1 00:00:02,720 --> 00:00:16,400 Speaker 1: Bloomberg Audio Studios, Podcasts, radio News. 2 00:00:17,960 --> 00:00:21,759 Speaker 2: Hello and welcome to another episode of the Odd Lots podcast. 3 00:00:21,840 --> 00:00:24,200 Speaker 3: I'm Joe Wisenthal and I'm Tracy Alloway. 4 00:00:24,520 --> 00:00:26,160 Speaker 2: So, Tracy, I think like maybe a year and a 5 00:00:26,160 --> 00:00:29,240 Speaker 2: half ago, we did that episode with Joel Wertheimer talking 6 00:00:29,240 --> 00:00:34,200 Speaker 2: about AI and law. It does feel like law specifically 7 00:00:35,600 --> 00:00:38,920 Speaker 2: is one of those areas where it's very easy for 8 00:00:39,000 --> 00:00:42,760 Speaker 2: the normal person to imagine how AI could be very disruptive. 9 00:00:43,080 --> 00:00:47,000 Speaker 3: Would you say that's fair, yes, I would say, however, 10 00:00:47,760 --> 00:00:50,560 Speaker 3: and full disclaimer, So my husband used to be a lawyer, 11 00:00:50,560 --> 00:00:54,000 Speaker 3: a corporate lawyer. I would say that, like, standardized forms 12 00:00:54,000 --> 00:00:57,240 Speaker 3: and templates have existed in the legal profession for many, 13 00:00:57,280 --> 00:01:00,120 Speaker 3: many years, and law has also gone through many many 14 00:01:00,160 --> 00:01:04,520 Speaker 3: technological revolution so we've gone from like scriveners, you know, 15 00:01:06,200 --> 00:01:10,800 Speaker 3: to like keyboardists. The advent of like typis was supposed 16 00:01:10,800 --> 00:01:13,440 Speaker 3: to be this massive hit for not a hit, but 17 00:01:13,560 --> 00:01:17,720 Speaker 3: like this massive deal for the legal industry, and everything 18 00:01:17,800 --> 00:01:19,760 Speaker 3: ended up like navigating through it pretty much. 19 00:01:19,840 --> 00:01:22,600 Speaker 2: I guess they probably each one of those waves. There 20 00:01:22,680 --> 00:01:28,120 Speaker 2: probably was some extreme period of anxiety about disruption and 21 00:01:28,319 --> 00:01:31,480 Speaker 2: how would billing still happen? And of course one of 22 00:01:31,480 --> 00:01:33,880 Speaker 2: the questions that comes to mind and the legal profession 23 00:01:34,000 --> 00:01:36,160 Speaker 2: is like, why would the lawyers want to reduce the 24 00:01:36,240 --> 00:01:38,920 Speaker 2: number of hours that they can, you know, do on 25 00:01:39,000 --> 00:01:41,880 Speaker 2: a case, given that famously they're paid by the hour. 26 00:01:42,000 --> 00:01:44,200 Speaker 2: I think, actually, are they like paid by the ten minutes? 27 00:01:44,560 --> 00:01:48,280 Speaker 2: Technically sometimes yeah, but it does feel like we're gods. 28 00:01:48,440 --> 00:01:51,040 Speaker 2: So maybe, like I'm not saying, and I don't even 29 00:01:51,080 --> 00:01:54,040 Speaker 2: have an intuition per se that a I would like upend, 30 00:01:54,280 --> 00:01:57,000 Speaker 2: like a reduce the amount of legal work that needs 31 00:01:57,000 --> 00:02:00,880 Speaker 2: to get done. But those technological revolutions that you mentioned 32 00:02:00,880 --> 00:02:03,200 Speaker 2: in the past probably did change the industry quite a bit. 33 00:02:03,480 --> 00:02:07,080 Speaker 2: So regardless of like total employment or job prospects, et cetera, 34 00:02:07,480 --> 00:02:09,760 Speaker 2: it feels inevitable that change is going to come at 35 00:02:09,760 --> 00:02:11,920 Speaker 2: one point. It is probably like a skill to be 36 00:02:11,919 --> 00:02:14,240 Speaker 2: able to know case law or know how to search 37 00:02:14,320 --> 00:02:16,920 Speaker 2: lexus and nexus like that almost certainly is going to change. 38 00:02:16,960 --> 00:02:19,160 Speaker 3: So one thing I've been thinking about, and this gets 39 00:02:19,160 --> 00:02:21,880 Speaker 3: into the productivity debate, is I'm starting to come around 40 00:02:21,919 --> 00:02:24,720 Speaker 3: to the idea of a Jeffins paradox in just general 41 00:02:24,840 --> 00:02:25,919 Speaker 3: for work bureaucracy. 42 00:02:26,080 --> 00:02:27,200 Speaker 2: Yeah, basically, so. 43 00:02:27,240 --> 00:02:30,800 Speaker 3: You already have AI that's like being used by insurance 44 00:02:30,880 --> 00:02:34,480 Speaker 3: companies to push back against claims, and then you have 45 00:02:34,919 --> 00:02:40,239 Speaker 3: claimants using AI to then push back against the insurance companies. 46 00:02:40,520 --> 00:02:43,000 Speaker 3: And I feel like the future could just be a 47 00:02:43,000 --> 00:02:45,440 Speaker 3: bunch of bots like talking to each other and filing 48 00:02:45,560 --> 00:02:47,000 Speaker 3: different claims, and. 49 00:02:46,760 --> 00:02:48,880 Speaker 2: I guess he will just sort it all out at 50 00:02:48,919 --> 00:02:49,240 Speaker 2: the end. 51 00:02:49,440 --> 00:02:51,520 Speaker 3: Yeah, And maybe I don't get more productive. 52 00:02:51,560 --> 00:02:51,960 Speaker 4: I don't know. 53 00:02:52,040 --> 00:02:54,680 Speaker 2: Other people have said this. Everyone. A lot of people 54 00:02:54,720 --> 00:02:58,919 Speaker 2: like intec coding all of these computer jobs. Right now, 55 00:02:59,560 --> 00:03:02,400 Speaker 2: a lot of people seem to feel very stressed overworked. 56 00:03:02,440 --> 00:03:05,160 Speaker 2: Like I don't think you like, look at a situation 57 00:03:05,360 --> 00:03:07,680 Speaker 2: someone who has a job that primarily involves in their computer. 58 00:03:07,760 --> 00:03:10,840 Speaker 2: It's like, oh, I'm feeling like I'm really avid eazy. 59 00:03:10,919 --> 00:03:14,880 Speaker 2: These days, everyone feels very stressed. But anyway, I'm very 60 00:03:14,880 --> 00:03:17,520 Speaker 2: interested in this topic. Was a big story several months ago. 61 00:03:17,600 --> 00:03:20,320 Speaker 2: One of the major law firms, Kirkland and ellis like 62 00:03:20,360 --> 00:03:24,040 Speaker 2: talking about like building out more in house infrastructure. So 63 00:03:24,120 --> 00:03:25,920 Speaker 2: I think it's like a good time to sort of 64 00:03:26,240 --> 00:03:29,360 Speaker 2: get a state of the market of like, okay, like 65 00:03:29,639 --> 00:03:33,400 Speaker 2: we're what almost four years into the post chat GPT world, 66 00:03:34,200 --> 00:03:37,480 Speaker 2: where do we actually stand in terms of what what 67 00:03:37,800 --> 00:03:40,320 Speaker 2: AI could do for legal work, and then also just 68 00:03:40,360 --> 00:03:42,440 Speaker 2: like how it's changing the profession if at all. Yeah, 69 00:03:42,520 --> 00:03:44,160 Speaker 2: let's do it all right. I'm very excited to say 70 00:03:44,320 --> 00:03:46,840 Speaker 2: we do, in fact have the perfect guest, someone who's 71 00:03:46,880 --> 00:03:49,560 Speaker 2: been talking a lot about this, someone who's written about 72 00:03:49,600 --> 00:03:54,360 Speaker 2: this topic, someone who is at a very substantial law 73 00:03:54,360 --> 00:03:56,440 Speaker 2: firm where we can actually see some of this taking 74 00:03:56,480 --> 00:03:59,040 Speaker 2: place in real times. We're gonna be speaking with Gary Wingins. 75 00:03:59,040 --> 00:04:01,880 Speaker 2: He is the chair of Lowenstein Sandler. So, Gary, thank 76 00:04:01,880 --> 00:04:03,120 Speaker 2: you so much for coming on odd. 77 00:04:02,960 --> 00:04:05,520 Speaker 4: Lots delighted to be here. Thanks for having me. 78 00:04:05,920 --> 00:04:08,200 Speaker 2: Just to set the scene. Why don't you give us 79 00:04:08,240 --> 00:04:11,040 Speaker 2: a little bit of an idea of what the firm is, 80 00:04:11,160 --> 00:04:13,240 Speaker 2: how big it is, your role there, and what it's 81 00:04:13,280 --> 00:04:15,360 Speaker 2: sort of special, what kind of work gets done there? 82 00:04:15,440 --> 00:04:18,560 Speaker 4: Again, thanks for having me. Loan Science Sandler about four 83 00:04:18,640 --> 00:04:23,560 Speaker 4: hundred lawyers, primarily located in the New York area. 84 00:04:23,680 --> 00:04:25,680 Speaker 2: But it's four hundred big law at that point. 85 00:04:25,720 --> 00:04:29,359 Speaker 4: Four hundred is big law. It's not huge law. You 86 00:04:29,400 --> 00:04:33,800 Speaker 4: mentioned Kirkland earlier. They're much larger. We very firmly focus 87 00:04:33,880 --> 00:04:38,520 Speaker 4: on clients in the private capital sector, private equity, venture 88 00:04:39,120 --> 00:04:42,960 Speaker 4: hedge funds, technology companies and life sciences businesses. 89 00:04:43,320 --> 00:04:46,840 Speaker 3: Why has the billable hours thing lasted as long as 90 00:04:46,880 --> 00:04:49,359 Speaker 3: it has because everyone seems to complain about this, Like 91 00:04:49,360 --> 00:04:52,440 Speaker 3: the lawyers complain that it's terrible, the clients complain that 92 00:04:52,480 --> 00:04:56,240 Speaker 3: it's terrible, and it does seem to lead to sometimes 93 00:04:56,279 --> 00:04:58,240 Speaker 3: not the most efficient outcomes. 94 00:04:58,560 --> 00:05:02,279 Speaker 4: Everyone agrees it's terrible, as you said, yet it remains 95 00:05:02,279 --> 00:05:05,560 Speaker 4: the dominant model. I have never met a client who 96 00:05:05,560 --> 00:05:07,960 Speaker 4: has come to me asking to buy billable hours. 97 00:05:08,240 --> 00:05:08,400 Speaker 3: Right. 98 00:05:08,520 --> 00:05:10,599 Speaker 4: Nobody ever wants to buy billable hours. They want to 99 00:05:10,600 --> 00:05:13,919 Speaker 4: buy business solutions, and it happens that billable hours is 100 00:05:13,960 --> 00:05:17,279 Speaker 4: how we measure what the bill will turn out to be. 101 00:05:17,920 --> 00:05:20,919 Speaker 4: There has been, at least for the past fifteen years, 102 00:05:21,360 --> 00:05:25,240 Speaker 4: there's been a pretty strong movement toward what a're called 103 00:05:25,279 --> 00:05:30,160 Speaker 4: or alternative fear arrangements, where you know you, project based pricing, 104 00:05:30,440 --> 00:05:35,480 Speaker 4: cap fees, collars, things like that has not really taken hold, 105 00:05:36,120 --> 00:05:39,480 Speaker 4: which is surprising. We are I think law is one 106 00:05:39,480 --> 00:05:43,240 Speaker 4: of the last segments in professional services firms that has 107 00:05:43,279 --> 00:05:49,239 Speaker 4: not moved to some kind of value or project based pricing. Perhaps, 108 00:05:49,279 --> 00:05:52,960 Speaker 4: and I'm not exactly sure why, Perhaps it's because lawyers 109 00:05:52,960 --> 00:05:56,120 Speaker 4: and clients don't really trust each other that much. And 110 00:05:57,080 --> 00:06:00,400 Speaker 4: when we propose a you know, here's how much a 111 00:06:00,440 --> 00:06:05,720 Speaker 4: cost to do this kind of public offering, sometimes clients say, well, 112 00:06:05,800 --> 00:06:08,440 Speaker 4: if you're going to propose that amount, I'd rather pay 113 00:06:08,480 --> 00:06:10,760 Speaker 4: you by the hour because I want to see what 114 00:06:10,760 --> 00:06:14,120 Speaker 4: you're actually doing. On the other hand, when clients ask 115 00:06:14,200 --> 00:06:16,839 Speaker 4: for alternative fear arrangements, we almost always offer them. 116 00:06:17,680 --> 00:06:22,880 Speaker 2: Is there a understanding though of like, okay, maybe it's 117 00:06:22,960 --> 00:06:26,800 Speaker 2: billable hours and the alternative fear arrangement doesn't get formally 118 00:06:26,839 --> 00:06:30,160 Speaker 2: written down, But is there a general taskit understanding of 119 00:06:30,160 --> 00:06:33,480 Speaker 2: like okay, here is a IPO we're going to raise 120 00:06:33,520 --> 00:06:36,520 Speaker 2: a billion dollars, or here is a you know, venture 121 00:06:36,560 --> 00:06:38,599 Speaker 2: capital deal where a company is going to ran five 122 00:06:39,160 --> 00:06:42,920 Speaker 2: raise five hundred million dollars. How much that is there 123 00:06:42,960 --> 00:06:45,560 Speaker 2: a cost? Like is it well understood what this is? 124 00:06:45,600 --> 00:06:47,720 Speaker 2: How much this kind of job cost? And if it's 125 00:06:47,920 --> 00:06:51,520 Speaker 2: wildly different, then some alarm bells would go of yeah. 126 00:06:51,560 --> 00:06:56,120 Speaker 4: So the IPO market is easy because everybody as part 127 00:06:56,120 --> 00:06:59,359 Speaker 4: of your disclosure in doing an IPO, you disclose how 128 00:06:59,400 --> 00:07:02,200 Speaker 4: much you pay the life and so there is a 129 00:07:02,600 --> 00:07:06,280 Speaker 4: known market and there's a known comparative price, you know, 130 00:07:06,360 --> 00:07:09,640 Speaker 4: a venture. A round is a little more challenging because 131 00:07:09,680 --> 00:07:12,360 Speaker 4: there are lots of different details, but there's there's a 132 00:07:12,440 --> 00:07:16,360 Speaker 4: common understanding about the range of rates and then things 133 00:07:16,440 --> 00:07:18,720 Speaker 4: go from there if there are funky tax issues or 134 00:07:18,760 --> 00:07:21,040 Speaker 4: things like that. But for most areas, there's kind of 135 00:07:21,040 --> 00:07:22,240 Speaker 4: a common range. 136 00:07:22,760 --> 00:07:26,800 Speaker 3: So getting back to the topic at hand, AI, How 137 00:07:26,880 --> 00:07:29,360 Speaker 3: much of a step change is this in terms of 138 00:07:29,360 --> 00:07:32,880 Speaker 3: the actual work process for a corporate lawyer? Because as 139 00:07:32,880 --> 00:07:37,520 Speaker 3: I mentioned before, like standardized templates exist. Law firms have had, 140 00:07:37,640 --> 00:07:40,520 Speaker 3: you know, software for due diligence for ages. A lot 141 00:07:40,600 --> 00:07:44,640 Speaker 3: of them have knowledge libraries where they sort of share information, 142 00:07:44,800 --> 00:07:49,320 Speaker 3: or you have experts that act like little encyclopedias for law. 143 00:07:49,720 --> 00:07:50,880 Speaker 3: How big a change is this? 144 00:07:51,920 --> 00:07:56,320 Speaker 4: I think it is a pretty huge change. So there 145 00:07:56,320 --> 00:08:01,280 Speaker 4: are two ways it's changing. One is what you're talking 146 00:08:01,320 --> 00:08:05,400 Speaker 4: about kind of the the process and systems and using 147 00:08:05,960 --> 00:08:10,240 Speaker 4: AI to become more efficient, just like when you know, 148 00:08:10,280 --> 00:08:14,160 Speaker 4: when Microsoft Word came out, came out and it was 149 00:08:14,280 --> 00:08:16,680 Speaker 4: became much more efficient for you to kind of do 150 00:08:16,800 --> 00:08:19,160 Speaker 4: your own documents or compare Right. When I was a 151 00:08:19,200 --> 00:08:21,800 Speaker 4: first year associate, I'm old. When I was a first 152 00:08:21,840 --> 00:08:25,680 Speaker 4: year associate, I was hand black lining documents and clients 153 00:08:25,680 --> 00:08:28,440 Speaker 4: were paying me by the hour to hand black line documents. 154 00:08:28,680 --> 00:08:32,600 Speaker 4: Within two years, you know, we had black lining software 155 00:08:32,720 --> 00:08:34,400 Speaker 4: and they no longer paid me by the hour, and 156 00:08:34,440 --> 00:08:37,240 Speaker 4: it took roughly ninety seconds to black line a document. 157 00:08:37,559 --> 00:08:42,160 Speaker 4: So there's there's the efficiency piece which definitely is starting 158 00:08:42,240 --> 00:08:45,240 Speaker 4: to make it to make the things that clients want 159 00:08:45,320 --> 00:08:46,920 Speaker 4: to buy less expensive. 160 00:08:47,120 --> 00:08:48,280 Speaker 2: Were you ever did you. 161 00:08:48,280 --> 00:08:51,720 Speaker 3: Ever miss a comma that resulted in like five million 162 00:08:51,760 --> 00:08:52,720 Speaker 3: dollars in extra. 163 00:08:52,559 --> 00:08:54,719 Speaker 2: Chargees, Not that I'm aware of. 164 00:08:54,760 --> 00:08:58,199 Speaker 3: Okay, as that's a famous case, right right, right. 165 00:08:58,000 --> 00:09:01,640 Speaker 4: But I've definitely missed commas every body has, but I'm 166 00:09:01,679 --> 00:09:04,680 Speaker 4: not aware that it's you know, result in anything bad happening. 167 00:09:05,960 --> 00:09:09,040 Speaker 4: Hopefully nobody listening to this has the other side of that. 168 00:09:10,600 --> 00:09:14,520 Speaker 4: The so, but the other way AI is working in 169 00:09:14,600 --> 00:09:17,360 Speaker 4: law firms right now is it's it's acting as a 170 00:09:17,400 --> 00:09:21,440 Speaker 4: thought partner, and it is making us better at our 171 00:09:21,559 --> 00:09:26,640 Speaker 4: jobs and to use a term, acting as a co 172 00:09:26,800 --> 00:09:31,920 Speaker 4: pilot for all of our lawyers in helping think through problems. 173 00:09:32,240 --> 00:09:35,440 Speaker 4: So as that thought partner, we've never had that before. 174 00:09:35,760 --> 00:09:35,880 Speaker 3: Right. 175 00:09:35,920 --> 00:09:42,199 Speaker 4: That's a dramatic difference in that AI is providing that 176 00:09:42,400 --> 00:09:45,600 Speaker 4: these other technological advances did not. They only went to 177 00:09:45,640 --> 00:09:49,920 Speaker 4: the efficiency piece and allowed us to produce what clients 178 00:09:49,960 --> 00:09:52,800 Speaker 4: want to buy a deal or alligation for a lower 179 00:09:53,000 --> 00:09:57,200 Speaker 4: total cost. But we were, now I believe, producing better 180 00:09:57,320 --> 00:09:58,760 Speaker 4: work product from the get go. 181 00:09:58,920 --> 00:10:03,679 Speaker 2: Okay, let's go back to a VC financing around. So 182 00:10:03,760 --> 00:10:08,720 Speaker 2: why don't you explain, even to say, pre AI, what 183 00:10:08,840 --> 00:10:11,160 Speaker 2: was the role of the lawyer or the law firm 184 00:10:11,559 --> 00:10:13,199 Speaker 2: in a typical financing around? 185 00:10:13,280 --> 00:10:13,439 Speaker 4: What? 186 00:10:13,559 --> 00:10:15,480 Speaker 2: Why? Why are their lawyers evolved? What were they what 187 00:10:15,520 --> 00:10:16,200 Speaker 2: are they brought. 188 00:10:15,960 --> 00:10:16,320 Speaker 4: In to do? 189 00:10:16,800 --> 00:10:19,559 Speaker 2: And then maybe sort of give us a concrete example 190 00:10:19,679 --> 00:10:23,000 Speaker 2: of today what it means or how a lawyer who's 191 00:10:23,040 --> 00:10:25,440 Speaker 2: in one of these deals can use AI as a 192 00:10:25,480 --> 00:10:26,480 Speaker 2: quote thought partner. 193 00:10:26,600 --> 00:10:28,480 Speaker 4: Yeah. So, so first I have to give you a 194 00:10:28,520 --> 00:10:30,920 Speaker 4: bit of a disclaimer. I'm a structured finance lawyer, not 195 00:10:31,040 --> 00:10:34,240 Speaker 4: a venture capital lawyer. But we have one of the 196 00:10:34,320 --> 00:10:37,160 Speaker 4: largest venture practices in the United States in our firm, 197 00:10:37,240 --> 00:10:42,760 Speaker 4: so I'm you know, I'm familiar. Okay, Right, what's the 198 00:10:42,840 --> 00:10:46,920 Speaker 4: role of the lawyer? So often when our client is 199 00:10:47,040 --> 00:10:51,439 Speaker 4: the company side or the founders, they've never done anything 200 00:10:51,480 --> 00:10:54,679 Speaker 4: like this before, and we are educating our client on 201 00:10:54,800 --> 00:10:58,760 Speaker 4: how deals work. They rely on us for market knowledge 202 00:10:59,200 --> 00:11:02,199 Speaker 4: of kind of what typical deal terms are in the 203 00:11:02,200 --> 00:11:05,439 Speaker 4: adventure financing, whether it's a serious C, the A B C. 204 00:11:07,240 --> 00:11:10,320 Speaker 4: And we because we are so active in the space, 205 00:11:10,440 --> 00:11:12,560 Speaker 4: we usually know the other you know, we usually know 206 00:11:12,640 --> 00:11:15,960 Speaker 4: the funds that are investing, we know their council, and 207 00:11:16,160 --> 00:11:19,600 Speaker 4: we can add I believe a lot of value to 208 00:11:20,320 --> 00:11:22,920 Speaker 4: the client in the in the way we can negotiate 209 00:11:23,040 --> 00:11:27,120 Speaker 4: and structure the deal knowing also what the next round 210 00:11:27,160 --> 00:11:28,800 Speaker 4: is going to look like, and the round after that 211 00:11:29,040 --> 00:11:32,160 Speaker 4: and their ultimate exit, whether it's to an IPO or 212 00:11:32,200 --> 00:11:34,960 Speaker 4: an M and A deal, And we can help them 213 00:11:35,280 --> 00:11:38,120 Speaker 4: get their structure right from the get go so that 214 00:11:38,200 --> 00:11:42,360 Speaker 4: they'll get all those future rounds right. And when clients 215 00:11:42,360 --> 00:11:45,960 Speaker 4: come to us with kind of their chat kept produced documents, 216 00:11:46,280 --> 00:11:49,200 Speaker 4: I think, go, you know, they're they're missing all of 217 00:11:49,200 --> 00:11:53,720 Speaker 4: that nuance, the knowledge of the market and the other 218 00:11:53,880 --> 00:11:59,000 Speaker 4: humans in the deal and what they need for future transactions. 219 00:11:59,440 --> 00:12:02,280 Speaker 3: So a lot of clients coming to you with chat 220 00:12:02,320 --> 00:12:03,840 Speaker 3: GPT produced stuff. 221 00:12:03,520 --> 00:12:09,160 Speaker 4: In the venture space, Yeah, I've right, And in a 222 00:12:09,240 --> 00:12:12,839 Speaker 4: number of areas. I mean, we have some very sophisticated 223 00:12:12,840 --> 00:12:17,520 Speaker 4: clients who have the same AI tools we have in 224 00:12:17,559 --> 00:12:21,200 Speaker 4: the legal profession and some fortune for example, a Fortune 225 00:12:21,280 --> 00:12:26,680 Speaker 4: fifty client that basically produces their first drafts of whether 226 00:12:26,720 --> 00:12:30,320 Speaker 4: it's a contract or a complaint or an answer in 227 00:12:30,400 --> 00:12:34,280 Speaker 4: a litigation. They all produce that using their AI tools 228 00:12:34,440 --> 00:12:38,640 Speaker 4: first send it to us and want us to take 229 00:12:38,679 --> 00:12:41,760 Speaker 4: that and run with it. On the other end. In 230 00:12:41,800 --> 00:12:44,600 Speaker 4: another area of our practice, we do a fair amount 231 00:12:44,600 --> 00:12:48,280 Speaker 4: of patent prosecution work, particularly out of our West Coast offices. 232 00:12:49,720 --> 00:12:54,360 Speaker 4: We represent some extremely sophisticated technology companies who will have 233 00:12:54,600 --> 00:12:58,720 Speaker 4: their AI tools reviewing our work and providing us with 234 00:12:59,200 --> 00:13:04,240 Speaker 4: AI produced comments. Now we're also using AI tools to 235 00:13:04,400 --> 00:13:09,280 Speaker 4: create these patents with full knowledge of our clients, so 236 00:13:09,559 --> 00:13:12,160 Speaker 4: we're almost at the point where their AI agent is 237 00:13:12,200 --> 00:13:15,959 Speaker 4: talking to our a AI agent. That's an interesting development 238 00:13:16,080 --> 00:13:16,760 Speaker 4: as well. 239 00:13:17,080 --> 00:13:21,280 Speaker 3: For sure on the AI copilot ideas, since your expertise 240 00:13:21,320 --> 00:13:25,080 Speaker 3: is in structured finance, how far does this actually go? 241 00:13:25,559 --> 00:13:31,079 Speaker 3: As in, could you envision asking an LLM to produce 242 00:13:31,240 --> 00:13:34,840 Speaker 3: like a brand new structure for I don't know, ABS 243 00:13:35,000 --> 00:13:40,760 Speaker 3: or cnbs or pick your structured finance poison. Like when 244 00:13:40,800 --> 00:13:44,560 Speaker 3: you talk about AI out with strategy and complexity. 245 00:13:44,040 --> 00:13:46,600 Speaker 4: I don't believe that it can come up with a 246 00:13:46,640 --> 00:13:52,040 Speaker 4: new structure. Okay, But I have a partner who is 247 00:13:52,200 --> 00:13:57,120 Speaker 4: a real expert in international tax for example, and he 248 00:13:57,280 --> 00:14:01,240 Speaker 4: will use AI tools to and he's always coming up 249 00:14:01,280 --> 00:14:05,599 Speaker 4: and he represents a lot of family offices and global businesses, 250 00:14:06,840 --> 00:14:09,360 Speaker 4: and he will come up with and he spends all 251 00:14:09,400 --> 00:14:12,960 Speaker 4: of his time structuring you know, how these family offices work. 252 00:14:13,040 --> 00:14:15,679 Speaker 4: How do you get money from country A to country 253 00:14:15,760 --> 00:14:19,400 Speaker 4: C with a minimum number of tax hops along the way. 254 00:14:19,800 --> 00:14:23,360 Speaker 4: He will use AI to test out some of his 255 00:14:23,560 --> 00:14:29,000 Speaker 4: theories and it will give him some feedback, and he'll 256 00:14:29,040 --> 00:14:32,160 Speaker 4: work with the AI to come up with new structures. 257 00:14:32,440 --> 00:14:36,600 Speaker 4: He will then hand the report that he gets out 258 00:14:36,600 --> 00:14:41,680 Speaker 4: of say Claude, and hand it to an associate and say, okay, 259 00:14:41,960 --> 00:14:45,920 Speaker 4: now run this down and actually do the research to 260 00:14:45,960 --> 00:14:49,000 Speaker 4: figure out if this is right. But let's work on 261 00:14:49,040 --> 00:14:53,640 Speaker 4: this structure together. So it increases the horizons. It validates 262 00:14:53,640 --> 00:14:58,000 Speaker 4: some stuff. What's the opposite of valid not validate? Knocks 263 00:14:58,000 --> 00:15:03,360 Speaker 4: out some ideas valid right, and then we still have 264 00:15:03,680 --> 00:15:06,480 Speaker 4: you know, a real human who has a law degree, 265 00:15:06,800 --> 00:15:10,080 Speaker 4: run it down and double check and see if we 266 00:15:10,080 --> 00:15:13,360 Speaker 4: can make it even better. And usually the iterative process 267 00:15:13,640 --> 00:15:16,800 Speaker 4: between the AI and the human and the associate and 268 00:15:16,880 --> 00:15:20,840 Speaker 4: the AI and the partner come up with some really 269 00:15:20,840 --> 00:15:23,440 Speaker 4: cool structures that they wouldn't have thought of on their own, 270 00:15:23,760 --> 00:15:25,840 Speaker 4: and that the AI wouldn't have thought of on its own. 271 00:15:25,920 --> 00:15:27,600 Speaker 2: This is gonna sound like a rude question, how do 272 00:15:27,640 --> 00:15:30,200 Speaker 2: you know? Because the question, like, I think it's like 273 00:15:30,440 --> 00:15:33,840 Speaker 2: I could see going back and forth with the chat 274 00:15:33,920 --> 00:15:37,160 Speaker 2: pot as a way to like stress test ideas or 275 00:15:37,280 --> 00:15:39,000 Speaker 2: sort of like do you think that you know just 276 00:15:39,040 --> 00:15:42,160 Speaker 2: sort of like you generate these new ideas, and so, like, 277 00:15:42,280 --> 00:15:46,400 Speaker 2: I guess on some level, I'm not surprised that by interacting, 278 00:15:46,480 --> 00:15:49,400 Speaker 2: let's say Claude, a really good tax lawyer, could like 279 00:15:49,720 --> 00:15:54,040 Speaker 2: explore some new potentials or like feel out this potential possibility. 280 00:15:54,120 --> 00:15:57,680 Speaker 2: But on the other hand, it's also easy to go 281 00:15:57,760 --> 00:15:59,960 Speaker 2: back and forth with the chatbot and create the illusion 282 00:16:00,040 --> 00:16:03,040 Speaker 2: that you're finding something new that you also could have 283 00:16:03,080 --> 00:16:06,240 Speaker 2: like intuitively arrived at yourself, because if you were one 284 00:16:06,240 --> 00:16:08,400 Speaker 2: of the most brilliant tax lawyers in the world, So 285 00:16:08,480 --> 00:16:10,640 Speaker 2: how do you sort of establish that this is in 286 00:16:10,760 --> 00:16:16,240 Speaker 2: fact adding value as opposed to just work ish or 287 00:16:16,280 --> 00:16:16,800 Speaker 2: work light. 288 00:16:17,280 --> 00:16:20,520 Speaker 4: So in the example, in the tax example, I just 289 00:16:20,640 --> 00:16:23,800 Speaker 4: gave you I have no idea, how okay, but I 290 00:16:23,840 --> 00:16:26,000 Speaker 4: can tell you because I was out in our Palo 291 00:16:26,040 --> 00:16:30,760 Speaker 4: Alto office just the week before last and was talking 292 00:16:30,800 --> 00:16:33,760 Speaker 4: to our law I'll go back to the patent example, 293 00:16:33,800 --> 00:16:37,600 Speaker 4: because that's discrete, right, and it's and and there are 294 00:16:37,800 --> 00:16:42,280 Speaker 4: discrete set of steps. We are using an AI tool 295 00:16:42,640 --> 00:16:47,680 Speaker 4: to help us to help our lawyers draft patent patent applications. 296 00:16:48,360 --> 00:16:50,280 Speaker 4: They have been telling me, we've been doing this for 297 00:16:50,320 --> 00:16:53,800 Speaker 4: a little over six months, that the primary benefit is 298 00:16:53,840 --> 00:16:59,280 Speaker 4: that it creates a better patent application because while almost 299 00:16:59,320 --> 00:17:03,760 Speaker 4: all of our patent lawyers are or engineers of some type, 300 00:17:03,840 --> 00:17:09,680 Speaker 4: mostly software electrical engineers, the knowledge that the AI tool 301 00:17:09,760 --> 00:17:13,120 Speaker 4: has is of all engineering, right, and of all sciences 302 00:17:13,160 --> 00:17:17,560 Speaker 4: and arts, and it's able to bring in a biologist's perspective, 303 00:17:18,000 --> 00:17:22,480 Speaker 4: perhaps just to use something and create a broader application. 304 00:17:22,880 --> 00:17:24,960 Speaker 4: I was like, Okay, that's kind of cool. I was 305 00:17:25,000 --> 00:17:27,800 Speaker 4: then visiting with a couple of our clients out there 306 00:17:28,240 --> 00:17:30,960 Speaker 4: who we do patent work for and on their own. 307 00:17:31,240 --> 00:17:34,880 Speaker 4: They have told us that they really appreciate how we're 308 00:17:34,960 --> 00:17:38,040 Speaker 4: using the technology, that there were that we are ahead 309 00:17:38,040 --> 00:17:39,879 Speaker 4: of most of the other firms that they're using that 310 00:17:39,920 --> 00:17:43,200 Speaker 4: they use right now, and that they have noticed over 311 00:17:43,200 --> 00:17:45,840 Speaker 4: the past six months, how our applications have gotten better 312 00:17:45,920 --> 00:17:50,080 Speaker 4: and better. So they I didn't ask them that question unprompted, 313 00:17:50,200 --> 00:17:51,960 Speaker 4: they have said that, so that I was I was 314 00:17:52,040 --> 00:17:55,320 Speaker 4: really excited about that because that was validation because I 315 00:17:55,359 --> 00:17:56,840 Speaker 4: asked the same question that you do. It is like, 316 00:17:56,920 --> 00:17:58,399 Speaker 4: how do you know it makes it any better? And 317 00:17:58,440 --> 00:18:03,600 Speaker 4: do clients notice? Clients noticing you know makes a difference. 318 00:18:19,320 --> 00:18:22,320 Speaker 3: One thing I'm really interested in is the impact of 319 00:18:22,359 --> 00:18:25,879 Speaker 3: AI on actual pricing power for law firms. And you 320 00:18:25,920 --> 00:18:28,800 Speaker 3: wrote a really good piece for our colleagues over at 321 00:18:28,800 --> 00:18:32,680 Speaker 3: Bloomberg Law called AI will give Junior Lawyers better work 322 00:18:33,520 --> 00:18:36,960 Speaker 3: one of our legal insights, and you mentioned a specific 323 00:18:37,080 --> 00:18:40,520 Speaker 3: number in there. You cite a project where you were 324 00:18:40,640 --> 00:18:43,240 Speaker 3: going to do it but you decided not to because 325 00:18:43,320 --> 00:18:46,919 Speaker 3: it was too expensive. But then a year later or so, 326 00:18:47,800 --> 00:18:50,040 Speaker 3: with the assistance of AI, you said that the cost 327 00:18:50,080 --> 00:18:54,560 Speaker 3: had come down seventy percent, which is absolutely huge. But 328 00:18:54,640 --> 00:18:57,760 Speaker 3: I guess my question is if costs are coming down 329 00:18:57,880 --> 00:19:02,840 Speaker 3: by that much, then why don't clients accrue all of 330 00:19:02,880 --> 00:19:06,480 Speaker 3: those cost savings? How do lawyers actually protect their margins 331 00:19:06,840 --> 00:19:07,800 Speaker 3: in that scenario? 332 00:19:09,160 --> 00:19:13,960 Speaker 4: So a couple a couple of ways. First off, the 333 00:19:14,080 --> 00:19:19,280 Speaker 4: clients don't necessarily have the same knowledge, deep industry knowledge 334 00:19:19,520 --> 00:19:24,440 Speaker 4: that they're outside lawyers do in reviewing the work because well, 335 00:19:24,480 --> 00:19:26,479 Speaker 4: the cost has come down seventy percent hasn't come down 336 00:19:26,480 --> 00:19:30,640 Speaker 4: one hundred percent, right, So there we're still doing thirty 337 00:19:30,680 --> 00:19:33,720 Speaker 4: percent of the hours that we might have done a 338 00:19:33,720 --> 00:19:38,560 Speaker 4: few years ago or something like that, but those hours 339 00:19:38,560 --> 00:19:41,080 Speaker 4: that we're putting in it are much higher level hours. 340 00:19:41,280 --> 00:19:43,840 Speaker 4: The project that I was talking about in that article 341 00:19:44,400 --> 00:19:48,520 Speaker 4: is basically a due diligence project where we were where 342 00:19:48,560 --> 00:19:52,919 Speaker 4: the assignment was to review thousands of trust agreements for 343 00:19:53,119 --> 00:19:59,440 Speaker 4: the obligations of various parties in the agreements usually laborious, 344 00:20:00,119 --> 00:20:05,480 Speaker 4: kind of boring, tedious. And the price we quoted or 345 00:20:05,520 --> 00:20:08,920 Speaker 4: the cost that we quoted to do it I think 346 00:20:08,920 --> 00:20:12,760 Speaker 4: it was like three years ago. Was based on you know, 347 00:20:12,920 --> 00:20:16,800 Speaker 4: humans doing all of the work and a QC team 348 00:20:16,880 --> 00:20:20,040 Speaker 4: on top of the first round. We can basically take 349 00:20:20,119 --> 00:20:24,399 Speaker 4: out that first layer or review and assign that to 350 00:20:24,640 --> 00:20:28,320 Speaker 4: the robot which then puts it all in this beautiful 351 00:20:28,480 --> 00:20:32,360 Speaker 4: one hundred one hundred columns spreadsheet to show us all 352 00:20:32,400 --> 00:20:35,880 Speaker 4: of the fields. And we're doing the QC layer, and 353 00:20:36,280 --> 00:20:41,639 Speaker 4: the QC layer requires legal knowledge of how these transactions work, 354 00:20:42,080 --> 00:20:46,840 Speaker 4: and some experience that you can spot where the output 355 00:20:46,920 --> 00:20:50,679 Speaker 4: doesn't make sense and you go back and double check it. 356 00:20:51,160 --> 00:20:55,760 Speaker 4: So I think the clients still value having a law 357 00:20:55,800 --> 00:21:01,200 Speaker 4: firm basically certify or you know, say this is good 358 00:21:01,200 --> 00:21:03,719 Speaker 4: output in a way that they wouldn't want to do themselves. 359 00:21:03,760 --> 00:21:06,080 Speaker 4: They don't have the staffing to do it themselves, and 360 00:21:06,440 --> 00:21:08,199 Speaker 4: they want outside eyes on it. 361 00:21:08,560 --> 00:21:13,280 Speaker 2: So what does this mean specifically though? For early career lawyers? 362 00:21:13,320 --> 00:21:15,760 Speaker 2: All right, like the fear is or one version of 363 00:21:15,800 --> 00:21:18,320 Speaker 2: this is, like, that's great for the senior lawyers, pull 364 00:21:18,400 --> 00:21:22,399 Speaker 2: up the ladder, hire few lawgreds, et cetera, because you 365 00:21:22,440 --> 00:21:26,040 Speaker 2: can now do this, Like is that? What does that 366 00:21:26,119 --> 00:21:28,720 Speaker 2: mean for the people who three years ago would have 367 00:21:28,760 --> 00:21:31,800 Speaker 2: been doing this? You know, as you said, not the 368 00:21:31,800 --> 00:21:33,200 Speaker 2: most exciting, tedious work. 369 00:21:33,320 --> 00:21:37,280 Speaker 4: Right again, when I started, I was blacklining documents by hand, 370 00:21:37,800 --> 00:21:41,280 Speaker 4: and somehow, when the machines took that job away from me, 371 00:21:41,320 --> 00:21:43,240 Speaker 4: I still had plenty of work to do, and first 372 00:21:43,320 --> 00:21:45,159 Speaker 4: your lawyers behind me had plenty of work to do. 373 00:21:45,800 --> 00:21:47,480 Speaker 4: But my life got more interesting. 374 00:21:48,080 --> 00:21:49,520 Speaker 2: So I think. 375 00:21:49,320 --> 00:21:54,720 Speaker 4: That the tedious jobs are going away. First, junior lawyers, 376 00:21:54,920 --> 00:21:57,760 Speaker 4: I think are going to be doing more interesting work sooner. 377 00:21:58,480 --> 00:22:01,160 Speaker 4: We have to. We have not fully figured out yet 378 00:22:01,240 --> 00:22:04,520 Speaker 4: and we will solve, but we haven't full yet how 379 00:22:04,560 --> 00:22:08,359 Speaker 4: to train people without going through that tedious work. 380 00:22:09,840 --> 00:22:11,920 Speaker 2: So that are you still are you adding? Are you 381 00:22:12,040 --> 00:22:16,760 Speaker 2: hiring one l's and two or whatever ls? Are they 382 00:22:16,800 --> 00:22:19,400 Speaker 2: in law school? We are screds like right now. 383 00:22:19,320 --> 00:22:21,719 Speaker 4: We are hiring summer We have a you know, we 384 00:22:21,760 --> 00:22:25,760 Speaker 4: have summer associates. This summer we will have a full 385 00:22:26,040 --> 00:22:28,600 Speaker 4: slate of first year associates who join us in the fall. 386 00:22:29,119 --> 00:22:32,320 Speaker 4: We have already hired our summer associate class for the 387 00:22:32,359 --> 00:22:37,880 Speaker 4: summer of twenty twenty seven and we have not changed 388 00:22:37,960 --> 00:22:41,399 Speaker 4: or hiring patterns we have. We are starting to change 389 00:22:41,520 --> 00:22:46,320 Speaker 4: the skill set we're looking for and what you need 390 00:22:46,600 --> 00:22:49,080 Speaker 4: to be successful in this profession. I think we'll change 391 00:22:49,119 --> 00:22:49,640 Speaker 4: a little bit. 392 00:22:49,920 --> 00:22:52,359 Speaker 2: Well, can you talk Dan about what is that skill 393 00:22:52,400 --> 00:22:55,360 Speaker 2: set and what are they going to be doing when 394 00:22:55,359 --> 00:22:57,520 Speaker 2: they get when they arrived there on their first day 395 00:22:57,560 --> 00:23:00,400 Speaker 2: after law school? What is this skill set they need 396 00:23:00,440 --> 00:23:01,800 Speaker 2: to first get in the door, and then what are 397 00:23:01,800 --> 00:23:03,040 Speaker 2: they going to do once they're in the door. 398 00:23:03,160 --> 00:23:05,080 Speaker 4: I think, no matter what, you're going to need some 399 00:23:05,280 --> 00:23:10,320 Speaker 4: training on the area of practice you're in and there's 400 00:23:10,359 --> 00:23:14,080 Speaker 4: going to be less of the gruntwork kind of training 401 00:23:14,680 --> 00:23:18,160 Speaker 4: and we're ultimately going to have more simulator type training 402 00:23:18,359 --> 00:23:21,760 Speaker 4: or you know what we've already started doing. You know, 403 00:23:21,960 --> 00:23:25,080 Speaker 4: a very basic, a very standard like first year training 404 00:23:25,119 --> 00:23:28,199 Speaker 4: curriculum at a law firm is like the uh the 405 00:23:28,200 --> 00:23:31,120 Speaker 4: anatomy of an M and a transaction where you actually 406 00:23:31,160 --> 00:23:35,359 Speaker 4: go through the provisions of a merger agreement or an 407 00:23:35,359 --> 00:23:38,280 Speaker 4: A or a stock purchase agreement or an asset purchase 408 00:23:38,320 --> 00:23:41,280 Speaker 4: agreement and you go through it, you know those provisions. 409 00:23:41,680 --> 00:23:44,720 Speaker 4: We're now going to What we've started doing is make 410 00:23:44,760 --> 00:23:47,360 Speaker 4: sure everybody brings their laptop with them to the training 411 00:23:47,880 --> 00:23:51,199 Speaker 4: and that they have their AI tools open at the 412 00:23:51,480 --> 00:23:55,240 Speaker 4: at the same time and as part of the training, 413 00:23:55,560 --> 00:23:58,480 Speaker 4: you're saying, Okay, now take this provision and put it 414 00:23:58,560 --> 00:24:02,040 Speaker 4: in this tool and see what the responses are. See 415 00:24:02,600 --> 00:24:05,719 Speaker 4: tell it you're representing a seller, and give me a 416 00:24:05,800 --> 00:24:08,600 Speaker 4: seller favorable provision to negotiate it. So we're going to 417 00:24:08,640 --> 00:24:11,600 Speaker 4: do a lot more of that training hands on with 418 00:24:11,800 --> 00:24:15,960 Speaker 4: the tools where people will get to see that we 419 00:24:16,000 --> 00:24:19,040 Speaker 4: will have they will have in front of them that 420 00:24:19,080 --> 00:24:23,440 Speaker 4: they don't have now, but they will very soon our 421 00:24:23,680 --> 00:24:27,400 Speaker 4: knowledge set of every similar deal that we have done, 422 00:24:27,840 --> 00:24:30,359 Speaker 4: say in the past couple of years, that have been 423 00:24:30,400 --> 00:24:33,000 Speaker 4: fully negotiated. They will be able to see the terms 424 00:24:33,160 --> 00:24:36,520 Speaker 4: of every deal we've recently done. Clients care a lot 425 00:24:36,680 --> 00:24:40,600 Speaker 4: about our market knowledge. They care what are the reasons 426 00:24:40,880 --> 00:24:44,240 Speaker 4: that they're willing to pay law firms at the exorbitant 427 00:24:44,320 --> 00:24:48,800 Speaker 4: rates we charge. Is because you have at the market 428 00:24:48,800 --> 00:24:51,960 Speaker 4: knowledge that they can't possibly have on their own, because 429 00:24:52,119 --> 00:24:55,119 Speaker 4: for many of them this is a one time deal. 430 00:24:55,640 --> 00:24:58,080 Speaker 4: For us, we do it every day, and they expect 431 00:24:58,160 --> 00:25:01,560 Speaker 4: us to know what the market terms are. Using this technology, 432 00:25:01,720 --> 00:25:06,040 Speaker 4: I can make sure every junior lawyer has access to 433 00:25:06,320 --> 00:25:10,439 Speaker 4: all of the knowledge of our entire firm in doing 434 00:25:10,520 --> 00:25:15,360 Speaker 4: similar types of deals or litigation or bankruptcy proceedings. As 435 00:25:15,359 --> 00:25:18,080 Speaker 4: opposed to just having a walk through the halls and 436 00:25:18,119 --> 00:25:21,679 Speaker 4: try to find somebody who's done it before and just 437 00:25:21,760 --> 00:25:24,160 Speaker 4: talk to them about what they've done. You will now 438 00:25:24,240 --> 00:25:29,400 Speaker 4: have access to basically a dashboard with all with all 439 00:25:29,400 --> 00:25:32,760 Speaker 4: prior documents, and I think that makes you much smarter faster. 440 00:25:34,600 --> 00:25:37,400 Speaker 4: We could have a separate conversation about the cognitive load 441 00:25:37,480 --> 00:25:39,879 Speaker 4: from that right. There was a lot of value in 442 00:25:39,920 --> 00:25:43,280 Speaker 4: my having downtime to black line, right, it was like relaxing. 443 00:25:43,359 --> 00:25:45,320 Speaker 3: This was going to be my next question, which is 444 00:25:45,640 --> 00:25:47,879 Speaker 3: lawyers have to be very detail oriented. 445 00:25:48,000 --> 00:25:48,600 Speaker 2: Yes, right. 446 00:25:48,640 --> 00:25:51,160 Speaker 3: And one of the arguments for having to do all 447 00:25:51,160 --> 00:25:54,880 Speaker 3: this grunt work like hand deliver is to agreements or 448 00:25:55,680 --> 00:25:58,520 Speaker 3: review trusts and things like that was that, like it 449 00:25:58,560 --> 00:26:01,560 Speaker 3: teaches you to focus, Yeah, and it teaches you to 450 00:26:01,720 --> 00:26:04,320 Speaker 3: look out for that missing comma that's gonna cost you 451 00:26:04,359 --> 00:26:05,639 Speaker 3: like five million dollars. You know. 452 00:26:05,960 --> 00:26:07,879 Speaker 2: I was thinking about that as you were saying that 453 00:26:08,200 --> 00:26:11,120 Speaker 2: dead documentary on Netflix or like fifteen years ago about 454 00:26:11,160 --> 00:26:13,399 Speaker 2: the sushi guy or it's like all these like chefs 455 00:26:13,400 --> 00:26:16,040 Speaker 2: she It's like, Okay, I'm gonna teach you to make sushi. 456 00:26:16,320 --> 00:26:20,080 Speaker 2: First spend fifteen years like learning how like clean rice 457 00:26:20,200 --> 00:26:22,040 Speaker 2: or whatever it is, and then we like get to sushi. 458 00:26:22,320 --> 00:26:24,760 Speaker 2: Or like think about musicians, and it's like it's no 459 00:26:24,920 --> 00:26:28,359 Speaker 2: fun to do scales, right, but you do it and 460 00:26:28,400 --> 00:26:31,720 Speaker 2: you build some sort of like deep understanding, and it's 461 00:26:31,720 --> 00:26:33,520 Speaker 2: like that feels like the equivalent of what like the 462 00:26:33,520 --> 00:26:34,280 Speaker 2: black lining is. 463 00:26:34,359 --> 00:26:37,600 Speaker 3: Yeah, So what happens if people aren't doing as many 464 00:26:37,640 --> 00:26:38,680 Speaker 3: repetitive tasks? 465 00:26:40,520 --> 00:26:44,480 Speaker 4: I worry about that as well. And I'm not sure 466 00:26:44,920 --> 00:26:49,920 Speaker 4: what the answer is from a behavioral psychology perspective, because 467 00:26:49,960 --> 00:26:53,840 Speaker 4: I think you're right, doing those scales is important. I 468 00:26:53,840 --> 00:26:58,400 Speaker 4: don't know, you know, we're talking about simulators to simulate 469 00:26:58,440 --> 00:27:02,560 Speaker 4: doing a deal. There's know, you know, there's nothing like 470 00:27:02,760 --> 00:27:07,600 Speaker 4: actually doing a deal right and actually negotiating with somebody 471 00:27:07,600 --> 00:27:09,560 Speaker 4: on the other side who's a jerk, or somebody on 472 00:27:09,560 --> 00:27:12,640 Speaker 4: the other side who's really nice but you know, bamboozles 473 00:27:12,680 --> 00:27:16,640 Speaker 4: you by being so nice. And the human element of 474 00:27:16,680 --> 00:27:21,520 Speaker 4: this all is still really important, and the human interaction 475 00:27:22,000 --> 00:27:25,480 Speaker 4: with your client with the adversary, you know, you need 476 00:27:25,560 --> 00:27:27,760 Speaker 4: to get that experience, and you need to have the 477 00:27:27,840 --> 00:27:32,920 Speaker 4: touch and feel. With the advances in AI, just over 478 00:27:32,960 --> 00:27:35,359 Speaker 4: the past, you know, six months or a year, you 479 00:27:35,400 --> 00:27:40,200 Speaker 4: can see a time when the drafting is largely delegated 480 00:27:40,240 --> 00:27:43,000 Speaker 4: to the machine. You still have to be careful, as 481 00:27:43,040 --> 00:27:45,040 Speaker 4: you said, the commas. You know, I can see AI 482 00:27:45,160 --> 00:27:48,120 Speaker 4: screwing up commas. Ais don't know how to add right 483 00:27:48,160 --> 00:27:50,280 Speaker 4: and they make silly mistakes all the time, So you 484 00:27:50,359 --> 00:27:55,040 Speaker 4: got to learn to be careful. I do worry about 485 00:27:55,600 --> 00:27:59,159 Speaker 4: it becoming too easy to just fall into the trap 486 00:27:59,240 --> 00:28:04,480 Speaker 4: of trusting what it tells you rather than thinking above it. 487 00:28:04,600 --> 00:28:06,840 Speaker 4: So I don't know the answer to your question, but 488 00:28:06,920 --> 00:28:08,280 Speaker 4: I have the same concern you do. 489 00:28:09,240 --> 00:28:11,800 Speaker 2: I want to get into like the various tools and 490 00:28:11,920 --> 00:28:14,600 Speaker 2: infrastructure you have, you know, because we're Markets and Finance 491 00:28:14,640 --> 00:28:17,480 Speaker 2: podcast if people want to know what the trade is obviously, 492 00:28:17,840 --> 00:28:23,360 Speaker 2: but before we get into that specifically, you know, when 493 00:28:23,359 --> 00:28:27,080 Speaker 2: we did our last episode about AI law, our guest 494 00:28:27,480 --> 00:28:30,360 Speaker 2: Joel Werdheimer, he's a civil rights lawyer in New York City, 495 00:28:30,359 --> 00:28:31,960 Speaker 2: and one of the points that he made is that 496 00:28:32,080 --> 00:28:37,119 Speaker 2: sometimes clients come with him with cases, often maybe like 497 00:28:37,200 --> 00:28:40,880 Speaker 2: suing the city over like police misconduct or something like that, 498 00:28:41,320 --> 00:28:43,840 Speaker 2: and the client has a good case, might have a 499 00:28:43,880 --> 00:28:46,239 Speaker 2: good case on or you're aware of a client might 500 00:28:46,280 --> 00:28:49,440 Speaker 2: have a good case on paper, but the potential damages 501 00:28:49,480 --> 00:28:52,640 Speaker 2: are so small that it's an uneconomical to pursue it, 502 00:28:52,680 --> 00:28:57,640 Speaker 2: even though clearly there's a legitimate case. If we're sort 503 00:28:57,640 --> 00:29:00,960 Speaker 2: of as Tracy mentioned, the Jeff's parad of all this, 504 00:29:01,880 --> 00:29:05,000 Speaker 2: do you see this, it's like, okay, you collapse the 505 00:29:05,040 --> 00:29:08,920 Speaker 2: price of certain types of legal work. Does that expand 506 00:29:09,680 --> 00:29:13,000 Speaker 2: the number of theoretical cases? Or I guess you're not 507 00:29:13,200 --> 00:29:16,800 Speaker 2: deals that. Oh let's take a look at this deal. 508 00:29:16,920 --> 00:29:19,840 Speaker 2: There's some like minimum upfront, cause that's going to be 509 00:29:19,840 --> 00:29:21,960 Speaker 2: the same. Let's take a look at this deal that 510 00:29:22,080 --> 00:29:25,440 Speaker 2: then compensates for the reduced number of hours that you charge. 511 00:29:25,600 --> 00:29:30,200 Speaker 4: I think so, I think that the Deevins paradox applies 512 00:29:30,440 --> 00:29:35,680 Speaker 4: to legal In the example I gave you with the 513 00:29:35,720 --> 00:29:38,280 Speaker 4: client who had thousands of trust agreements that needs to 514 00:29:38,280 --> 00:29:41,160 Speaker 4: be reviewed. When we gave them the first quote, we 515 00:29:41,240 --> 00:29:43,640 Speaker 4: didn't say we didn't want to do it. The client said, 516 00:29:43,680 --> 00:29:46,120 Speaker 4: at that price, we're not going to do it because 517 00:29:46,600 --> 00:29:49,719 Speaker 4: the risk, the downside risk isn't high enough to justify 518 00:29:49,720 --> 00:29:52,680 Speaker 4: the price. But when the price came down by seventy percent, 519 00:29:53,000 --> 00:29:55,400 Speaker 4: they said, oh yeah, at that price, we'll do it. 520 00:29:55,920 --> 00:30:00,760 Speaker 4: So at and let's just use fake numbers. Ten million 521 00:30:00,800 --> 00:30:03,239 Speaker 4: dollars they said, no, we're not going to do it, 522 00:30:03,360 --> 00:30:07,080 Speaker 4: and we had our revenue was zero. When it came 523 00:30:07,080 --> 00:30:09,880 Speaker 4: down to three million, they said, that's worth it, and 524 00:30:09,920 --> 00:30:11,240 Speaker 4: we had three million of revenue. 525 00:30:11,720 --> 00:30:11,840 Speaker 1: Uh. 526 00:30:12,120 --> 00:30:16,080 Speaker 4: So that's the Deevins paradox in action. Over the past 527 00:30:16,120 --> 00:30:21,320 Speaker 4: twenty five years, E Discovery has ballooned the cost of litigation. 528 00:30:21,480 --> 00:30:25,440 Speaker 4: To your point, right, Electronic electronic discovery run amok has 529 00:30:26,040 --> 00:30:31,600 Speaker 4: has made it virtually impossible to bring a sophisticated litigation 530 00:30:31,720 --> 00:30:34,200 Speaker 4: for under a few million dollars in legal fees and 531 00:30:34,280 --> 00:30:37,520 Speaker 4: in just the discovery expenses alone, or millions of dollars. 532 00:30:37,880 --> 00:30:42,440 Speaker 4: If we can bring say two million of discovery costs 533 00:30:42,440 --> 00:30:47,400 Speaker 4: down to two hundred thousand, it does change the calculus 534 00:30:47,760 --> 00:30:49,840 Speaker 4: for both on this. 535 00:30:50,120 --> 00:30:52,920 Speaker 2: On this, lawsuits are a lot more economical to file. 536 00:30:53,040 --> 00:30:54,800 Speaker 2: Everyone good, just what America. 537 00:30:54,880 --> 00:30:57,760 Speaker 4: Everyone is gonna love that aspect of Isn't that fantastic? 538 00:30:59,480 --> 00:31:03,280 Speaker 4: But you know, you raised it in a virtuous, you 539 00:31:03,320 --> 00:31:08,200 Speaker 4: know example with a civil rights lawyer. But I think 540 00:31:08,200 --> 00:31:12,960 Speaker 4: it's also true, you know, for large corporates who were 541 00:31:13,000 --> 00:31:16,840 Speaker 4: abandoning things before. So yeah, you can debate whether it's 542 00:31:16,880 --> 00:31:19,920 Speaker 4: a social good to have more litigation, But if I'm 543 00:31:19,960 --> 00:31:22,400 Speaker 4: speaking on behalf of a law firm that has a 544 00:31:22,440 --> 00:31:28,520 Speaker 4: large litigation department, I do think that the increased the 545 00:31:28,560 --> 00:31:34,920 Speaker 4: potential for increased litigation does outweigh the loss of really boring, 546 00:31:35,160 --> 00:31:39,160 Speaker 4: mundane hours reviewing documents in the I'll go back to 547 00:31:39,160 --> 00:31:43,040 Speaker 4: the patent example I gave you earlier, where we do see, 548 00:31:43,080 --> 00:31:45,920 Speaker 4: over time, the cost of being able to prosecute a 549 00:31:45,920 --> 00:31:49,880 Speaker 4: patent obligation coming down dramatically through the use of AI, 550 00:31:50,680 --> 00:31:53,320 Speaker 4: and that that same client who told me two weeks 551 00:31:53,320 --> 00:31:55,920 Speaker 4: ago that they've really noticed an increase in the quality 552 00:31:55,920 --> 00:31:59,840 Speaker 4: of a work product also told us that because they're 553 00:32:00,040 --> 00:32:05,080 Speaker 4: using AI in their own work, they're seeing a quadrupling 554 00:32:05,280 --> 00:32:09,200 Speaker 4: in the number of inventions that their engineers and scientists 555 00:32:09,240 --> 00:32:14,040 Speaker 4: are coming up with, and the number of patent requests 556 00:32:14,080 --> 00:32:18,800 Speaker 4: that they're getting internally from their engineering team has quadrupled, 557 00:32:19,000 --> 00:32:21,680 Speaker 4: and we're going to ultimately be able to lower the 558 00:32:21,720 --> 00:32:25,640 Speaker 4: cost of each of those, so the clients are able 559 00:32:25,680 --> 00:32:30,080 Speaker 4: to do more and create more economic activity, which is virtuous, 560 00:32:30,320 --> 00:32:33,920 Speaker 4: and a lot of that will require some legal work, 561 00:32:34,080 --> 00:32:37,160 Speaker 4: even though each unit of legal work will cost less, 562 00:32:38,360 --> 00:32:41,680 Speaker 4: I expect we'll be doing more of it, so you 563 00:32:41,720 --> 00:32:45,640 Speaker 4: know that should preserve the ability of lawyers to do 564 00:32:45,760 --> 00:32:47,800 Speaker 4: okay economically. 565 00:32:48,360 --> 00:32:51,440 Speaker 3: Well, how do you see AI shifting? I guess the 566 00:32:51,440 --> 00:32:55,040 Speaker 3: balance of power or balance of work between external law 567 00:32:55,080 --> 00:32:58,320 Speaker 3: firms versus in house lawyers. Because you could imagine a 568 00:32:58,360 --> 00:33:01,480 Speaker 3: scenario where I'm a big company, I have a couple 569 00:33:02,080 --> 00:33:07,120 Speaker 3: or maybe five big ex big law lawyers who work 570 00:33:07,160 --> 00:33:09,280 Speaker 3: for me, and because so much of the work is 571 00:33:09,280 --> 00:33:12,120 Speaker 3: now automated. They can use AI. They don't need to 572 00:33:12,120 --> 00:33:14,400 Speaker 3: go to a big law firm to do a lot 573 00:33:14,440 --> 00:33:16,680 Speaker 3: of the grunt work as you put it earlier. 574 00:33:17,120 --> 00:33:21,840 Speaker 4: Yeah, and I think you're going to see the the uh. 575 00:33:22,720 --> 00:33:25,160 Speaker 4: So the other side of the Devons paradox is you 576 00:33:25,200 --> 00:33:29,440 Speaker 4: will see larger clients that have dedicated legal teams keeping 577 00:33:29,560 --> 00:33:33,640 Speaker 4: more of the routine legal work in house. You know, 578 00:33:33,720 --> 00:33:38,040 Speaker 4: you mentioned is DOES before. Most banks have, for example, 579 00:33:38,120 --> 00:33:41,520 Speaker 4: have brought all their ISDA and derivatives work in house. 580 00:33:41,840 --> 00:33:44,640 Speaker 4: You'll probably see more of that, uh and you'll see 581 00:33:44,640 --> 00:33:48,959 Speaker 4: more end users on the fun side doing that internally 582 00:33:49,080 --> 00:33:51,400 Speaker 4: rather than sending it out. Although we have a terrific 583 00:33:51,600 --> 00:33:54,760 Speaker 4: you know, derivatives group. You may see that in some 584 00:33:54,880 --> 00:33:58,680 Speaker 4: other areas as well. But again, where you really need 585 00:33:59,040 --> 00:34:03,960 Speaker 4: market preadth and you need market knowledge, I think outside 586 00:34:04,040 --> 00:34:06,000 Speaker 4: law firms will still be able to provide a lot 587 00:34:06,000 --> 00:34:09,360 Speaker 4: of value and a lot of outside kind of independent 588 00:34:09,440 --> 00:34:28,600 Speaker 4: perspective and independent advice that it's hard to do inside. 589 00:34:30,360 --> 00:34:33,680 Speaker 2: Let's talk about your tech stack so to speak. You 590 00:34:33,719 --> 00:34:37,279 Speaker 2: mentioned lawyers talking to Claude. I know that there are 591 00:34:37,360 --> 00:34:41,400 Speaker 2: applications or companies like Harvey which are sort of like 592 00:34:41,719 --> 00:34:46,000 Speaker 2: built on top of some of the frontier models. There 593 00:34:46,000 --> 00:34:49,359 Speaker 2: are questions about do you need these application layers. I'm 594 00:34:49,400 --> 00:34:54,080 Speaker 2: also interested in whether, like when it comes to ingesting 595 00:34:54,120 --> 00:34:58,440 Speaker 2: your own internal data, whether like open source AI models 596 00:34:58,480 --> 00:34:59,920 Speaker 2: will play a role in that. What do you give 597 00:35:00,120 --> 00:35:03,560 Speaker 2: us like the overview of your firm, Like what is 598 00:35:03,600 --> 00:35:06,399 Speaker 2: the sort of essence of your tech stack that you're using? 599 00:35:09,800 --> 00:35:11,360 Speaker 4: So you actually want me to name products? 600 00:35:11,440 --> 00:35:12,239 Speaker 2: Yeah? Sure? Yeah? 601 00:35:12,239 --> 00:35:12,520 Speaker 4: Why not? 602 00:35:12,960 --> 00:35:15,080 Speaker 2: Okay, yeah, you mentioned Claude, but I'm really I did 603 00:35:15,200 --> 00:35:15,880 Speaker 2: miss Claude. 604 00:35:15,960 --> 00:35:18,520 Speaker 4: But we also so you know, we use a lot 605 00:35:18,560 --> 00:35:21,880 Speaker 4: of different products at the moment, probably too many, and 606 00:35:22,040 --> 00:35:26,240 Speaker 4: over time that will narrow. We are we are Harvey shop, Okay, 607 00:35:27,200 --> 00:35:31,680 Speaker 4: we uh, almost all of our users, our lawyers use 608 00:35:31,760 --> 00:35:35,760 Speaker 4: Harvey regularly. We also use Microsoft Copilot in the outlook 609 00:35:35,840 --> 00:35:37,000 Speaker 4: suite or the office. 610 00:35:37,040 --> 00:35:41,200 Speaker 2: Can you for our non lawyer listeners in me and 611 00:35:41,200 --> 00:35:43,719 Speaker 2: I assume most of us, Like, what does a Harvey 612 00:35:43,960 --> 00:35:47,640 Speaker 2: offer that is not offered when someone interfaces with the 613 00:35:47,680 --> 00:35:48,840 Speaker 2: model directly. 614 00:35:49,600 --> 00:35:51,960 Speaker 4: With the model being a frontier model. 615 00:35:52,920 --> 00:35:57,520 Speaker 2: Instead of or GPT five point five. What happens when 616 00:35:57,520 --> 00:36:00,200 Speaker 2: I use a model that sort of like that powers. 617 00:35:59,840 --> 00:36:01,840 Speaker 4: Hard within Harvey and I think this is true of 618 00:36:01,960 --> 00:36:05,480 Speaker 4: Legora as well, the other big one. Yeah, you can 619 00:36:05,520 --> 00:36:08,000 Speaker 4: pick which frontier model you want to use. You can 620 00:36:08,120 --> 00:36:12,680 Speaker 4: choose Chat or Claude for example, and I think Gemini 621 00:36:12,719 --> 00:36:16,200 Speaker 4: and some of them as well. But what a Harvey 622 00:36:16,880 --> 00:36:21,359 Speaker 4: ads is number one, a security layer that is much 623 00:36:21,400 --> 00:36:26,800 Speaker 4: more robust than the other models, and that is super 624 00:36:26,800 --> 00:36:31,560 Speaker 4: important for a law firm. We're able to control our 625 00:36:31,640 --> 00:36:34,319 Speaker 4: own data set and make sure that it's not going 626 00:36:34,400 --> 00:36:37,520 Speaker 4: up to the Internet, and that our client information is 627 00:36:37,600 --> 00:36:40,400 Speaker 4: not going out to the internet at all. We have 628 00:36:40,480 --> 00:36:42,880 Speaker 4: sessions that do not even connect to the Internet. It 629 00:36:43,080 --> 00:36:45,960 Speaker 4: is much more robust. And one of the things that 630 00:36:46,000 --> 00:36:50,200 Speaker 4: our clients often don't realize is that if they're using, 631 00:36:50,320 --> 00:36:54,800 Speaker 4: especially a consumer grade Claude or Chat GPT, they often 632 00:36:54,880 --> 00:36:59,680 Speaker 4: are losing atturning client privilege by asking a consumer model. 633 00:37:00,080 --> 00:37:03,239 Speaker 4: There are questions. We are number one focused on the 634 00:37:03,239 --> 00:37:07,160 Speaker 4: security layer and the ethics layer, and that's super important. 635 00:37:07,160 --> 00:37:10,880 Speaker 4: And a Harvey or a Legora or you know, West 636 00:37:10,920 --> 00:37:16,719 Speaker 4: Law has co Council all are very you know, security conscious. 637 00:37:16,800 --> 00:37:18,719 Speaker 4: So it gives us that it allows us to have 638 00:37:18,800 --> 00:37:24,080 Speaker 4: playbooks that we can share among among our team so 639 00:37:24,120 --> 00:37:27,319 Speaker 4: that people can see the projects that we're building, which 640 00:37:27,320 --> 00:37:29,480 Speaker 4: I think is harder to share in some of these 641 00:37:29,520 --> 00:37:34,239 Speaker 4: other models directly. And what at least Harvey provides is 642 00:37:34,239 --> 00:37:38,560 Speaker 4: basically a rag layer on top of the AI frontier. 643 00:37:38,120 --> 00:37:42,880 Speaker 2: Model augmented generation of Wow. Okay, I don't know, yeah. 644 00:37:42,960 --> 00:37:45,440 Speaker 4: I don't know if the I barely even know what 645 00:37:45,480 --> 00:37:47,440 Speaker 4: it means. But I think it's good to say it's 646 00:37:47,440 --> 00:37:50,520 Speaker 4: a rag layer, right that the that has been trained 647 00:37:50,560 --> 00:37:54,320 Speaker 4: on legal stuff, right, so it's much more fine tune 648 00:37:54,400 --> 00:37:56,640 Speaker 4: to law and out of the box. It comes with 649 00:37:56,719 --> 00:37:59,000 Speaker 4: some of the things lawyers like to do, like look 650 00:37:59,040 --> 00:38:03,160 Speaker 4: at clauses a merger and acquisition agreement. It has a 651 00:38:03,160 --> 00:38:08,760 Speaker 4: better understanding of nuance in reviewing deposition transcripts. But we're seeing, 652 00:38:09,239 --> 00:38:12,759 Speaker 4: you know, some of the main uses are creating a 653 00:38:12,760 --> 00:38:15,480 Speaker 4: project for if you're a litigator. I'll give some litigators 654 00:38:15,480 --> 00:38:19,000 Speaker 4: some airtime too here, not just talk about deals. Loading 655 00:38:19,160 --> 00:38:23,879 Speaker 4: all of the pleatings, the briefs, all of the transcripts, 656 00:38:23,960 --> 00:38:27,320 Speaker 4: and the back and forth between lawyers into a project 657 00:38:27,600 --> 00:38:30,520 Speaker 4: in a Harvey, and then when you are drafting a 658 00:38:30,600 --> 00:38:35,000 Speaker 4: pleting or a brief, it's able to retrieve for you 659 00:38:35,040 --> 00:38:38,960 Speaker 4: the quotes in the deposition transcripts that support what you're 660 00:38:38,960 --> 00:38:43,000 Speaker 4: trying to argue. So rather than spending you know, tens 661 00:38:43,120 --> 00:38:47,600 Speaker 4: or hundreds of hours trying to find those, it's using 662 00:38:47,719 --> 00:38:50,480 Speaker 4: the intelligence to find it for you. Claude could do 663 00:38:50,520 --> 00:38:53,040 Speaker 4: that as well, but Harvey does it I think a 664 00:38:53,080 --> 00:38:57,680 Speaker 4: little better, and it's collecting knowledge and allowing you to 665 00:38:57,760 --> 00:39:00,000 Speaker 4: share the playbooks more easily. 666 00:39:00,760 --> 00:39:02,719 Speaker 3: Yeah, I wanted to ask you about the security and 667 00:39:02,840 --> 00:39:06,960 Speaker 3: privacy aspect of all of this. So how much of 668 00:39:07,040 --> 00:39:11,520 Speaker 3: AI adoption in the legal industry is actually I mean, 669 00:39:11,719 --> 00:39:16,160 Speaker 3: law is a very heavily regulated industry itself, So how 670 00:39:16,239 --> 00:39:20,480 Speaker 3: much AI adoption is I guess constrained by things like 671 00:39:20,800 --> 00:39:27,600 Speaker 3: malpractice risk, you know, data concerns, accountability. Is that a 672 00:39:27,640 --> 00:39:28,759 Speaker 3: real limitation for you? 673 00:39:29,680 --> 00:39:35,960 Speaker 4: So two years ago the UH and I in my role, 674 00:39:36,200 --> 00:39:40,759 Speaker 4: I talked to our malpractice carriers and our underwriters out 675 00:39:40,800 --> 00:39:43,400 Speaker 4: of Lloyd's of London and go visit with them, and 676 00:39:43,440 --> 00:39:46,760 Speaker 4: they always ask questions about the things that they're worried about. 677 00:39:47,320 --> 00:39:51,440 Speaker 4: Two years ago, they were worried about, are you you know, 678 00:39:51,520 --> 00:39:55,719 Speaker 4: are you letting your lawyers use AI? And you know, 679 00:39:55,840 --> 00:39:56,640 Speaker 4: is that creating risks? 680 00:39:56,640 --> 00:39:56,920 Speaker 3: For us? 681 00:39:57,320 --> 00:40:01,439 Speaker 4: Now they're asking you like your lawyers use AI right 682 00:40:01,520 --> 00:40:04,120 Speaker 4: to make sure they're doing things right, right, because it's 683 00:40:04,160 --> 00:40:08,279 Speaker 4: becoming part of mainstream and it is becoming an assumption 684 00:40:09,080 --> 00:40:11,719 Speaker 4: that you're not going to be doing some of these 685 00:40:11,760 --> 00:40:16,120 Speaker 4: tasks without having AI assist you in doing them. So 686 00:40:16,760 --> 00:40:21,480 Speaker 4: it's going from a A insurers saying, oh, don't use 687 00:40:21,520 --> 00:40:24,080 Speaker 4: that stuff until it's proven too you have to be 688 00:40:24,160 --> 00:40:26,279 Speaker 4: using it, and the clients are doing the same thing right. 689 00:40:26,320 --> 00:40:28,520 Speaker 4: Clients at eighteen months ago were saying, don't use AI 690 00:40:28,640 --> 00:40:30,600 Speaker 4: for our work and now saying, well, you've got to 691 00:40:30,600 --> 00:40:32,680 Speaker 4: be using AI to try to cut down the cost 692 00:40:32,719 --> 00:40:36,480 Speaker 4: of the work. So you know, we've seen that shift 693 00:40:36,600 --> 00:40:40,520 Speaker 4: over the past two years pretty dramatically. That has changed. 694 00:40:40,880 --> 00:40:45,239 Speaker 4: But clearly, you know, we emphasize to our lawyers all 695 00:40:45,280 --> 00:40:49,160 Speaker 4: the time that you got to be double checking this stuff. 696 00:40:49,520 --> 00:40:54,360 Speaker 4: And there's no excuse for filing a brief or pleading 697 00:40:54,400 --> 00:40:58,480 Speaker 4: in a court that sites you know, made up hallucinated cases. 698 00:40:59,160 --> 00:41:00,440 Speaker 3: They're just toobarrassing. 699 00:41:00,640 --> 00:41:05,319 Speaker 4: It is embarrassing. We have too many tools at our 700 00:41:05,760 --> 00:41:09,759 Speaker 4: immediate disposal to be able to catch catch that, and yeah, 701 00:41:09,840 --> 00:41:13,600 Speaker 4: systems break down and and somebody is trying to draft 702 00:41:13,640 --> 00:41:17,239 Speaker 4: something and file it you know, in an hour and 703 00:41:17,280 --> 00:41:21,200 Speaker 4: they get hacinate cases slip in, but it's not. 704 00:41:21,880 --> 00:41:23,160 Speaker 2: So I want to ask you a question I asked 705 00:41:23,440 --> 00:41:27,040 Speaker 2: Goldenman David Solomons on an episode. One of the things 706 00:41:27,040 --> 00:41:30,840 Speaker 2: we're seeing actually in a lot of industries, including the 707 00:41:30,920 --> 00:41:35,040 Speaker 2: legal industry, is like the rise of like these superstar hires, 708 00:41:35,120 --> 00:41:38,080 Speaker 2: right and you see them like reported and various publications, 709 00:41:38,360 --> 00:41:41,480 Speaker 2: some big name lawyer gets some crazy salary to go 710 00:41:41,520 --> 00:41:44,680 Speaker 2: to or signing bonus or whatever kind of like you know, 711 00:41:44,719 --> 00:41:47,320 Speaker 2: like we see in some of the hedge fund space. 712 00:41:48,000 --> 00:41:50,239 Speaker 2: When I see these things, and when I think about AI, 713 00:41:50,320 --> 00:41:52,759 Speaker 2: and you mentioned, it's like, oh, here's this knowledge pool 714 00:41:52,840 --> 00:41:55,239 Speaker 2: or here's every We're gonna the junior lawyers are going 715 00:41:55,320 --> 00:41:57,880 Speaker 2: to get access to all these templates. And I'm thinking, 716 00:41:57,920 --> 00:42:00,920 Speaker 2: like if I'm this guy, like do I want to 717 00:42:00,960 --> 00:42:04,000 Speaker 2: share all my templates and knowledge with the firm? But 718 00:42:04,080 --> 00:42:07,800 Speaker 2: if I don't, if everyone thinks that, then these tools 719 00:42:07,840 --> 00:42:08,439 Speaker 2: don't really work. 720 00:42:08,520 --> 00:42:08,680 Speaker 1: Right. 721 00:42:08,680 --> 00:42:11,440 Speaker 2: If you don't have a culture of like contributing into 722 00:42:11,480 --> 00:42:14,680 Speaker 2: sight and knowledge back to the firm, there's no franchise 723 00:42:14,800 --> 00:42:16,960 Speaker 2: value and you're definitely not going to get anything out 724 00:42:17,000 --> 00:42:19,520 Speaker 2: of AI or in terms of like meaningful. I would think, 725 00:42:19,680 --> 00:42:22,480 Speaker 2: maybe we'll get some Does this create any tensions? And 726 00:42:22,520 --> 00:42:24,719 Speaker 2: when you're thinking about the firm, maybe not this year, 727 00:42:24,760 --> 00:42:27,319 Speaker 2: but down the road. How to make sure that the 728 00:42:27,360 --> 00:42:30,080 Speaker 2: interests of the firm are aligned with the interests of 729 00:42:30,080 --> 00:42:30,640 Speaker 2: the partners. 730 00:42:30,920 --> 00:42:35,040 Speaker 4: That is a huge issue. It's not a technological issue, 731 00:42:35,040 --> 00:42:36,000 Speaker 4: it's a human issue. 732 00:42:36,040 --> 00:42:36,239 Speaker 1: Right. 733 00:42:36,800 --> 00:42:42,000 Speaker 4: Lawyers tend to be highly autonomous creatures, right, and successful 734 00:42:42,040 --> 00:42:46,560 Speaker 4: law firm partners tend to be even more highly autonomous 735 00:42:46,840 --> 00:42:48,799 Speaker 4: and generally take the view you're not going to tell 736 00:42:48,840 --> 00:42:49,879 Speaker 4: me how to practice law. 737 00:42:50,200 --> 00:42:52,600 Speaker 3: My husband used to get calls from when he was 738 00:42:52,600 --> 00:42:54,400 Speaker 3: a junior lawyer. He would get calls from one of 739 00:42:54,400 --> 00:42:56,320 Speaker 3: the partners out on their fishing boat. 740 00:42:56,480 --> 00:42:57,880 Speaker 2: Oh yeah, on weekend. 741 00:42:58,040 --> 00:43:00,600 Speaker 3: It's assigning work autonomous. 742 00:43:00,680 --> 00:43:01,799 Speaker 4: Your husband worked at our firm? 743 00:43:02,040 --> 00:43:05,279 Speaker 2: No not. Do you think like an evil you kill 744 00:43:05,400 --> 00:43:10,719 Speaker 2: environment in many cases? Like yeah, right? 745 00:43:11,880 --> 00:43:15,319 Speaker 4: And you know, many people would say, especially at our firm, 746 00:43:15,360 --> 00:43:18,399 Speaker 4: that one of the secrets to our success is that 747 00:43:18,719 --> 00:43:21,919 Speaker 4: we a term we use oftense. We let the horses run, right, 748 00:43:22,000 --> 00:43:25,920 Speaker 4: you let people go do their thing, and clients love 749 00:43:26,000 --> 00:43:30,520 Speaker 4: that and love the entrepreneurial energy. And now with AI 750 00:43:30,760 --> 00:43:33,480 Speaker 4: were saying we needed to share all your knowledge. So 751 00:43:33,520 --> 00:43:38,560 Speaker 4: the culture the culture clash there is significant interesting and 752 00:43:38,640 --> 00:43:41,640 Speaker 4: I think that's I think that's a real challenge. Now 753 00:43:41,840 --> 00:43:45,560 Speaker 4: when you talk about the NBA type hires or NFL 754 00:43:45,680 --> 00:43:49,080 Speaker 4: type you know salaries that you're seeing, perhaps some of 755 00:43:49,080 --> 00:43:53,839 Speaker 4: the motivation for that is getting their knowledge into their 756 00:43:54,320 --> 00:43:57,600 Speaker 4: into the systems. Right if you can, if you can 757 00:43:57,640 --> 00:43:59,960 Speaker 4: get some of these superstars and kind of put their 758 00:44:00,160 --> 00:44:04,320 Speaker 4: name on the UH on your AI tools and say 759 00:44:04,360 --> 00:44:07,400 Speaker 4: that you know, so and so recently did a deal 760 00:44:07,920 --> 00:44:11,520 Speaker 4: that had these terms. You know, how's the client's going 761 00:44:11,600 --> 00:44:14,040 Speaker 4: to be super impressed by that? And maybe that's part 762 00:44:14,040 --> 00:44:18,120 Speaker 4: of the value add UH and maybe that creates even 763 00:44:18,360 --> 00:44:20,239 Speaker 4: higher comp numbers for superstars. 764 00:44:20,719 --> 00:44:22,520 Speaker 3: I want to go back to I guess my first 765 00:44:22,600 --> 00:44:26,000 Speaker 3: question about billable hours in the age of AI. If 766 00:44:26,000 --> 00:44:28,839 Speaker 3: we're seeing pricing actually compressed, if we're seeing a sort 767 00:44:28,880 --> 00:44:35,240 Speaker 3: of maybe democratization of legal tools, maybe uncertainty over case 768 00:44:35,280 --> 00:44:40,239 Speaker 3: results starts to collapse as well, would you maybe finally 769 00:44:40,320 --> 00:44:43,279 Speaker 3: see a shift away from the billable hours thing? 770 00:44:43,920 --> 00:44:48,040 Speaker 4: Maybe so, UH, and you should, right, we should be 771 00:44:48,200 --> 00:44:54,080 Speaker 4: able to move to project based pricing or outcome based pricing. 772 00:44:54,719 --> 00:44:58,239 Speaker 4: I think that that that will finally take off, there 773 00:44:58,280 --> 00:45:01,000 Speaker 4: will still be a lot of instance is where hourly 774 00:45:01,080 --> 00:45:01,760 Speaker 4: rates apply. 775 00:45:02,480 --> 00:45:03,600 Speaker 2: And I think you're. 776 00:45:03,440 --> 00:45:07,120 Speaker 4: Going to see a continuing huge increase in hourly rates 777 00:45:07,200 --> 00:45:10,200 Speaker 4: at top tier firms. And I think we were already 778 00:45:10,719 --> 00:45:13,720 Speaker 4: seeing it. If you look at data from twenty twenty five, 779 00:45:13,800 --> 00:45:17,719 Speaker 4: as published by the American Lawyer, average hourly rates in 780 00:45:17,760 --> 00:45:20,000 Speaker 4: twenty twenty five went up by ten point one percent 781 00:45:20,440 --> 00:45:24,000 Speaker 4: at the largest firms in the United States, and at 782 00:45:24,040 --> 00:45:28,040 Speaker 4: the top the top twenty firms by profitability, they went 783 00:45:28,120 --> 00:45:32,040 Speaker 4: up even more. And that doesn't make sense in a 784 00:45:32,120 --> 00:45:34,960 Speaker 4: year when CPI went up by around three percent, right, 785 00:45:35,000 --> 00:45:38,040 Speaker 4: I've never seen that kind of gap between CPI and 786 00:45:38,120 --> 00:45:41,720 Speaker 4: average rates. One of the things that might explain it 787 00:45:41,800 --> 00:45:45,200 Speaker 4: is that the billable hour has actually become more productive 788 00:45:45,239 --> 00:45:48,759 Speaker 4: and more valuable because of AI tools, and I think 789 00:45:48,840 --> 00:45:51,960 Speaker 4: you'll see that continuing at an ever faster pace. So 790 00:45:52,040 --> 00:45:55,160 Speaker 4: I would expect those hourly rates will continue to go up. 791 00:45:55,640 --> 00:45:58,280 Speaker 4: Yet the price for the things that clients will actually 792 00:45:58,320 --> 00:46:01,040 Speaker 4: want to buy, which are solution so the business problems 793 00:46:01,080 --> 00:46:03,080 Speaker 4: are going to come down because they will take far 794 00:46:03,200 --> 00:46:04,239 Speaker 4: fewer hours. 795 00:46:04,600 --> 00:46:06,799 Speaker 2: I want to ask you one more question about your 796 00:46:07,200 --> 00:46:11,840 Speaker 2: your tech stack. You know, there's a couple other things 797 00:46:11,920 --> 00:46:17,360 Speaker 2: that are sort of happening. One is obviously this idea 798 00:46:17,520 --> 00:46:20,680 Speaker 2: that well, if you use an open source model unlike 799 00:46:20,840 --> 00:46:26,080 Speaker 2: say some claude, then you can really like train bake 800 00:46:26,160 --> 00:46:30,359 Speaker 2: all your knowledge into the model itself and so you're 801 00:46:30,400 --> 00:46:33,520 Speaker 2: not just sort of doing the retrieval augmented generation where 802 00:46:33,560 --> 00:46:37,000 Speaker 2: the model essentially is like combined with a search engine, 803 00:46:37,080 --> 00:46:39,440 Speaker 2: but it's actually like baked in and you can actually 804 00:46:39,440 --> 00:46:41,840 Speaker 2: do that. And you see, it was a really interesting 805 00:46:42,080 --> 00:46:45,600 Speaker 2: article paper from Bridgewater et cetera recently where they talked 806 00:46:45,600 --> 00:46:49,759 Speaker 2: about doing this. And then of course, like just in 807 00:46:49,800 --> 00:46:53,200 Speaker 2: general setting aside open source versus closed source and model, 808 00:46:53,280 --> 00:46:55,799 Speaker 2: it's like, we do see this thing, and we mentioned 809 00:46:55,880 --> 00:46:58,840 Speaker 2: Kirkland Ellis in the beginning. Eventually you hit a scale 810 00:46:58,840 --> 00:47:00,600 Speaker 2: where it's like no, you want to I just you know, 811 00:47:00,640 --> 00:47:04,560 Speaker 2: you don't just want to be build your own Harvey 812 00:47:04,680 --> 00:47:07,319 Speaker 2: or whatever it is and actually internalize some of the 813 00:47:07,360 --> 00:47:12,040 Speaker 2: infrastructure or spend and like actually truly customize this, etcetera. 814 00:47:12,080 --> 00:47:15,520 Speaker 2: And arguably it's never been easier to build software et cetera. 815 00:47:15,920 --> 00:47:19,400 Speaker 2: I'm curious, like, okay, as a four hundred person firm, 816 00:47:19,880 --> 00:47:22,759 Speaker 2: obviously in your view, getting a lot of value already 817 00:47:22,840 --> 00:47:26,879 Speaker 2: out of AI, what comes the point where you would think, 818 00:47:26,960 --> 00:47:30,279 Speaker 2: you know what, we want to start building some technology 819 00:47:30,480 --> 00:47:36,640 Speaker 2: in house, including perhaps a homegrown version of one of 820 00:47:36,680 --> 00:47:40,480 Speaker 2: the models trained on your proprietary data. 821 00:47:40,640 --> 00:47:44,680 Speaker 4: So we definitely want to connect these proprietary models to 822 00:47:44,840 --> 00:47:48,280 Speaker 4: our data and use that to help train the model. 823 00:47:48,680 --> 00:47:53,520 Speaker 4: But you know, training from scratch a model does it. 824 00:47:53,840 --> 00:47:57,640 Speaker 4: Kirkland announced they're spending five hundred million dollars over five years. Right, 825 00:47:58,040 --> 00:47:59,920 Speaker 4: It costs them like a billion and a half dollars 826 00:48:00,000 --> 00:48:02,600 Speaker 4: to train a model one and a half billion, right, 827 00:48:03,480 --> 00:48:06,560 Speaker 4: So even that would take Kirkland fifteen years of investment. 828 00:48:06,800 --> 00:48:10,520 Speaker 4: I don't see us really training our own models. I 829 00:48:10,560 --> 00:48:15,280 Speaker 4: see us doing some customized solutions on top of existing models. 830 00:48:16,239 --> 00:48:18,759 Speaker 4: I see us certainly doing our playbooks so that they'll 831 00:48:18,800 --> 00:48:22,879 Speaker 4: be the Lowenstein Sandler M and a playbook for that's 832 00:48:22,960 --> 00:48:26,840 Speaker 4: customized for our firm and our market knowledge. But I 833 00:48:26,880 --> 00:48:30,600 Speaker 4: don't I don't think we're gonna really build our own 834 00:48:30,800 --> 00:48:33,279 Speaker 4: as much as modify and customize. 835 00:48:35,360 --> 00:48:36,080 Speaker 2: That's not happening. 836 00:48:36,120 --> 00:48:39,560 Speaker 4: We've always I mean, you know, we might white label 837 00:48:39,680 --> 00:48:41,719 Speaker 4: an existing one to come up with a fancy name, 838 00:48:42,000 --> 00:48:45,440 Speaker 4: but I don't think that's realistic for a firm of 839 00:48:45,480 --> 00:48:48,280 Speaker 4: our size. I don't think it's even realistic for a Kirkland. 840 00:48:48,360 --> 00:48:50,839 Speaker 4: So well, you know, we'll see how that shakes out, 841 00:48:51,640 --> 00:48:53,839 Speaker 4: but I don't see us building our own I do 842 00:48:53,960 --> 00:48:56,640 Speaker 4: want to mention one thing though, that's related to both 843 00:48:56,640 --> 00:48:59,680 Speaker 4: of your points. All of these models that law firms 844 00:48:59,719 --> 00:49:03,160 Speaker 4: have been using have been at the enterprise level, have 845 00:49:03,239 --> 00:49:07,160 Speaker 4: been basically all you can eat pricing. Right, they're all 846 00:49:07,200 --> 00:49:10,160 Speaker 4: going to be switching to token based pricing. You know 847 00:49:10,280 --> 00:49:11,520 Speaker 4: Claude is doing it first. 848 00:49:12,680 --> 00:49:15,759 Speaker 2: But do you think you could see sticker shocks like, oh, 849 00:49:16,200 --> 00:49:17,840 Speaker 2: you know what I thought we were getting that. I 850 00:49:17,880 --> 00:49:19,560 Speaker 2: thought we were able to do all these trusts at 851 00:49:19,600 --> 00:49:21,640 Speaker 2: thirty percent of the cost, but it turned out that 852 00:49:21,719 --> 00:49:24,440 Speaker 2: was subsidized. It turns out there's actually gonna be one 853 00:49:24,480 --> 00:49:26,520 Speaker 2: hundred and thirty percent of the cost or ninety percent 854 00:49:26,560 --> 00:49:27,000 Speaker 2: of the cost. 855 00:49:27,120 --> 00:49:29,640 Speaker 4: So I don't think we yet know what the actual 856 00:49:29,760 --> 00:49:33,600 Speaker 4: cost is. That's not if it's not subsidized by outside investors. 857 00:49:34,040 --> 00:49:34,239 Speaker 1: Right. 858 00:49:34,640 --> 00:49:39,560 Speaker 4: I don't know that we really know what the optimal 859 00:49:39,640 --> 00:49:44,680 Speaker 4: combination of labor and capital and you know, robot capital 860 00:49:44,880 --> 00:49:48,960 Speaker 4: is because we don't know the true cost, especially if we, 861 00:49:49,360 --> 00:49:53,040 Speaker 4: you know, kind of hit this energy grid wall where 862 00:49:53,320 --> 00:49:55,520 Speaker 4: it just gets more and more expensive. 863 00:49:55,280 --> 00:49:58,479 Speaker 2: And today run these July eighth, you do not yet 864 00:49:58,520 --> 00:50:01,239 Speaker 2: have visibility from any of your partners into like the 865 00:50:01,280 --> 00:50:03,200 Speaker 2: true cost of this stuff. I mean, if it's like 866 00:50:03,440 --> 00:50:06,480 Speaker 2: if if they have to make a profit per token. 867 00:50:06,600 --> 00:50:11,880 Speaker 4: Right, we don't know. You know, I know from from 868 00:50:11,960 --> 00:50:13,840 Speaker 4: my IT person who I was talking to you yesterday 869 00:50:14,200 --> 00:50:17,200 Speaker 4: that so far this month, I've my use of Claude 870 00:50:17,200 --> 00:50:20,920 Speaker 4: has used up thirty six dollars. But that's what July 871 00:50:21,000 --> 00:50:23,440 Speaker 4: fourth weekend in there, right, And it was only this 872 00:50:23,719 --> 00:50:26,080 Speaker 4: on the seventh, there only three business days that I 873 00:50:26,120 --> 00:50:28,920 Speaker 4: was actually using it. But so that's not not not so. 874 00:50:28,960 --> 00:50:32,160 Speaker 2: Much tillable time. Yeah, exactly exactly. 875 00:50:32,400 --> 00:50:35,440 Speaker 4: But I don't know where that pricing is ultimately going 876 00:50:35,480 --> 00:50:38,600 Speaker 4: to go, and uh so we don't we don't know 877 00:50:38,600 --> 00:50:40,680 Speaker 4: how that story is going to play out two years 878 00:50:40,680 --> 00:50:41,120 Speaker 4: down the road. 879 00:50:41,880 --> 00:50:44,000 Speaker 2: He Garry wingoins, thank you so much. That was a 880 00:50:44,000 --> 00:50:46,319 Speaker 2: really helpful conversation. Really appreciate it. 881 00:50:46,400 --> 00:50:48,319 Speaker 4: Great being here, So thanks for having me. 882 00:51:02,000 --> 00:51:04,640 Speaker 2: Tracy. It's really interesting that we really don't even know 883 00:51:04,680 --> 00:51:07,680 Speaker 2: anything about the economics of AI yet because of like 884 00:51:07,800 --> 00:51:10,759 Speaker 2: how much like the you know, these all you can 885 00:51:10,840 --> 00:51:13,040 Speaker 2: eat models and like what how that's all going to 886 00:51:13,040 --> 00:51:14,160 Speaker 2: shake out? I don't know. 887 00:51:14,280 --> 00:51:16,279 Speaker 3: Well, it reminds me a lot of you know, what 888 00:51:16,440 --> 00:51:20,440 Speaker 3: happened with Uber and the food delivery apps, where basically 889 00:51:21,000 --> 00:51:25,080 Speaker 3: venture capital was subsidizing everyone's ability to order food at home, 890 00:51:25,200 --> 00:51:27,920 Speaker 3: and then when it actually came time to prove the 891 00:51:27,960 --> 00:51:30,880 Speaker 3: business model and generate some profit, we saw the cost 892 00:51:30,920 --> 00:51:34,240 Speaker 3: of food delivery go up and use of it go down. 893 00:51:34,800 --> 00:51:38,200 Speaker 2: Yeah. Also same with oil. Yeah, and the fact that 894 00:51:38,239 --> 00:51:41,400 Speaker 2: like capital markets subsidized losses for years and years with 895 00:51:41,520 --> 00:51:45,560 Speaker 2: oil and now we're getting you know, similar examples there. 896 00:51:45,880 --> 00:51:48,759 Speaker 2: And then on the flip side, so let's say that 897 00:51:48,960 --> 00:51:52,840 Speaker 2: like you know, the historical prices, like token prices continue 898 00:51:52,840 --> 00:51:56,000 Speaker 2: to drop at a very high rate. On the flip side, 899 00:51:56,080 --> 00:51:58,640 Speaker 2: we get a lot more lawsuits and so on the 900 00:51:58,640 --> 00:52:03,080 Speaker 2: flip side, like this is like, okay, this is you know, 901 00:52:03,120 --> 00:52:06,359 Speaker 2: the Jeffens paradox is Like, good news, it turns out 902 00:52:06,360 --> 00:52:08,640 Speaker 2: there's going to be plenty of human labor in the future. 903 00:52:08,920 --> 00:52:11,320 Speaker 2: Bad news, there's going to be one hundred more frivolous 904 00:52:11,360 --> 00:52:14,960 Speaker 2: lawsuits and patent counter challenges that you never had to do, 905 00:52:15,000 --> 00:52:16,040 Speaker 2: and that's what we're doing now. 906 00:52:16,160 --> 00:52:18,040 Speaker 3: This is what I worry about on a wider scale, 907 00:52:18,120 --> 00:52:21,799 Speaker 3: because if we talk about a AI as this productivity enhancement, 908 00:52:21,920 --> 00:52:24,800 Speaker 3: you could have a situation where AI is just used 909 00:52:24,840 --> 00:52:29,200 Speaker 3: to expand bureaucracy forever and ever and ever, so not 910 00:52:29,280 --> 00:52:33,399 Speaker 3: just frivolous lawsuits, but like maybe human resources things like that. 911 00:52:33,440 --> 00:52:36,279 Speaker 3: I always wanted to write an article about how like 912 00:52:36,400 --> 00:52:38,440 Speaker 3: human resources ate the US economy. 913 00:52:38,600 --> 00:52:42,480 Speaker 2: Yeah, well think about like, you know, you could imagine like, okay, 914 00:52:42,560 --> 00:52:45,480 Speaker 2: every worker at a firm gets access to claude or 915 00:52:45,520 --> 00:52:48,439 Speaker 2: something like that, and on day one, it's like, oh, 916 00:52:48,440 --> 00:52:51,160 Speaker 2: this is great, Like I just condensed eight hours of 917 00:52:51,239 --> 00:52:54,680 Speaker 2: labor into two hours of labor, et cetera. But then suddenly, 918 00:52:54,719 --> 00:52:57,520 Speaker 2: like all of the other people that they're working with, 919 00:52:57,560 --> 00:53:01,120 Speaker 2: who are fighting for the same promotions also condensed eight 920 00:53:01,160 --> 00:53:04,040 Speaker 2: hours of labor into two hours, and so it's like, well, 921 00:53:04,120 --> 00:53:06,560 Speaker 2: I'm trying to get that promotion and you just like 922 00:53:06,719 --> 00:53:10,799 Speaker 2: find more and more work that scenario in which like 923 00:53:11,080 --> 00:53:14,120 Speaker 2: all of us suddenly feel like life is easier because 924 00:53:14,160 --> 00:53:17,279 Speaker 2: of AI. Like I gotta say that feels like you 925 00:53:17,280 --> 00:53:19,440 Speaker 2: know what I did actually have a I needed to 926 00:53:19,800 --> 00:53:24,320 Speaker 2: change a flight recently and change a round trip ticket 927 00:53:24,320 --> 00:53:26,719 Speaker 2: to a three way ticket at the last minute, and 928 00:53:26,760 --> 00:53:28,640 Speaker 2: that was like a very daunting thing to me to 929 00:53:28,680 --> 00:53:32,319 Speaker 2: do on the Delta dot com website and so like 930 00:53:32,440 --> 00:53:35,799 Speaker 2: I asked, like Chadubt the steps like what on and 931 00:53:36,040 --> 00:53:38,480 Speaker 2: like was really helpful is like go to this page 932 00:53:38,520 --> 00:53:40,160 Speaker 2: and the bottom right there will be a thing. This 933 00:53:40,200 --> 00:53:42,640 Speaker 2: is what you click to like change one leg of 934 00:53:42,680 --> 00:53:45,799 Speaker 2: the thing, and like it worked exactly as it So 935 00:53:45,840 --> 00:53:47,520 Speaker 2: it's like there was an example where it's like, Okay, 936 00:53:47,520 --> 00:53:52,440 Speaker 2: this actually eased some psychic attacks. But generally speaking, this 937 00:53:52,560 --> 00:53:55,439 Speaker 2: future where it's like life just gets much easier because 938 00:53:55,480 --> 00:53:57,879 Speaker 2: of AI, feels like at a minimum, it's a long 939 00:53:57,920 --> 00:53:58,279 Speaker 2: way off. 940 00:53:58,360 --> 00:54:01,560 Speaker 3: No, the future is a Jeffens parrot for like administrative 941 00:54:01,560 --> 00:54:05,080 Speaker 3: overhead and busy works. That's how I feel at the moment. 942 00:54:05,120 --> 00:54:05,879 Speaker 3: Shall we leave it there? 943 00:54:05,960 --> 00:54:06,680 Speaker 2: Let's leave it there. 944 00:54:06,880 --> 00:54:09,359 Speaker 3: This has been another episode of the Oudlots podcast. I'm 945 00:54:09,360 --> 00:54:12,400 Speaker 3: Tracy Alloway. You can follow me at Tracy Alloway. 946 00:54:12,120 --> 00:54:14,680 Speaker 2: And I'm Jill Wisenthal. You can follow me at the Stalwart. 947 00:54:14,960 --> 00:54:18,160 Speaker 2: Follow our producers Carmen Rodriguez at Kerman armand dash Ol 948 00:54:18,160 --> 00:54:21,719 Speaker 2: Bennett at dashbot, cal Brooks at Kilbrooks and Kevin Lozano 949 00:54:21,800 --> 00:54:24,560 Speaker 2: at Kevin Lloyd Losana and from our Odd Lots content. 950 00:54:24,600 --> 00:54:26,640 Speaker 2: Go to Bloomberg dot com slash odd Lots or of 951 00:54:26,640 --> 00:54:29,040 Speaker 2: a daily newsletter and all of our episodes and you 952 00:54:29,040 --> 00:54:31,200 Speaker 2: could chat about all these topics twenty four seven in 953 00:54:31,320 --> 00:54:34,480 Speaker 2: our discord Discord dot gig slash onlines. 954 00:54:34,760 --> 00:54:36,960 Speaker 3: And if you enjoy Oddlots, if you like it when 955 00:54:36,960 --> 00:54:39,080 Speaker 3: we talk about the future of the legal industry, then 956 00:54:39,120 --> 00:54:42,400 Speaker 3: please leave us a positive review on your favorite podcast platform. 957 00:54:42,760 --> 00:54:45,080 Speaker 3: And remember, if you are a Bloomberg subscriber, you can 958 00:54:45,120 --> 00:54:48,239 Speaker 3: listen to all of our episodes absolutely ad free. All 959 00:54:48,239 --> 00:54:50,200 Speaker 3: you need to do is find the Bloomberg channel on 960 00:54:50,239 --> 00:55:11,600 Speaker 3: Apple Podcasts and follow the instructions there. Thanks for listening 961 00:55:03,520 --> 00:55:03,560 Speaker 3: it