942 But there are a few States now where, you could, deploy this in a sandbox. 349 00:34:11,243 --> 00:34:15,716 And I would do this incrementally with a small portion. 350 00:34:15,716 --> 00:34:25,566 I think I outlined 20%, as a maybe starting point, because I agree with you that it's not, you're not going to be able to take. 351 00:34:25,566 --> 00:34:37,257 You know, there's legacy um barriers, legacy baggage in these existing organizations that will make it very difficult. 352 00:34:37,257 --> 00:34:44,964 So starting clean with a traditional governance model, you know, that's one of the reasons why I think that law firms aren't able to scale. 353 00:34:44,964 --> 00:34:46,225 It's their governance model. 354 00:34:46,225 --> 00:34:51,500 There's one, there's many, but one of them, you know, in a traditional C-corp, 355 00:34:51,500 --> 00:34:53,351 You have a group of investors. 356 00:34:53,811 --> 00:34:54,632 elect a board. 357 00:34:54,632 --> 00:35:09,580 That board installs professional management that are fully empowered and accountable and heavily invested with dollars in the success um of that stakeholder value. 358 00:35:09,580 --> 00:35:14,403 um That's not the case in law firms. 359 00:35:14,403 --> 00:35:18,445 Non-legal professionals are typically not always empowered. 360 00:35:18,445 --> 00:35:20,288 um 361 00:35:20,288 --> 00:35:33,070 it's difficult from a recruiting perspective to take a CTO from Amazon and hire him at a law firm where he's got stock options and he's fully empowered. 362 00:35:33,070 --> 00:35:41,287 um so I think, I think we kind of need a little bit of a uh reset in a number of different areas, including the governance model. 363 00:35:41,287 --> 00:35:41,647 I don't know. 364 00:35:41,647 --> 00:35:42,678 Do you agree? 365 00:35:43,790 --> 00:35:47,730 I think it's definitely a fascinating idea. 366 00:35:47,730 --> 00:35:50,270 I think that it has some value. 367 00:35:51,310 --> 00:36:03,070 I would still be wary if I were the CTO of Amazon of joining a startup like that. 368 00:36:04,210 --> 00:36:10,170 It's still sort of beholden to the lawyers as stakeholders. 369 00:36:10,190 --> 00:36:12,358 And no matter how 370 00:36:12,446 --> 00:36:25,758 much separation you want to build between the traditional law firm and this startup, um the imprint uh of the law firm mentality is still going to be on it. 371 00:36:25,758 --> 00:36:34,465 um Obviously, the law firm will have a vested interest in the services. 372 00:36:35,310 --> 00:36:46,870 being rendered by this startup being sort of top-notch and not reflecting poorly on the bigger brand. 373 00:36:47,310 --> 00:37:03,030 So if you were to say, there's sort of the bottom 20 % of the revenue in your example, you cannot afford having a situation where even a small portion of it is not delivered to the 374 00:37:03,030 --> 00:37:03,930 same story. 375 00:37:04,498 --> 00:37:11,103 same or at least similar standard that your clients are used to. 376 00:37:11,103 --> 00:37:29,015 uh And even identifying that 20 % of revenues is difficult because some of it may well be coming from the same clients that are included in the top 10 % of your revenue. 377 00:37:29,015 --> 00:37:34,238 So you really can't, if you've got a platinum 378 00:37:34,270 --> 00:37:41,915 or uh a client that is uh rated very highly within the organization. 379 00:37:43,497 --> 00:37:47,209 We always regard pretty much all the clients as important. 380 00:37:47,209 --> 00:38:02,810 But um let's say the client that generates the most revenue for the firm, if they generate a lot of revenue and you take sort of a small portion of their work, which is 381 00:38:03,064 --> 00:38:12,678 simple low-level work and you want to be doing it out of the startup, that's where you really get that reputational concern. 382 00:38:12,678 --> 00:38:29,444 um Not to say that oh none of this could be made, and in fact, some law firms are going maybe not in exactly this direction, but they're uh doing something similar. 383 00:38:29,444 --> 00:38:31,845 Cleary comes to mind with ClearyX. 384 00:38:32,309 --> 00:38:32,689 Right. 385 00:38:32,689 --> 00:38:39,771 uh And it's, it's this idea of a playground. 386 00:38:39,771 --> 00:38:44,533 don't, I don't know if it necessarily has to be a separate uh entity. 387 00:38:44,533 --> 00:38:45,433 It could be. 388 00:38:45,433 --> 00:38:47,864 It doesn't have to be. 389 00:38:47,864 --> 00:38:59,597 The idea of a playground where those experiments happen and adoption happens and happens faster than trying to roll it out to the entire firm. 390 00:39:00,101 --> 00:39:02,884 There that I definitely agree with you. 391 00:39:02,925 --> 00:39:19,045 I absolutely agree that that's the way for big law to sort of transform itself and use that pocket of competences um as the um 392 00:39:20,809 --> 00:39:23,571 as the guiding light, right? 393 00:39:23,571 --> 00:39:35,711 If that pocket of competences, technological competences is successful and it does generate revenue and it does keep the clients happy and clients swear by it and say, 394 00:39:35,711 --> 00:39:49,172 actually, you know what, what used to cost us a hundred K now costs us five and we are still getting the same quality and it's much faster and it's 395 00:39:49,770 --> 00:39:52,512 it's technologically advanced. 396 00:39:52,512 --> 00:39:55,794 We understand what you're doing there. 397 00:39:55,854 --> 00:40:12,426 We want to work with you because not only are we getting the um expert advice from your partners on the matters where we need expert advice, we are also getting the grant work, 398 00:40:12,426 --> 00:40:16,168 if you will, handled faster and more efficiently. 399 00:40:16,168 --> 00:40:17,449 That's when 400 00:40:17,505 --> 00:40:21,307 people will absolutely buy into it. 401 00:40:21,368 --> 00:40:25,831 We are talking about the advisory part of the work. 402 00:40:25,831 --> 00:40:40,460 uh The thing that I feel is often overlooked is how um time consuming the administrative part of uh law firm life is. 403 00:40:41,001 --> 00:40:47,465 And that is another area where I think transformation could really be happening. 404 00:40:47,469 --> 00:41:06,954 uh law firms could adopt technology much faster when they're not weighing the gains against um less billable hours uh for the sake of it, but rather um when they are 405 00:41:07,614 --> 00:41:15,607 essentially saying we are saving, you know, all the time saved or most of the time saved is non-billable time, right? 406 00:41:15,607 --> 00:41:17,357 We actually have more 407 00:41:17,357 --> 00:41:19,900 time to dedicate to our clients. 408 00:41:20,242 --> 00:41:23,206 And that's a winning proposition. 409 00:41:23,427 --> 00:41:29,774 The business of law uh through digitization is definitely a winning proposition. 410 00:41:29,774 --> 00:41:30,854 Yeah. 411 00:41:30,894 --> 00:41:45,134 So speaking of that transition from kind of hourly billing to AFA work, you know, I have heard people throw out, oh, the answer is outcome-based billing or it's value-based 412 00:41:45,134 --> 00:41:46,054 billing. 413 00:41:46,474 --> 00:41:48,154 That's not the end of the conversation. 414 00:41:48,154 --> 00:41:50,494 That's the beginning of the conversation. 415 00:41:50,974 --> 00:41:51,654 Yes. 416 00:41:51,654 --> 00:41:55,774 Because there's so much devil in the details around that. 417 00:41:55,774 --> 00:41:57,914 So you can't just say, 418 00:41:58,199 --> 00:42:09,625 I will, we will take some reasonable percentage of the value that we deliver because what that neglects is supply and demand. 419 00:42:09,926 --> 00:42:15,909 Where do you draw that line of, let's say you've delivered a million dollars worth of value. 420 00:42:15,909 --> 00:42:17,370 Well, we're going to ask for 20%. 421 00:42:17,370 --> 00:42:24,774 Okay, oh well, what's the law firm down the street going to ask for who does comparable? 422 00:42:24,942 --> 00:42:25,952 comparable quality. 423 00:42:25,952 --> 00:42:29,423 There's just so many variables. 424 00:42:29,583 --> 00:42:31,874 I don't see any real easy answers. 425 00:42:31,874 --> 00:42:42,407 And I read an article um about tech fees associated with um AI infrastructure. 426 00:42:42,407 --> 00:42:45,248 That's a whole other set of problems. 427 00:42:45,248 --> 00:42:50,370 um You they talked about day-to-day tools like maybe Harvey and Legora. 428 00:42:50,370 --> 00:42:51,910 um 429 00:42:51,916 --> 00:42:54,368 You know, you treat those like you do Microsoft Word. 430 00:42:54,368 --> 00:43:09,421 You don't bill your clients for Microsoft Word, but if you're doing some specialized task that is very resource or tech intensive, maybe that could be something that you bill for 431 00:43:09,421 --> 00:43:11,083 separately in addition to your hours. 432 00:43:11,083 --> 00:43:18,249 So the pricing thing is really fascinating to me and I really don't feel that there are lots any easy answers out there. 433 00:43:18,249 --> 00:43:18,569 I don't know. 434 00:43:18,569 --> 00:43:20,190 What is your take on that? 435 00:43:21,503 --> 00:43:30,680 There will be multiple models and I do not think that there is one right or wrong answer really. 436 00:43:30,680 --> 00:43:32,100 I completely agree with you, Ted. 437 00:43:32,100 --> 00:43:50,533 I think that uh lawyers love analogies, so you indulge me and allow me to say that law firms have offered, have always offered other services in addition to um 438 00:43:50,655 --> 00:43:52,185 There are primary services. 439 00:43:52,185 --> 00:43:57,167 um In some cases, that's actually part of the revenue. 440 00:43:57,167 --> 00:44:12,090 In some cases, that's a value add that is provided to the clients uh for free um to build goodwill and strengthen the relationship, all of that. 441 00:44:12,090 --> 00:44:19,972 uh For as long as those services were uh ancillary, 442 00:44:19,986 --> 00:44:35,902 to the main work and did not directly impact on the service delivery, it certainly made sense that most of those were provided for free. 443 00:44:36,262 --> 00:44:41,084 Now I think we're entering into a different world where 444 00:44:41,768 --> 00:44:54,834 It's almost like clients might have a choice, and in fact they do have a choice in certain situations where you might say to the clients, we can do this uh due diligence document 445 00:44:54,834 --> 00:44:57,775 review the traditional way. 446 00:44:57,775 --> 00:45:01,637 We'll put an army of uh junior lawyers on this. 447 00:45:01,637 --> 00:45:06,199 They will sort of bill X number of hours. 448 00:45:06,199 --> 00:45:10,020 We will negotiate with you and that's going to be 449 00:45:11,747 --> 00:45:15,208 $500,000 for the sake of the argument, right? 450 00:45:17,028 --> 00:45:27,628 Or you can have it done by a tool that we've bought, that we are using, we are subscribed for it. 451 00:45:28,228 --> 00:45:31,568 We think it's a good tool, we like it. 452 00:45:31,968 --> 00:45:34,068 The accuracy is there. 453 00:45:34,068 --> 00:45:40,630 And then you can either ask us to double check, do a selective 454 00:45:40,630 --> 00:45:55,397 verification process, or you might say, you know what, none of this matters to me, I'm just happy for you uh to run the tool with no additional uh lawyer time added on top of 455 00:45:55,397 --> 00:45:56,287 that fee. 456 00:45:56,287 --> 00:45:58,308 um 457 00:45:59,923 --> 00:46:10,816 I will posit, I will sort argue that in these three situations, you have, know, and you still have the tool in the background. 458 00:46:10,816 --> 00:46:24,449 So if you're running that analysis, that due diligence, the traditional way, you might still enable your team to use the tool to do sort of the first run through the due 459 00:46:24,449 --> 00:46:26,559 diligence materials, right? 460 00:46:26,559 --> 00:46:29,660 ah Do you... 461 00:46:31,365 --> 00:46:41,573 Do you charge the client less for the traditional services because your people spent 80 hours as opposed to 100 hours on it? 462 00:46:41,573 --> 00:46:50,721 ah Do you charge the client more for the time it takes to review in the second scenario? 463 00:46:50,721 --> 00:47:00,012 Because you're like, actually, that review needs to be done not by junior lawyers, but by people who are more senior who can actually spot 464 00:47:00,012 --> 00:47:03,754 ah the things that the solution might have missed. 465 00:47:03,754 --> 00:47:20,091 ah And do you charge sort of, do you pass on the service, the tool at cost in the third scenario, or do you actually say, well, we are losing the revenue from the traditional 466 00:47:20,091 --> 00:47:20,501 way. 467 00:47:20,501 --> 00:47:27,944 um So that tool will cost you something, or you can go ahead and. 468 00:47:27,944 --> 00:47:31,584 get a subscription yourself and run it on the data room. 469 00:47:31,927 --> 00:47:33,588 Those are the... 470 00:47:33,729 --> 00:47:37,272 And once again, we're just scratching the surface here, right? 471 00:47:37,272 --> 00:47:50,062 You literally have all kinds of possibilities and all kinds of business models, and some firms will be heavily reliant on the technological tools and others will not. 472 00:47:50,062 --> 00:47:56,287 They will prefer to uh stick to the traditional way for as long as they can. 473 00:47:56,287 --> 00:47:57,388 And, you know, that... 474 00:47:57,388 --> 00:48:03,388 that has to do with the risk aversion and the... 475 00:48:04,940 --> 00:48:10,665 traditionalism and conservatism, healthy conservatism of the legal profession. 476 00:48:10,668 --> 00:48:12,539 Yeah, that makes sense. 477 00:48:12,539 --> 00:48:15,060 Well, I hate that we're out of time. 478 00:48:15,060 --> 00:48:18,562 I have thoroughly enjoyed this conversation. 479 00:48:19,983 --> 00:48:34,381 it's not, we have lots of lawyers on the show, but I think your role as having been a longtime partner at a major firm and now investing in the space gives some interesting 480 00:48:34,381 --> 00:48:35,972 perspective there. 481 00:48:35,972 --> 00:48:40,314 Because I see varying levels of buy-in. 482 00:48:40,450 --> 00:48:46,913 on just how transformational this tech is going to be in the industry. 483 00:48:46,913 --> 00:48:49,054 And I feel like you get it. 484 00:48:49,054 --> 00:48:54,155 um have lots of friends in the space having been around for a while. 485 00:48:54,356 --> 00:49:07,461 And ones who are really smart and who I respect to think I'm over indexing on the transformational magnitude, I don't think that I am. 486 00:49:07,461 --> 00:49:09,964 um 487 00:49:09,964 --> 00:49:17,194 You know, law firms have been successful being the way that they've been for a really long time. 488 00:49:17,194 --> 00:49:19,668 I just think, yeah, exactly. 489 00:49:19,668 --> 00:49:21,119 I think that gig is up though. 490 00:49:21,119 --> 00:49:21,879 I really do. 491 00:49:21,879 --> 00:49:23,620 Now, how soon? 492 00:49:23,874 --> 00:49:28,863 I had a guest on yesterday um who runs a VC. 493 00:49:28,863 --> 00:49:31,994 He's a partner at TLTF. 494 00:49:31,994 --> 00:49:36,226 And he was hesitant to lay out a timeline. 495 00:49:36,745 --> 00:49:41,217 I have less to lose and if I'm wrong, who cares? 496 00:49:41,217 --> 00:49:46,479 But I think that there will be no impact on revenues this year. 497 00:49:46,479 --> 00:49:54,842 I think very modest in 2026, but I think in 2027, there will be non-trivial, there will be real impact. 498 00:49:54,842 --> 00:50:02,765 um And that is based on how fast I see this technology moving. 499 00:50:02,765 --> 00:50:05,506 And I speak to a lot of 500 00:50:05,666 --> 00:50:15,789 GCs and I they there is so much pent up demand for efficiency From their law firm partners. 501 00:50:15,789 --> 00:50:28,633 I mean, it's they're screaming for it, you know as rates continue to go up so I don't know last thought does that seem like a reasonable timeline in terms of When we'll see impact 502 00:50:28,633 --> 00:50:30,734 or you think it'll be sooner longer 503 00:50:34,031 --> 00:50:37,392 My prophetic capabilities are limited. 504 00:50:37,731 --> 00:50:43,211 I don't necessarily know what the timeline is. 505 00:50:43,211 --> 00:50:47,057 It sort of does depend on a lot of factors. 506 00:50:47,537 --> 00:50:53,960 I will agree with you wholeheartedly that the demand from the clients is there. 507 00:50:53,960 --> 00:50:59,947 um I will also say that there will be an impact. 508 00:50:59,947 --> 00:51:12,038 ah on revenues, law firm revenues, uh I might be a bit more contrarian in terms of the gig is up. 509 00:51:12,038 --> 00:51:25,991 I actually think that there may well be some scenarios where law firms will improve their revenues, or at least some law firms will improve their revenues by quite a bit. 510 00:51:26,435 --> 00:51:28,356 utilizing technology. 511 00:51:28,437 --> 00:51:35,462 it's more a question of if you are doing nothing with it, then you could be in trouble. 512 00:51:35,462 --> 00:51:51,254 ah If you are actually actively monitoring what's happening in the space, and you're looking for ways to uh bring technology, sort of enlist the services of technology and um 513 00:51:51,254 --> 00:51:53,175 get it right, 514 00:51:53,189 --> 00:51:57,041 then you might be one of the beneficiaries. 515 00:51:57,041 --> 00:52:00,642 And there will be more than one sort group of beneficiaries. 516 00:52:00,822 --> 00:52:06,014 So look, I wish I had the proverbial crystal ball. 517 00:52:06,014 --> 00:52:07,505 I don't. 518 00:52:07,505 --> 00:52:14,648 But um the impact will be felt within the next couple of years, definitely. 519 00:52:14,648 --> 00:52:21,340 Whether it's going to be sort of a negative impact or a positive impact, I really do think it will matter. 520 00:52:21,340 --> 00:52:22,811 Either it will depend on 521 00:52:23,251 --> 00:52:27,354 how people approach uh this whole situation. 522 00:52:28,435 --> 00:52:31,052 The ostrich approach is not the right approach. 523 00:52:31,052 --> 00:52:33,163 Well, and that's what I mean by the gig is up. 524 00:52:33,163 --> 00:52:44,528 The law firms not across the board, but Many many law firms have been extremely resistant to tech, even the cloud. 525 00:52:44,528 --> 00:52:45,828 And that gig is up. 526 00:52:45,828 --> 00:52:46,458 Yeah. 527 00:52:46,458 --> 00:52:49,460 There's that, that is no longer going to fly that I'm certain of. 528 00:52:49,460 --> 00:52:52,070 I'll bet you my, I'll bet you the company. 529 00:52:52,391 --> 00:52:53,091 Yeah. 530 00:52:53,091 --> 00:52:54,012 Yeah. 531 00:52:54,012 --> 00:52:58,873 So yeah, it's been, we're going to make changes here. 532 00:52:58,873 --> 00:53:00,414 It's going to happen. 533 00:53:00,416 --> 00:53:04,847 and law firms are going to have to make investments, it's not optional. 534 00:53:04,847 --> 00:53:13,190 um When the chickens will come home to roost, that's up for debate, but they're coming. 535 00:53:13,190 --> 00:53:17,671 um So it's a fun time to be around. 536 00:53:17,671 --> 00:53:19,352 So this is a great conversation. 537 00:53:19,352 --> 00:53:21,372 really, I really enjoyed it. 538 00:53:21,372 --> 00:53:27,444 For folks that want to find out more about, you know, I think you have some great commentary on LinkedIn. 539 00:53:27,444 --> 00:53:28,524 I haven't seen 540 00:53:28,672 --> 00:53:32,513 actual posts, but I've really enjoyed some of the comments that you made. 541 00:53:32,877 --> 00:53:36,085 Is that the best way for people to connect with you? 542 00:53:36,764 --> 00:53:38,105 Or they can just write to me. 543 00:53:38,105 --> 00:53:40,143 My email is on LinkedIn as well. 544 00:53:40,143 --> 00:53:44,798 I'm always happy to speak and thank you so much for having me on the podcast. 545 00:53:44,798 --> 00:53:46,671 I really enjoyed talking to you. 546 00:53:46,671 --> 00:53:47,751 Awesome. 547 00:53:47,851 --> 00:53:48,231 All right. 548 00:53:48,231 --> 00:53:50,731 Well, I'm sure we'll bump into each other. 549 00:53:51,731 --> 00:53:57,411 I do a lot of conferences out there, so I'm sure we'll bump into each other in person. 550 00:53:57,531 --> 00:54:01,471 But until then, thank you very much, and we'll chat again soon. 551 00:54:02,890 --> 00:54:03,311 All righty. 552 00:54:03,311 --> 00:54:04,251 Take care. 00:00:04,692 Lev, how are you this afternoon? 2 00:00:05,263 --> 00:00:06,649 I'm good, Ted. 3 00:00:06,649 --> 00:00:07,780 Yourself? 4 00:00:07,918 --> 00:00:08,959 I'm great. 5 00:00:08,959 --> 00:00:10,439 I'm doing great. 6 00:00:10,560 --> 00:00:13,521 I'm really looking forward to this conversation. 7 00:00:13,861 --> 00:00:26,869 You and I, this is very often the case on how I find guests for the podcast is I see content or I have interactions on LinkedIn that align with what I know our audience likes 8 00:00:26,869 --> 00:00:27,709 to hear about. 9 00:00:27,709 --> 00:00:30,231 And you and I were on a thread. 10 00:00:30,231 --> 00:00:36,224 think it was, was that the Andrew Wang um meme or yeah. 11 00:00:37,292 --> 00:00:45,287 where he basically said, you know, AI is already replacing third year associate level work. 12 00:00:45,287 --> 00:00:51,910 And there was a lot of debate, um good debate, but I liked a lot of the things you were saying. 13 00:00:51,910 --> 00:00:57,653 So I'm looking forward to the conversation, but before we jump in, let's get you introduced. 14 00:00:57,653 --> 00:01:05,037 So you were a former &A partner at Linklater's for 20 plus years and doing corporate law work. 15 00:01:05,037 --> 00:01:06,126 um 16 00:01:06,126 --> 00:01:08,326 You're an economist by training. 17 00:01:08,546 --> 00:01:12,606 You got a lot of deep expertise in digital transformation. 18 00:01:13,006 --> 00:01:14,406 Fill in the gaps for us. 19 00:01:14,406 --> 00:01:19,406 Tell us about what your background is and what you're up to today. 20 00:01:19,731 --> 00:01:21,451 Not much to add. 21 00:01:21,451 --> 00:01:24,250 You summed it up perfectly, really. 22 00:01:24,271 --> 00:01:26,371 I am a lawyer. 23 00:01:26,371 --> 00:01:31,391 I spent 20 years in the legal profession working at Sullivan and Cromwell. 24 00:01:31,391 --> 00:01:32,711 That's where I started out. 25 00:01:32,711 --> 00:01:35,051 Then I moved to Link Laters. 26 00:01:35,311 --> 00:01:38,931 I did cross-border work, M &A. 27 00:01:39,211 --> 00:01:43,595 Worked in London, worked in Moscow, worked in New York. 28 00:01:43,860 --> 00:01:56,640 And about a year ago, I resigned from LinkLators to become a legal tech investor, a legal tech angel investor for the time being. 29 00:01:56,860 --> 00:01:59,460 I invest in various legal tech startups. 30 00:01:59,460 --> 00:02:04,120 I advise legal tech startups, formally and informally. 31 00:02:04,120 --> 00:02:08,160 And I really like the space. 32 00:02:08,160 --> 00:02:09,460 What can I say? 33 00:02:09,520 --> 00:02:13,440 That's what brings me to all those threads about 34 00:02:13,795 --> 00:02:16,463 AI and legal and innovation. 35 00:02:17,102 --> 00:02:22,524 Well, so you've got an interesting perspective as somebody who practiced law for so many years. 36 00:02:22,524 --> 00:02:36,288 And um what is your take on AI's capabilities as it relates to the area where you spent most of your time in transactional work? 37 00:02:36,489 --> 00:02:40,050 Is it replacing? 38 00:02:40,406 --> 00:02:53,506 work done by attorneys today, if not how soon in the future will it be taking at least some of those more mundane tasks and completing the lion's share of the work. 39 00:02:54,449 --> 00:03:09,445 I genuinely think that the capabilities of technology, even today, and I think that we're in the early days, I think the capabilities are unimaginably impressive. 40 00:03:09,725 --> 00:03:11,186 Unimaginably impressive. 41 00:03:11,186 --> 00:03:20,010 ah I'm not thinking about it in terms of replacing lawyers or not replacing lawyers. 42 00:03:20,010 --> 00:03:22,351 That's a debate that we can have. 43 00:03:22,351 --> 00:03:23,663 ah 44 00:03:23,663 --> 00:03:37,447 It's more about can a system, can a tool ah do work that is comparable to a junior associate or a mid-level associate? 45 00:03:37,447 --> 00:03:42,148 And the answer to that is absolutely it can, right? 46 00:03:42,148 --> 00:03:43,448 That's the truth of it. 47 00:03:43,448 --> 00:03:53,591 The best solutions out there without any question can definitely do the same work that a junior associate can do. 48 00:03:53,907 --> 00:04:00,010 ah And many of the better ones can also do the work that a mid-level associate can do. 49 00:04:00,010 --> 00:04:07,494 um The pace of changes is what really strikes me. 50 00:04:07,494 --> 00:04:23,343 um I'd say two years ago ah when I just started looking at this, um it was difficult to imagine, difficult to fathom that uh 51 00:04:23,343 --> 00:04:25,845 we'd be where we are today, right? 52 00:04:25,845 --> 00:04:32,449 ah The early chatbots work and have hopelessly riddled with uh hallucinations. 53 00:04:32,449 --> 00:04:41,596 uh It felt like the guardrails would uh take years to make and put in place. 54 00:04:41,596 --> 00:04:46,179 ah The reality is different, right? 55 00:04:46,179 --> 00:04:50,462 Philosophically, what we do as lawyers is we work with text. 56 00:04:50,462 --> 00:04:52,663 We work with a lot of text. 57 00:04:52,849 --> 00:05:02,423 I'm thinking transactional, um Litigators work with all kinds of stuff, uh images, audio and all of that. 58 00:05:02,624 --> 00:05:17,051 But the products that I'm seeing, can do due diligence, for instance, ah much faster, much cheaper and more accurately than ah a big team of junior associates. 59 00:05:17,051 --> 00:05:18,091 So it's... 60 00:05:18,277 --> 00:05:29,260 you get a due diligence report in a matter of hours as opposed to a month of a big team with sort of numerous iterations. 61 00:05:29,742 --> 00:05:32,284 I don't know if that answers your question. 62 00:05:32,445 --> 00:05:33,903 What do you think about this? 63 00:05:33,903 --> 00:05:47,223 Yeah, so I have been using AI side by side on legal work for my business for about six months. 64 00:05:47,603 --> 00:05:51,383 And it started with a simple labor and employment matter. 65 00:05:51,383 --> 00:05:56,883 We were hiring somebody from a firm with a little bit of overlap, and they had some non-compete language. 66 00:05:56,883 --> 00:06:01,683 And we hired a L &E attorney to advise us. 67 00:06:01,702 --> 00:06:05,664 And in parallel, I asked all the questions. 68 00:06:05,664 --> 00:06:13,869 I wrote down all the questions that I asked the attorney about how we keep from running afoul of that non-compete. 69 00:06:14,129 --> 00:06:18,811 And I asked the attorney, and then I asked AI. 70 00:06:18,972 --> 00:06:24,715 And the answers were very close. 71 00:06:24,715 --> 00:06:26,296 the added benefit that I got. 72 00:06:26,296 --> 00:06:27,667 And this was my first experiment. 73 00:06:27,667 --> 00:06:30,278 I've got lots of examples, but. 74 00:06:30,610 --> 00:06:35,192 What the AI did better was roleplay. 75 00:06:35,292 --> 00:06:43,075 So I have a custom GPT that's trained on, I call it my co-CEO. 76 00:06:43,075 --> 00:06:55,801 So it has as training materials, it has our core values, it has our pitch deck, it has our quarterly all hands meeting notes and deck. 77 00:06:55,801 --> 00:06:59,062 has, we do uh periodic 78 00:06:59,062 --> 00:07:03,144 retreats with leadership where we map out the next quarter. 79 00:07:03,485 --> 00:07:06,126 It has all that information as training materials. 80 00:07:06,126 --> 00:07:07,767 So then I role played with it. 81 00:07:07,767 --> 00:07:08,908 It has our org chart. 82 00:07:08,908 --> 00:07:10,209 I mean, it has a bunch of stuff. 83 00:07:10,209 --> 00:07:12,030 It knows everything about us. 84 00:07:12,030 --> 00:07:16,152 It even has our operating agreement and corporate docs in there. 85 00:07:16,152 --> 00:07:20,595 I asked it the role playing part because it knew us so well. 86 00:07:20,595 --> 00:07:22,886 And this was the first time that we had used this attorney. 87 00:07:22,886 --> 00:07:24,417 And he did a great job, by the way. 88 00:07:24,417 --> 00:07:25,958 um 89 00:07:26,370 --> 00:07:30,772 But when we role played and said, OK, how can we construct this role? 90 00:07:30,772 --> 00:07:34,133 And again, it had the language from the non-compete. 91 00:07:34,133 --> 00:07:36,444 It did a fantastic job. 92 00:07:36,444 --> 00:07:40,196 It wrote the job description for me because it knew my business. 93 00:07:40,196 --> 00:07:41,936 It knew where we had needs. 94 00:07:41,936 --> 00:07:50,970 It even placed it on the org chart in the reporting structure that, made the most sense because I had the guy's CV that I also uploaded. 95 00:07:50,970 --> 00:07:52,721 Like, these are his capabilities. 96 00:07:52,801 --> 00:07:53,881 Because we want to be careful. 97 00:07:53,881 --> 00:07:56,202 Like, as a young company, last thing you want to get 98 00:07:56,291 --> 00:08:00,162 do is get a lawsuit against you um for something like that. 99 00:08:00,162 --> 00:08:04,613 So it was incredible with how the role playing part. 100 00:08:04,834 --> 00:08:18,758 And um I think it was a uh function of I was able to train it on our organization, whereas it would have taken a lot of time for me to train the attorney to really understand us. 101 00:08:20,326 --> 00:08:21,467 No doubt about that. 102 00:08:21,467 --> 00:08:24,789 And you're not uh the exception. 103 00:08:24,789 --> 00:08:40,828 There's a lot of clients who use uh either out-of-the-box tools or custom-made, know, customized GPTs uh to either double-check the legal work or, frankly, do the legal work 104 00:08:40,828 --> 00:08:42,018 themselves. 105 00:08:43,319 --> 00:08:49,162 And then they have uh law firms just review that and sign on. 106 00:08:49,555 --> 00:08:53,058 We're not talking about multi-billion dollar transactions. 107 00:08:53,058 --> 00:08:57,762 Fine, that's probably a few years away, right? 108 00:08:57,762 --> 00:09:15,036 But uh we are talking about uh basic uh questions that uh sometimes take people a week or two to get answers to, because attorneys are busy and especially if it's a small question, 109 00:09:15,036 --> 00:09:16,297 you may not get... 110 00:09:16,297 --> 00:09:17,418 uh 111 00:09:17,752 --> 00:09:21,924 the answer right away, right? 112 00:09:22,004 --> 00:09:25,715 You need to take a spot on the queue, right? 113 00:09:25,715 --> 00:09:31,308 And then by the time the attorney gets to your question, that's when you'll get the response. 114 00:09:31,308 --> 00:09:36,980 ah So people kind of are circumventing the queue, right? 115 00:09:36,980 --> 00:09:42,302 They circumvent, you know, they go around the queue and they get the answer. 116 00:09:42,302 --> 00:09:46,874 They ask the attorney to spend sort of a few minutes to just take a look at the 117 00:09:47,089 --> 00:09:52,521 response and bless it or curse it for that matter. 118 00:09:52,521 --> 00:10:08,517 uh the other thing is uh law firms, uh AI enabled law firms, Traditional ones uh have started using it to great effect. 119 00:10:08,517 --> 00:10:17,050 I just saw a post by uh Michael Huseby uh from the investment lawyers, the fund formation lawyers. 120 00:10:17,170 --> 00:10:29,950 who said that he routinely now checks at least two tools before sort of delving into the source material himself. 121 00:10:30,730 --> 00:10:34,670 it's not, you know, Michael is not the exception either. 122 00:10:34,670 --> 00:10:46,370 He's at the cutting edge of the profession, I would say, but it is becoming a lot more common for people to do that. 123 00:10:47,582 --> 00:10:49,523 And the responses are good. 124 00:10:49,523 --> 00:10:53,826 That's the wonderful thing and the scary thing, right? 125 00:10:53,826 --> 00:10:55,606 The responses are good. 126 00:10:55,606 --> 00:11:12,538 um The thing that really brought me to this space, one of the conversations that really brought me to this space was actually in the early days when um Chad GPC just came out, I 127 00:11:12,538 --> 00:11:13,508 had a... 128 00:11:13,818 --> 00:11:27,468 with a friend of mine, a very experienced capital markets partner, uh and I asked him whether he um had experimented with ChetGPC, and he told me, yeah, I was actually short on 129 00:11:27,468 --> 00:11:35,194 time, and I needed to prepare for this management interview, management due diligence uh interview. 130 00:11:35,194 --> 00:11:40,988 um So I asked ChetGPC if for 131 00:11:40,988 --> 00:11:42,739 for kicks, right, for the hell of it. 132 00:11:42,739 --> 00:11:55,774 um I asked that to prepare an agenda, right, a list of questions to ask without betraying any client confidences, without the name of the client just said, this is the company that 133 00:11:55,774 --> 00:11:59,985 is um about to issue some public debt. 134 00:11:59,985 --> 00:12:04,127 uh What kind of questions should I ask? 135 00:12:04,147 --> 00:12:08,168 I am an experienced attorney, you know, know the drill, right? 136 00:12:08,168 --> 00:12:10,349 The prompt is simple enough. 137 00:12:10,351 --> 00:12:14,525 And he said that out of the... 138 00:12:14,525 --> 00:12:16,967 he limited it to something like 20 questions. 139 00:12:16,967 --> 00:12:26,394 uh He said out of the 20 questions, 15 were completely on point, um things that he would have asked himself. 140 00:12:26,635 --> 00:12:32,519 Two or three were probably questions that he wouldn't ask, but they weren't horrible. 141 00:12:32,720 --> 00:12:39,265 And two or three were actually things that he didn't think about, but he should have asked. 142 00:12:39,901 --> 00:12:45,146 And that's something that he has been doing his entire life. 143 00:12:45,587 --> 00:12:51,092 If at the very early stage the technology is such that it can... 144 00:12:55,437 --> 00:13:04,450 suggests something new to very experienced lawyer, then there is really something to that technology that uh holds tremendous potential. 145 00:13:04,451 --> 00:13:11,971 And uh that was one of the reasons that kind of pushed me into this uh current role. 146 00:13:12,130 --> 00:13:12,751 Yeah. 147 00:13:12,751 --> 00:13:17,935 You know, another thing I used it on is, uh, I experiment with a lot of the different models. 148 00:13:17,935 --> 00:13:24,339 There's really kind of four in my rotation now, Grok, Claude, Gemini, and Chagy BT. 149 00:13:25,321 --> 00:13:29,684 and I used Grok for this and it did really well when Grok four came out. 150 00:13:29,684 --> 00:13:32,667 It was, uh, that's when I started paying more attention. 151 00:13:32,667 --> 00:13:41,794 I took our four nine a program, which is like around kind of shadow equity, um, kind of stock option incentives and 152 00:13:42,227 --> 00:13:54,290 that had been drawn up by a lawyer and I ran it through and asked, I used to be an auditor and I worked for Bank of America in corporate audit, doing anti-money laundering audit and 153 00:13:54,290 --> 00:13:55,650 technology audit. 154 00:13:55,650 --> 00:14:03,002 So I've got a risk management background and I coached it to go through the entire document. 155 00:14:03,002 --> 00:14:10,934 It's about 40 pages long and give me a risk priority number, like identify every risk in the 156 00:14:11,022 --> 00:14:17,587 in the program and then risk it something called RPN we use in the audit world, risk priority number. 157 00:14:17,587 --> 00:14:22,971 It's um what is the magnitude of the risk? 158 00:14:22,971 --> 00:14:24,712 Should it materialize? 159 00:14:24,712 --> 00:14:28,695 It's severity detection and likely of occurrence. 160 00:14:28,695 --> 00:14:34,900 So I had it multiply those three things and give it a numeric score and rank it in descending order. 161 00:14:34,900 --> 00:14:37,782 So the highest risk came up top. 162 00:14:37,974 --> 00:14:49,671 And it gave me detailed breakdown in each, you know, what the likelihood of severity, um what is the likelihood of occurrence and what is the likelihood of detection? 163 00:14:49,671 --> 00:14:53,003 How would you know if this risk were to materialize? 164 00:14:53,003 --> 00:14:55,785 In the risk management world, that's uh a thing. 165 00:14:55,785 --> 00:15:01,358 um And it did incredible work helping us. 166 00:15:01,358 --> 00:15:03,410 I took that output, handed it to my lawyer. 167 00:15:03,410 --> 00:15:04,410 He was like, 168 00:15:04,726 --> 00:15:12,650 Okay, yeah, this is, this is a good starting point because the top three things on here, we ought to, uh, we ought to remediate this, this and this. 169 00:15:12,650 --> 00:15:20,355 Don't worry about not really an issue given the context, but I was, this was an untrained, this was not a legal specific tool. 170 00:15:20,355 --> 00:15:27,739 wasn't using a custom GPT like in my, my co CEO that knows my business and it hit it out of the park. 171 00:15:27,739 --> 00:15:34,042 I've got dozens of examples, like just over the last six months, and I'm a small buyer of legal services. 172 00:15:34,126 --> 00:15:35,646 I don't know what we spend. 173 00:15:35,646 --> 00:15:40,226 mean, it's more than six figures a year, but it's not a huge, it's not millions of dollars. 174 00:15:40,326 --> 00:15:53,766 Um, and just in six months, I have probably, I don't know, I'm guessing eight to 10 really compelling examples of where, uh, it saved me money, time, or both. 175 00:15:55,670 --> 00:15:57,010 Not surprised at all. 176 00:15:57,010 --> 00:15:59,391 Not surprised at all. 177 00:16:00,551 --> 00:16:12,694 will say ah that foundational models, I'd be wary of using them uh for Betta Farm kinds of things. 178 00:16:12,934 --> 00:16:13,414 True. 179 00:16:13,414 --> 00:16:23,277 uh For uh basic level stuff, for stuff that is uh not Betta Farm. 180 00:16:24,497 --> 00:16:28,857 The way you are using it is extremely, extremely right. 181 00:16:28,877 --> 00:16:37,817 So you're basically asking questions, you're getting the input, and then you have it double-checked by your professionals. 182 00:16:38,037 --> 00:16:41,617 That is certainly the way to go. 183 00:16:42,757 --> 00:16:53,257 There are solutions either being developed or that have been developed that I would say take this a bit further. 184 00:16:53,385 --> 00:17:07,256 And it's no longer necessary for you to have a deep understanding of what you're looking for, what you actually want to take to your lawyer, to your outside counsel. 185 00:17:07,256 --> 00:17:11,359 ah It will be sort of automatic. 186 00:17:11,520 --> 00:17:13,802 It will ask the right questions. 187 00:17:13,802 --> 00:17:15,133 The playbooks are there. 188 00:17:15,133 --> 00:17:16,444 The workflows are there. 189 00:17:16,444 --> 00:17:22,989 uh All you need to do is provide the information to the tool. 190 00:17:23,057 --> 00:17:31,871 um Importantly enough, the better tools are kind of going the agents accrued. 191 00:17:31,871 --> 00:17:41,965 Everybody's talking about the agents accrued and they are flagging ah that something ought to be reviewed. 192 00:17:42,125 --> 00:17:42,585 Right? 193 00:17:42,585 --> 00:17:52,451 So if somebody gets a document in an email, the system will recognize that it's a legal document and it will say, do you want 194 00:17:52,451 --> 00:17:56,003 the system to kind of run some kind of an analysis. 195 00:17:56,003 --> 00:17:58,924 Do you want to set the parameters of the analysis? 196 00:17:58,924 --> 00:18:02,486 Do you want to rely on the preset parameters of the analysis? 197 00:18:03,007 --> 00:18:06,708 And those are impressive, very impressive. 198 00:18:06,708 --> 00:18:22,056 So uh high level analysis is definitely something that is already there and it's at least comparable to the average lower uh 199 00:18:22,242 --> 00:18:24,614 of the relevant seniority. 200 00:18:24,614 --> 00:18:33,831 ah The in-depth solutions are things that are still being built. 201 00:18:33,831 --> 00:18:40,317 The ecosystem is still in its nascency, It's infancy, I should say. 202 00:18:40,317 --> 00:18:42,168 It's very early days. 203 00:18:42,168 --> 00:18:51,515 um Despite all the headlines, despite all the great successes that we've seen, 204 00:18:51,696 --> 00:18:58,816 it's still not at all clear who the winners will be. 205 00:19:00,156 --> 00:19:08,876 It's still something that is very much an open race. 206 00:19:08,876 --> 00:19:09,766 Yeah. 207 00:19:09,807 --> 00:19:21,074 You know, when you and I talked last time, you brought up an interesting angle on something that I've been thinking about quite deeply. 208 00:19:21,074 --> 00:19:36,565 that is, and my listeners have heard many times me talk about the foundational fundamental characteristics of the law firm world that create headwinds towards innovation and 209 00:19:36,565 --> 00:19:37,545 transformation. 210 00:19:37,545 --> 00:19:38,886 uh 211 00:19:39,353 --> 00:19:43,686 Transformation and innovation um aren't necessarily the same thing. 212 00:19:43,686 --> 00:19:48,979 um Transformation has much greater magnitude. 213 00:19:48,979 --> 00:19:50,890 Innovation can lead to transformation. 214 00:19:50,890 --> 00:20:02,077 um But for real transformation, I feel like there is a lot of legacy, cultural barriers. 215 00:20:02,077 --> 00:20:03,616 um 216 00:20:03,616 --> 00:20:12,492 Again, fundamental business structures like the partnership model that very much prioritizes profit taking at year end. 217 00:20:12,713 --> 00:20:17,495 the concept of R &D in a law firm is a bit foreign. 218 00:20:17,556 --> 00:20:22,339 They don't tend to do traditional R &D. 219 00:20:22,339 --> 00:20:27,006 And you brought up a good point that I want to expand on a little bit. 220 00:20:27,006 --> 00:20:30,605 I guess let me also say the 221 00:20:32,150 --> 00:20:40,432 adversity towards risk that lawyers that mindset tends to have as they're trained to be risk averse. 222 00:20:40,612 --> 00:20:52,916 Um, and you brought up a good point, which I hadn't really thought about before, which was law firms make investments that are risky all the time in like opening up new offices and 223 00:20:52,916 --> 00:20:53,486 areas. 224 00:20:53,486 --> 00:20:55,557 And I thought that was a good counterpoint. 225 00:20:55,557 --> 00:21:00,918 I still think my, I still think my original point is valid. 226 00:21:01,098 --> 00:21:08,024 in terms of traditional R &D, like explain how law firms take risks. 227 00:21:08,024 --> 00:21:09,545 I mean, that was one good example. 228 00:21:09,545 --> 00:21:10,606 We can expand on that. 229 00:21:10,606 --> 00:21:15,720 There's maybe others like they do take risks, correct? 230 00:21:16,154 --> 00:21:17,104 They do, they do. 231 00:21:17,104 --> 00:21:22,066 And just to be clear, Ted, I was not disagreeing with you. 232 00:21:22,066 --> 00:21:25,427 Everything you said is valid, right? 233 00:21:25,427 --> 00:21:27,447 Lawyers are risk averse. 234 00:21:27,628 --> 00:21:30,529 The partnership model is the partnership model. 235 00:21:30,529 --> 00:21:40,953 uh Where I sort of diverged a little bit from what you and a lot of others are saying is, uh 236 00:21:40,953 --> 00:21:59,137 This notion that law firms are just uh short-termist uh profits at the end of the year is ah the ultimate goal, and there is sort of no recognition or no ability to make 237 00:21:59,137 --> 00:21:59,877 investments. 238 00:21:59,877 --> 00:22:10,820 um Law firms make investments all the time, whether it's opening up a new office, hiring a very 239 00:22:11,980 --> 00:22:16,643 impressive lateral partner or a group of lateral partners, right? 240 00:22:16,944 --> 00:22:27,151 And it's the same problem with consensus building that people are talking about when they're talking about uh technology adoption, right? 241 00:22:27,151 --> 00:22:40,610 um How do you bring on board somebody who doesn't see immediate benefits from that technology in his or her own practice, right? 242 00:22:40,760 --> 00:22:47,563 A valid point, true, some people will benefit more and some people will benefit less. 243 00:22:48,284 --> 00:23:00,571 However, it is absolutely the same thing when I'm at a global law firm and somebody says, we should open an office in Indonesia. 244 00:23:00,671 --> 00:23:09,776 If my practice is very much US centric or UK centric and my clients are not looking at Indonesia, 245 00:23:09,776 --> 00:23:15,578 to invest in that uh jurisdiction, how am I going to benefit? 246 00:23:15,578 --> 00:23:16,659 I won't. 247 00:23:17,259 --> 00:23:27,623 And yet there is the notion that somehow the firm is bigger than individual partners or individual practice groups. 248 00:23:27,623 --> 00:23:39,468 uh And it's um the same with uh lateral partners or electing the current cohort of partners, homegrown partners. 249 00:23:39,504 --> 00:23:40,284 It's the same thing. 250 00:23:40,284 --> 00:23:45,127 ah You don't necessarily know how many of them are going to be successful. 251 00:23:45,127 --> 00:23:53,610 You don't know if the investment, this is an investment, ah if that investment is going um to pay for itself. 252 00:23:54,251 --> 00:23:56,331 And yet law firms routinely do it. 253 00:23:56,331 --> 00:24:06,672 ah And the truth of the matter is uh when people imagine people on the outside, 254 00:24:06,672 --> 00:24:09,594 sort of not having worked at law firms. 255 00:24:09,594 --> 00:24:16,778 They imagined law firms to be almost like Greek democracies, right? 256 00:24:16,939 --> 00:24:26,865 Where everybody uh gets a vote and you need to convince pretty much everyone uh in order to get a decision through. 257 00:24:26,865 --> 00:24:30,967 That is just not the case for big law firms. 258 00:24:31,928 --> 00:24:34,890 Big law firms are run by management. 259 00:24:34,890 --> 00:24:36,391 um 260 00:24:37,121 --> 00:24:41,134 The management does consult with the rank and file. 261 00:24:41,134 --> 00:24:44,806 um There are multiple reporting lines. 262 00:24:44,806 --> 00:24:52,822 There's multiple ways in which the management solicits input um from the partnership. 263 00:24:52,822 --> 00:24:55,603 And they're doing a marvelous job of that. 264 00:24:55,884 --> 00:25:01,007 But ultimately, it is the management that needs to be convinced. 265 00:25:01,007 --> 00:25:06,573 And then once the management posits something to the partnership, the partnership 266 00:25:06,573 --> 00:25:11,694 tends to listen, which is not to say that it's a rubber stamp, right? 267 00:25:11,694 --> 00:25:21,527 There are situations where the partnership is going to be opposed and will say, well, actually, no, we don't like this idea. 268 00:25:21,807 --> 00:25:28,479 But by and large, it's more common for people to trust their management. 269 00:25:28,479 --> 00:25:30,269 Otherwise, they wouldn't have elected them, right? 270 00:25:30,269 --> 00:25:36,323 um So the management can and does um 271 00:25:36,323 --> 00:25:39,905 propose things that are sort of inherently risky. 272 00:25:39,905 --> 00:25:46,149 And it's, it's no different with technology adoption is what I'm saying, right? 273 00:25:46,229 --> 00:26:02,679 The management uh has been talking to the partnerships at pretty much all the big law firms in the world, telling them, guys, we have to be um at least cognizant of the risk. 274 00:26:03,212 --> 00:26:15,563 We have to be least cognizant of the risk that the poses to the profession, and we have to explore how we make best use of this technology without running afoul of our professional 275 00:26:15,563 --> 00:26:20,166 duties and without ah getting in trouble with our clients. 276 00:26:20,407 --> 00:26:24,049 And our clients demand it, we have to follow. 277 00:26:24,050 --> 00:26:28,454 That's the nature of the law firm environment. 278 00:26:28,454 --> 00:26:30,535 ah 279 00:26:31,140 --> 00:26:37,504 Hopefully, this once again, this answers your question, but let me know if you'd like me to expand on it. 280 00:26:37,504 --> 00:26:38,675 No, that makes sense. 281 00:26:38,675 --> 00:26:41,316 Let's have a little debate here. 282 00:26:41,316 --> 00:26:50,661 um I would argue that the law firm business model has limited ability to scale. 283 00:26:50,661 --> 00:26:55,464 um And I'll tell you what I'm basing that on. 284 00:26:55,464 --> 00:26:57,635 We talked a little bit about it the last time. 285 00:26:57,635 --> 00:27:07,030 The largest law firm in the world, K &E, roughly $8 billion in revenue, is one fifth the size of the smallest of the big four. 286 00:27:07,328 --> 00:27:16,443 If you add up all of the revenues of the top five law firms in the world, they have like 6 % of market share. 287 00:27:16,443 --> 00:27:23,467 If you look at the big four, they have like over 90 % of the audit work of all publicly traded companies. 288 00:27:23,467 --> 00:27:25,988 It's extremely fragmented. 289 00:27:26,088 --> 00:27:31,731 So no firm has been able to scale, even K &E, which is an outlier at 8 billion. 290 00:27:31,831 --> 00:27:37,184 The average revenue at an AMLAL 100 firm is around 1 billion. 291 00:27:37,874 --> 00:27:44,201 Why do we think the current model accommodates scale? 292 00:27:47,365 --> 00:27:58,955 There are multiple reasons why uh the legal market is more fragmented than the big four market, right, than the audit market. 293 00:27:58,955 --> 00:27:59,976 uh 294 00:28:00,888 --> 00:28:04,189 I'd say some of it has to do with barriers to entry. 295 00:28:04,229 --> 00:28:19,416 Some of it has to do with the fact that uh we are talking about different jurisdictions, some of which are less uh susceptible to be folded into your practice. 296 00:28:19,416 --> 00:28:25,658 ah And some of it is just uh the fact that lawyers 297 00:28:28,111 --> 00:28:30,733 Culture is probably a bit more important. 298 00:28:30,733 --> 00:28:33,114 I know Big Four uh fairly well. 299 00:28:33,114 --> 00:28:34,215 I've worked with them. 300 00:28:34,215 --> 00:28:38,457 They used to be clients of mine, uh great people working there. 301 00:28:38,457 --> 00:28:41,379 uh My wife actually worked for a Big Four firm. 302 00:28:41,379 --> 00:28:54,667 uh At a certain point, the Big Four became fairly, firms within the Big Four became... 303 00:28:55,415 --> 00:28:58,596 I'd say pretty much mirror images of one another. 304 00:28:59,037 --> 00:29:04,599 Culturally, they're very similar, um or so it seems from the outside. 305 00:29:04,939 --> 00:29:06,340 Law firms are different. 306 00:29:06,340 --> 00:29:08,361 uh that's the truth of it. 307 00:29:08,361 --> 00:29:17,604 um Integrating two big law firms is a major challenge. 308 00:29:17,625 --> 00:29:23,707 Integrating 10 big law firms is a task that very few 309 00:29:23,757 --> 00:29:25,849 have been able to do successfully. 310 00:29:25,849 --> 00:29:39,080 Actually, Linklater has done that um in the 90s when it merged the Alliance firms, as they were known at the time, and uh sort of expanded it seriously. 311 00:29:39,080 --> 00:29:46,086 um There have been some good examples, and there have been some bad examples, and everyone knows it. 312 00:29:46,086 --> 00:29:48,267 I'm not even going to talk about that. 313 00:29:48,268 --> 00:29:52,815 So um law firms are not... 314 00:29:52,815 --> 00:29:56,115 looking to be like the big four. 315 00:29:56,115 --> 00:30:03,915 Law firms do not have ambitions to be sort of a Fortune 500 company. 316 00:30:04,795 --> 00:30:08,615 Some of it has to do with the restriction of outside investments in law firms. 317 00:30:08,615 --> 00:30:18,155 Some of it has to do with the ethical and professional restrictions that are imposed on lawyers. 318 00:30:18,395 --> 00:30:21,314 But ultimately, the... 319 00:30:21,314 --> 00:30:29,936 Firms are not going to invest billions and billions of dollars um in legal AI. 320 00:30:29,936 --> 00:30:30,597 They won't. 321 00:30:30,597 --> 00:30:39,179 uh Solutions need to be sensible in the context of the law firm model that exists right now. 322 00:30:40,059 --> 00:30:50,742 That leaves open the question of, is it possible to build a technology company that becomes a law firm? 323 00:30:50,742 --> 00:31:01,611 right, the AI native law firms that Y Combinator has called for and that uh have sprung up, uh know, Crosby comes to mind and the others. 324 00:31:01,611 --> 00:31:05,854 uh And that's an experiment. 325 00:31:05,914 --> 00:31:11,398 That's an experiment that uh we are still waiting to see the outcome. 326 00:31:14,312 --> 00:31:30,653 If it does work, um the value of those businesses will be tremendously higher than ah that of a traditional law firm. 327 00:31:30,653 --> 00:31:33,115 They're going to be valued differently. 328 00:31:34,096 --> 00:31:36,037 The multiples will be higher. 329 00:31:36,337 --> 00:31:43,582 And in a sense, for some of the more ambitious law firms, 330 00:31:44,687 --> 00:31:52,151 probably outside of the Amlo 100, uh mid-sized law firms, smaller law firms. 331 00:31:52,291 --> 00:31:58,114 This is clearly a path that they can take, and they can compete with the big guys. 332 00:31:58,594 --> 00:32:05,618 And ah that's not an existential threat at the moment. 333 00:32:05,618 --> 00:32:09,420 I wouldn't call it that, but it is a threat down the line. 334 00:32:09,420 --> 00:32:13,542 And it is something that uh 335 00:32:13,674 --> 00:32:32,610 I would certainly consider if I were running a successful mid-market firm or a small firm of several partners with technology uh using the technological footprint, they can 336 00:32:32,610 --> 00:32:36,994 certainly compete with the big guys and that can scale quickly. 337 00:32:37,134 --> 00:32:41,878 For big firms, uh the legacy oh systems are 338 00:32:42,230 --> 00:32:55,995 difficult to replace uh and that's not a path that I think any big firm is probably going to explore, throwing out the baby with the bath water and um building completely from 339 00:32:55,995 --> 00:32:56,966 scratch. 340 00:32:57,228 --> 00:33:06,658 Yeah, I have laid out a blueprint for, I call it a big law 2.0 blueprint for transformation. 341 00:33:06,658 --> 00:33:17,974 I would, if I were in that position leading a large law firm, I think, I'm not a lawyer, but I've worked with more than half the AMLaw in roughly 20 years. 342 00:33:18,155 --> 00:33:20,277 And so I know a lot about how they work. 343 00:33:20,277 --> 00:33:27,082 I've seen a really good cross-sectional view and some of the opportunities and challenges inside these organizations. 344 00:33:27,574 --> 00:33:44,629 If I were running one of these, I would be advocating to deploy a plan like that blueprint, which is basically take a non-trivial amount of bottom line revenue and deploy 345 00:33:44,629 --> 00:33:56,481 it into an entirely new structure, not a partnership, uh one that's a C-corp led by professional management with a board and 346 00:33:56,481 --> 00:33:58,072 outside investors. 347 00:33:58,072 --> 00:34:04,718 And obviously we have, you know, the ABA model rule five four that prevents legal, non-legal ownership at the moment. 348 00:34:04,718 --> 00:34:10,942 But there are a few States now where, you could, deploy this in a sandbox. 349 00:34:11,243 --> 00:34:15,716 And I would do this incrementally with a small portion. 350 00:34:15,716 --> 00:34:25,566 I think I outlined 20%, as a maybe starting point, because I agree with you that it's not, you're not going to be able to take. 351 00:34:25,566 --> 00:34:37,257 You know, there's legacy um barriers, legacy baggage in these existing organizations that will make it very difficult. 352 00:34:37,257 --> 00:34:44,964 So starting clean with a traditional governance model, you know, that's one of the reasons why I think that law firms aren't able to scale. 353 00:34:44,964 --> 00:34:46,225 It's their governance model. 354 00:34:46,225 --> 00:34:51,500 There's one, there's many, but one of them, you know, in a traditional C-corp, 355 00:34:51,500 --> 00:34:53,351 You have a group of investors. 356 00:34:53,811 --> 00:34:54,632 elect a board. 357 00:34:54,632 --> 00:35:09,580 That board installs professional management that are fully empowered and accountable and heavily invested with dollars in the success um of that stakeholder value. 358 00:35:09,580 --> 00:35:14,403 um That's not the case in law firms. 359 00:35:14,403 --> 00:35:18,445 Non-legal professionals are typically not always empowered. 360 00:35:18,445 --> 00:35:20,288 um 361 00:35:20,288 --> 00:35:33,070 it's difficult from a recruiting perspective to take a CTO from Amazon and hire him at a law firm where he's got stock options and he's fully empowered. 362 00:35:33,070 --> 00:35:41,287 um so I think, I think we kind of need a little bit of a uh reset in a number of different areas, including the governance model. 363 00:35:41,287 --> 00:35:41,647 I don't know. 364 00:35:41,647 --> 00:35:42,678 Do you agree? 365 00:35:43,790 --> 00:35:47,730 I think it's definitely a fascinating idea. 366 00:35:47,730 --> 00:35:50,270 I think that it has some value. 367 00:35:51,310 --> 00:36:03,070 I would still be wary if I were the CTO of Amazon of joining a startup like that. 368 00:36:04,210 --> 00:36:10,170 It's still sort of beholden to the lawyers as stakeholders. 369 00:36:10,190 --> 00:36:12,358 And no matter how 370 00:36:12,446 --> 00:36:25,758 much separation you want to build between the traditional law firm and this startup, um the imprint uh of the law firm mentality is still going to be on it. 371 00:36:25,758 --> 00:36:34,465 um Obviously, the law firm will have a vested interest in the services. 372 00:36:35,310 --> 00:36:46,870 being rendered by this startup being sort of top-notch and not reflecting poorly on the bigger brand. 373 00:36:47,310 --> 00:37:03,030 So if you were to say, there's sort of the bottom 20 % of the revenue in your example, you cannot afford having a situation where even a small portion of it is not delivered to the 374 00:37:03,030 --> 00:37:03,930 same story. 375 00:37:04,498 --> 00:37:11,103 same or at least similar standard that your clients are used to. 376 00:37:11,103 --> 00:37:29,015 uh And even identifying that 20 % of revenues is difficult because some of it may well be coming from the same clients that are included in the top 10 % of your revenue. 377 00:37:29,015 --> 00:37:34,238 So you really can't, if you've got a platinum 378 00:37:34,270 --> 00:37:41,915 or uh a client that is uh rated very highly within the organization. 379 00:37:43,497 --> 00:37:47,209 We always regard pretty much all the clients as important. 380 00:37:47,209 --> 00:38:02,810 But um let's say the client that generates the most revenue for the firm, if they generate a lot of revenue and you take sort of a small portion of their work, which is 381 00:38:03,064 --> 00:38:12,678 simple low-level work and you want to be doing it out of the startup, that's where you really get that reputational concern. 382 00:38:12,678 --> 00:38:29,444 um Not to say that oh none of this could be made, and in fact, some law firms are going maybe not in exactly this direction, but they're uh doing something similar. 383 00:38:29,444 --> 00:38:31,845 Cleary comes to mind with ClearyX. 384 00:38:32,309 --> 00:38:32,689 Right. 385 00:38:32,689 --> 00:38:39,771 uh And it's, it's this idea of a playground. 386 00:38:39,771 --> 00:38:44,533 don't, I don't know if it necessarily has to be a separate uh entity. 387 00:38:44,533 --> 00:38:45,433 It could be. 388 00:38:45,433 --> 00:38:47,864 It doesn't have to be. 389 00:38:47,864 --> 00:38:59,597 The idea of a playground where those experiments happen and adoption happens and happens faster than trying to roll it out to the entire firm. 390 00:39:00,101 --> 00:39:02,884 There that I definitely agree with you. 391 00:39:02,925 --> 00:39:19,045 I absolutely agree that that's the way for big law to sort of transform itself and use that pocket of competences um as the um 392 00:39:20,809 --> 00:39:23,571 as the guiding light, right? 393 00:39:23,571 --> 00:39:35,711 If that pocket of competences, technological competences is successful and it does generate revenue and it does keep the clients happy and clients swear by it and say, 394 00:39:35,711 --> 00:39:49,172 actually, you know what, what used to cost us a hundred K now costs us five and we are still getting the same quality and it's much faster and it's 395 00:39:49,770 --> 00:39:52,512 it's technologically advanced. 396 00:39:52,512 --> 00:39:55,794 We understand what you're doing there. 397 00:39:55,854 --> 00:40:12,426 We want to work with you because not only are we getting the um expert advice from your partners on the matters where we need expert advice, we are also getting the grant work, 398 00:40:12,426 --> 00:40:16,168 if you will, handled faster and more efficiently. 399 00:40:16,168 --> 00:40:17,449 That's when 400 00:40:17,505 --> 00:40:21,307 people will absolutely buy into it. 401 00:40:21,368 --> 00:40:25,831 We are talking about the advisory part of the work. 402 00:40:25,831 --> 00:40:40,460 uh The thing that I feel is often overlooked is how um time consuming the administrative part of uh law firm life is. 403 00:40:41,001 --> 00:40:47,465 And that is another area where I think transformation could really be happening. 404 00:40:47,469 --> 00:41:06,954 uh law firms could adopt technology much faster when they're not weighing the gains against um less billable hours uh for the sake of it, but rather um when they are 405 00:41:07,614 --> 00:41:15,607 essentially saying we are saving, you know, all the time saved or most of the time saved is non-billable time, right? 406 00:41:15,607 --> 00:41:17,357 We actually have more 407 00:41:17,357 --> 00:41:19,900 time to dedicate to our clients. 408 00:41:20,242 --> 00:41:23,206 And that's a winning proposition. 409 00:41:23,427 --> 00:41:29,774 The business of law uh through digitization is definitely a winning proposition. 410 00:41:29,774 --> 00:41:30,854 Yeah. 411 00:41:30,894 --> 00:41:45,134 So speaking of that transition from kind of hourly billing to AFA work, you know, I have heard people throw out, oh, the answer is outcome-based billing or it's value-based 412 00:41:45,134 --> 00:41:46,054 billing. 413 00:41:46,474 --> 00:41:48,154 That's not the end of the conversation. 414 00:41:48,154 --> 00:41:50,494 That's the beginning of the conversation. 415 00:41:50,974 --> 00:41:51,654 Yes. 416 00:41:51,654 --> 00:41:55,774 Because there's so much devil in the details around that. 417 00:41:55,774 --> 00:41:57,914 So you can't just say, 418 00:41:58,199 --> 00:42:09,625 I will, we will take some reasonable percentage of the value that we deliver because what that neglects is supply and demand. 419 00:42:09,926 --> 00:42:15,909 Where do you draw that line of, let's say you've delivered a million dollars worth of value. 420 00:42:15,909 --> 00:42:17,370 Well, we're going to ask for 20%. 421 00:42:17,370 --> 00:42:24,774 Okay, oh well, what's the law firm down the street going to ask for who does comparable? 422 00:42:24,942 --> 00:42:25,952 comparable quality. 423 00:42:25,952 --> 00:42:29,423 There's just so many variables. 424 00:42:29,583 --> 00:42:31,874 I don't see any real easy answers. 425 00:42:31,874 --> 00:42:42,407 And I read an article um about tech fees associated with um AI infrastructure. 426 00:42:42,407 --> 00:42:45,248 That's a whole other set of problems. 427 00:42:45,248 --> 00:42:50,370 um You they talked about day-to-day tools like maybe Harvey and Legora. 428 00:42:50,370 --> 00:42:51,910 um 429 00:42:51,916 --> 00:42:54,368 You know, you treat those like you do Microsoft Word. 430 00:42:54,368 --> 00:43:09,421 You don't bill your clients for Microsoft Word, but if you're doing some specialized task that is very resource or tech intensive, maybe that could be something that you bill for 431 00:43:09,421 --> 00:43:11,083 separately in addition to your hours. 432 00:43:11,083 --> 00:43:18,249 So the pricing thing is really fascinating to me and I really don't feel that there are lots any easy answers out there. 433 00:43:18,249 --> 00:43:18,569 I don't know. 434 00:43:18,569 --> 00:43:20,190 What is your take on that? 435 00:43:21,503 --> 00:43:30,680 There will be multiple models and I do not think that there is one right or wrong answer really. 436 00:43:30,680 --> 00:43:32,100 I completely agree with you, Ted. 437 00:43:32,100 --> 00:43:50,533 I think that uh lawyers love analogies, so you indulge me and allow me to say that law firms have offered, have always offered other services in addition to um 438 00:43:50,655 --> 00:43:52,185 There are primary services. 439 00:43:52,185 --> 00:43:57,167 um In some cases, that's actually part of the revenue. 440 00:43:57,167 --> 00:44:12,090 In some cases, that's a value add that is provided to the clients uh for free um to build goodwill and strengthen the relationship, all of that. 441 00:44:12,090 --> 00:44:19,972 uh For as long as those services were uh ancillary, 442 00:44:19,986 --> 00:44:35,902 to the main work and did not directly impact on the service delivery, it certainly made sense that most of those were provided for free. 443 00:44:36,262 --> 00:44:41,084 Now I think we're entering into a different world where 444 00:44:41,768 --> 00:44:54,834 It's almost like clients might have a choice, and in fact they do have a choice in certain situations where you might say to the clients, we can do this uh due diligence document 445 00:44:54,834 --> 00:44:57,775 review the traditional way. 446 00:44:57,775 --> 00:45:01,637 We'll put an army of uh junior lawyers on this. 447 00:45:01,637 --> 00:45:06,199 They will sort of bill X number of hours. 448 00:45:06,199 --> 00:45:10,020 We will negotiate with you and that's going to be 449 00:45:11,747 --> 00:45:15,208 $500,000 for the sake of the argument, right? 450 00:45:17,028 --> 00:45:27,628 Or you can have it done by a tool that we've bought, that we are using, we are subscribed for it. 451 00:45:28,228 --> 00:45:31,568 We think it's a good tool, we like it. 452 00:45:31,968 --> 00:45:34,068 The accuracy is there. 453 00:45:34,068 --> 00:45:40,630 And then you can either ask us to double check, do a selective 454 00:45:40,630 --> 00:45:55,397 verification process, or you might say, you know what, none of this matters to me, I'm just happy for you uh to run the tool with no additional uh lawyer time added on top of 455 00:45:55,397 --> 00:45:56,287 that fee. 456 00:45:56,287 --> 00:45:58,308 um 457 00:45:59,923 --> 00:46:10,816 I will posit, I will sort argue that in these three situations, you have, know, and you still have the tool in the background. 458 00:46:10,816 --> 00:46:24,449 So if you're running that analysis, that due diligence, the traditional way, you might still enable your team to use the tool to do sort of the first run through the due 459 00:46:24,449 --> 00:46:26,559 diligence materials, right? 460 00:46:26,559 --> 00:46:29,660 ah Do you... 461 00:46:31,365 --> 00:46:41,573 Do you charge the client less for the traditional services because your people spent 80 hours as opposed to 100 hours on it? 462 00:46:41,573 --> 00:46:50,721 ah Do you charge the client more for the time it takes to review in the second scenario? 463 00:46:50,721 --> 00:47:00,012 Because you're like, actually, that review needs to be done not by junior lawyers, but by people who are more senior who can actually spot 464 00:47:00,012 --> 00:47:03,754 ah the things that the solution might have missed. 465 00:47:03,754 --> 00:47:20,091 ah And do you charge sort of, do you pass on the service, the tool at cost in the third scenario, or do you actually say, well, we are losing the revenue from the traditional 466 00:47:20,091 --> 00:47:20,501 way. 467 00:47:20,501 --> 00:47:27,944 um So that tool will cost you something, or you can go ahead and. 468 00:47:27,944 --> 00:47:31,584 get a subscription yourself and run it on the data room. 469 00:47:31,927 --> 00:47:33,588 Those are the... 470 00:47:33,729 --> 00:47:37,272 And once again, we're just scratching the surface here, right? 471 00:47:37,272 --> 00:47:50,062 You literally have all kinds of possibilities and all kinds of business models, and some firms will be heavily reliant on the technological tools and others will not. 472 00:47:50,062 --> 00:47:56,287 They will prefer to uh stick to the traditional way for as long as they can. 473 00:47:56,287 --> 00:47:57,388 And, you know, that... 474 00:47:57,388 --> 00:48:03,388 that has to do with the risk aversion and the... 475 00:48:04,940 --> 00:48:10,665 traditionalism and conservatism, healthy conservatism of the legal profession. 476 00:48:10,668 --> 00:48:12,539 Yeah, that makes sense. 477 00:48:12,539 --> 00:48:15,060 Well, I hate that we're out of time. 478 00:48:15,060 --> 00:48:18,562 I have thoroughly enjoyed this conversation. 479 00:48:19,983 --> 00:48:34,381 it's not, we have lots of lawyers on the show, but I think your role as having been a longtime partner at a major firm and now investing in the space gives some interesting 480 00:48:34,381 --> 00:48:35,972 perspective there. 481 00:48:35,972 --> 00:48:40,314 Because I see varying levels of buy-in. 482 00:48:40,450 --> 00:48:46,913 on just how transformational this tech is going to be in the industry. 483 00:48:46,913 --> 00:48:49,054 And I feel like you get it. 484 00:48:49,054 --> 00:48:54,155 um have lots of friends in the space having been around for a while. 485 00:48:54,356 --> 00:49:07,461 And ones who are really smart and who I respect to think I'm over indexing on the transformational magnitude, I don't think that I am. 486 00:49:07,461 --> 00:49:09,964 um 487 00:49:09,964 --> 00:49:17,194 You know, law firms have been successful being the way that they've been for a really long time. 488 00:49:17,194 --> 00:49:19,668 I just think, yeah, exactly. 489 00:49:19,668 --> 00:49:21,119 I think that gig is up though. 490 00:49:21,119 --> 00:49:21,879 I really do. 491 00:49:21,879 --> 00:49:23,620 Now, how soon? 492 00:49:23,874 --> 00:49:28,863 I had a guest on yesterday um who runs a VC. 493 00:49:28,863 --> 00:49:31,994 He's a partner at TLTF. 494 00:49:31,994 --> 00:49:36,226 And he was hesitant to lay out a timeline. 495 00:49:36,745 --> 00:49:41,217 I have less to lose and if I'm wrong, who cares? 496 00:49:41,217 --> 00:49:46,479 But I think that there will be no impact on revenues this year. 497 00:49:46,479 --> 00:49:54,842 I think very modest in 2026, but I think in 2027, there will be non-trivial, there will be real impact. 498 00:49:54,842 --> 00:50:02,765 um And that is based on how fast I see this technology moving. 499 00:50:02,765 --> 00:50:05,506 And I speak to a lot of 500 00:50:05,666 --> 00:50:15,789 GCs and I they there is so much pent up demand for efficiency From their law firm partners. 501 00:50:15,789 --> 00:50:28,633 I mean, it's they're screaming for it, you know as rates continue to go up so I don't know last thought does that seem like a reasonable timeline in terms of When we'll see impact 502 00:50:28,633 --> 00:50:30,734 or you think it'll be sooner longer 503 00:50:34,031 --> 00:50:37,392 My prophetic capabilities are limited. 504 00:50:37,731 --> 00:50:43,211 I don't necessarily know what the timeline is. 505 00:50:43,211 --> 00:50:47,057 It sort of does depend on a lot of factors. 506 00:50:47,537 --> 00:50:53,960 I will agree with you wholeheartedly that the demand from the clients is there. 507 00:50:53,960 --> 00:50:59,947 um I will also say that there will be an impact. 508 00:50:59,947 --> 00:51:12,038 ah on revenues, law firm revenues, uh I might be a bit more contrarian in terms of the gig is up. 509 00:51:12,038 --> 00:51:25,991 I actually think that there may well be some scenarios where law firms will improve their revenues, or at least some law firms will improve their revenues by quite a bit. 510 00:51:26,435 --> 00:51:28,356 utilizing technology. 511 00:51:28,437 --> 00:51:35,462 it's more a question of if you are doing nothing with it, then you could be in trouble. 512 00:51:35,462 --> 00:51:51,254 ah If you are actually actively monitoring what's happening in the space, and you're looking for ways to uh bring technology, sort of enlist the services of technology and um 513 00:51:51,254 --> 00:51:53,175 get it right, 514 00:51:53,189 --> 00:51:57,041 then you might be one of the beneficiaries. 515 00:51:57,041 --> 00:52:00,642 And there will be more than one sort group of beneficiaries. 516 00:52:00,822 --> 00:52:06,014 So look, I wish I had the proverbial crystal ball. 517 00:52:06,014 --> 00:52:07,505 I don't. 518 00:52:07,505 --> 00:52:14,648 But um the impact will be felt within the next couple of years, definitely. 519 00:52:14,648 --> 00:52:21,340 Whether it's going to be sort of a negative impact or a positive impact, I really do think it will matter. 520 00:52:21,340 --> 00:52:22,811 Either it will depend on 521 00:52:23,251 --> 00:52:27,354 how people approach uh this whole situation. 522 00:52:28,435 --> 00:52:31,052 The ostrich approach is not the right approach. 523 00:52:31,052 --> 00:52:33,163 Well, and that's what I mean by the gig is up. 524 00:52:33,163 --> 00:52:44,528 The law firms not across the board, but Many many law firms have been extremely resistant to tech, even the cloud. 525 00:52:44,528 --> 00:52:45,828 And that gig is up. 526 00:52:45,828 --> 00:52:46,458 Yeah. 527 00:52:46,458 --> 00:52:49,460 There's that, that is no longer going to fly that I'm certain of. 528 00:52:49,460 --> 00:52:52,070 I'll bet you my, I'll bet you the company. 529 00:52:52,391 --> 00:52:53,091 Yeah. 530 00:52:53,091 --> 00:52:54,012 Yeah. 531 00:52:54,012 --> 00:52:58,873 So yeah, it's been, we're going to make changes here. 532 00:52:58,873 --> 00:53:00,414 It's going to happen. 533 00:53:00,416 --> 00:53:04,847 and law firms are going to have to make investments, it's not optional. 534 00:53:04,847 --> 00:53:13,190 um When the chickens will come home to roost, that's up for debate, but they're coming. 535 00:53:13,190 --> 00:53:17,671 um So it's a fun time to be around. 536 00:53:17,671 --> 00:53:19,352 So this is a great conversation. 537 00:53:19,352 --> 00:53:21,372 really, I really enjoyed it. 538 00:53:21,372 --> 00:53:27,444 For folks that want to find out more about, you know, I think you have some great commentary on LinkedIn. 539 00:53:27,444 --> 00:53:28,524 I haven't seen 540 00:53:28,672 --> 00:53:32,513 actual posts, but I've really enjoyed some of the comments that you made. 541 00:53:32,877 --> 00:53:36,085 Is that the best way for people to connect with you? 542 00:53:36,764 --> 00:53:38,105 Or they can just write to me. 543 00:53:38,105 --> 00:53:40,143 My email is on LinkedIn as well. 544 00:53:40,143 --> 00:53:44,798 I'm always happy to speak and thank you so much for having me on the podcast. 545 00:53:44,798 --> 00:53:46,671 I really enjoyed talking to you. 546 00:53:46,671 --> 00:53:47,751 Awesome. 547 00:53:47,851 --> 00:53:48,231 All right. 548 00:53:48,231 --> 00:53:50,731 Well, I'm sure we'll bump into each other. 549 00:53:51,731 --> 00:53:57,411 I do a lot of conferences out there, so I'm sure we'll bump into each other in person. 550 00:53:57,531 --> 00:54:01,471 But until then, thank you very much, and we'll chat again soon. 551 00:54:02,890 --> 00:54:03,311 All righty. 552 00:54:03,311 --> 00:54:04,251 Take care. -->