1 00:00:00,151 --> 00:00:02,334 Richard Trumans, how are you this afternoon? 2 00:00:02,597 --> 00:00:04,721 ah Good, good, good. 3 00:00:04,721 --> 00:00:12,334 Slightly frazzled around the edges, but I've been turned over and I'm a sunny side up again now, so I'm pretty good. 4 00:00:12,334 --> 00:00:14,156 uh 5 00:00:14,156 --> 00:00:14,527 good. 6 00:00:14,527 --> 00:00:16,123 That's better than sunny side down. 7 00:00:16,123 --> 00:00:17,593 um 8 00:00:17,593 --> 00:00:18,194 I don't like that. 9 00:00:18,194 --> 00:00:20,225 mean, yeah, it seems that will go scrambled. 10 00:00:20,225 --> 00:00:20,945 There you go. 11 00:00:20,945 --> 00:00:27,589 Yeah, I'm a over medium kind of guy, whatever floats your boat. 12 00:00:27,589 --> 00:00:35,326 So I know you, think most of our audience knows you, but for the ones that don't, just give us a quick introduction about who you are, what you do, and where you do it. 13 00:00:35,326 --> 00:00:38,238 Yeah, well, I'll give you the sensible two minute version. 14 00:00:38,238 --> 00:00:47,874 So working in the legal sector for about 25 years, started off as a proper journalist, went to journalism school for my sins, ended up at a magazine called Legal Week, which 15 00:00:47,874 --> 00:00:52,837 doesn't exist anymore, which was bought eventually by law.com, focused on the international world. 16 00:00:52,837 --> 00:00:55,278 Then I became a foreign correspondent. 17 00:00:55,278 --> 00:01:02,122 After a couple of years decided that I'd much rather focus on the business of law, ended up in the city. 18 00:01:02,170 --> 00:01:05,450 Chem and management consultant helped to merge law firms together. 19 00:01:05,450 --> 00:01:19,745 It was very much focused on research of the market, big macroeconomic picture, did a lot of work over the years on profitability, business structure, business culture, and all of 20 00:01:19,745 --> 00:01:20,506 those things. 21 00:01:20,506 --> 00:01:24,869 And really, if I had been happy doing that, you probably would never have met me. 22 00:01:24,869 --> 00:01:28,873 In fact, no one in the legal tech world would ever, ever have met me. 23 00:01:28,873 --> 00:01:30,734 Artificial law would not exist. 24 00:01:30,790 --> 00:01:38,960 and I would be living in Surrey in a nice little cottage and commuting into London to do my little bits of consulting work with big law firms. 25 00:01:38,960 --> 00:01:42,043 And I'd still be wearing a suit and all of that kind of stuff. 26 00:01:42,043 --> 00:01:48,479 But, fatefully, in 2015, I went and started getting interested in technology. 27 00:01:48,479 --> 00:01:53,793 And I became fascinated with AI and I've always been interested in science in my free time. 28 00:01:53,915 --> 00:01:59,247 And I, in 2016, it bubbled up to the point where I suddenly realized, my God, it's going to happen. 29 00:01:59,247 --> 00:02:00,768 It's really going to happen. 30 00:02:00,768 --> 00:02:02,169 AI is going to change everything. 31 00:02:02,169 --> 00:02:10,373 And I launched Artificial Lawyer, which combined a bit of my business analysis and macro picture skills with my good old fashioned journalistic skills. 32 00:02:10,373 --> 00:02:18,776 People kind of think that I went from being a journalist to running Artificial Lawyer, but there's like an enormous gap in the middle of running around the city in a suit, trying to 33 00:02:18,776 --> 00:02:21,667 advise people on business, which no one ever talks about. 34 00:02:21,667 --> 00:02:23,580 understandably, it wasn't very exciting. 35 00:02:23,580 --> 00:02:29,905 It was it was to begin with it is I mean, honestly, when you've been a reporter, going, tell me about your profitability. 36 00:02:29,905 --> 00:02:38,693 And then you get invited inside the inner sanctum and they show you your your you know, they show you the spreadsheet and you see everything you like, wow, I you get to talk to 37 00:02:38,693 --> 00:02:41,216 managing partners and practice heads and everything. 38 00:02:41,216 --> 00:02:44,019 You get a completely different sense of what's going on. 39 00:02:44,019 --> 00:02:44,579 And 40 00:02:44,579 --> 00:02:53,259 That for me is why I think those two things when AI came around, I just thought, yes, this is going to change everything. 41 00:02:53,259 --> 00:02:59,799 And then of course it didn't, but I hung in there like a fool. 42 00:02:59,799 --> 00:03:01,379 hung in there and guess what? 43 00:03:01,379 --> 00:03:03,902 In 2022, it became real. 44 00:03:03,902 --> 00:03:04,502 Yeah. 45 00:03:04,502 --> 00:03:09,482 And well, and you were, I've talked about this on a previous episode with you, you were early. 46 00:03:09,562 --> 00:03:23,351 So the, really the birth of generative AI as I know it was, you know, the attention is all you need paper from Google, which I think was 2017, right? 47 00:03:23,351 --> 00:03:30,591 Around that time, that time, we were just having things because I'm just reading, you know, book about Empire of AI, great book. 48 00:03:31,071 --> 00:03:33,591 Some bits that aren't so great, but broadly. 49 00:03:35,111 --> 00:03:39,271 Let's not get into the whole looking for victims kind of thing. 50 00:03:39,430 --> 00:03:45,051 But it reminds me of that South Park episode about Megan and Harry. 51 00:03:45,871 --> 00:03:49,091 CSI Special Victims Unit, you know. 52 00:03:49,320 --> 00:03:53,130 There's a lot of sour grapes and whining and that, but yeah. 53 00:03:53,130 --> 00:03:53,661 Yeah. 54 00:03:53,661 --> 00:03:54,071 right. 55 00:03:54,071 --> 00:03:54,891 We'll edit that out. 56 00:03:54,891 --> 00:03:56,143 This will be censored. 57 00:03:56,143 --> 00:04:03,291 This will be sent to the George Orwell School of video editing to be sanitized before publication. 58 00:04:03,291 --> 00:04:09,597 But the interesting thing is that Google were one of the, if not the one, that really helped to pioneer the transformer model. 59 00:04:09,597 --> 00:04:11,489 And they didn't really take it that far. 60 00:04:11,489 --> 00:04:17,352 And then OpenAI, Altman and the team around him and Musk as well. 61 00:04:17,352 --> 00:04:20,132 deserves a lot of credit for backing them early on. 62 00:04:20,132 --> 00:04:21,432 They just picked it up and wrapped them up. 63 00:04:21,432 --> 00:04:25,552 This is one of the cool things, right, about him I saw in the book, which is really fascinating to me. 64 00:04:25,552 --> 00:04:27,232 They started off doing everything. 65 00:04:27,232 --> 00:04:28,732 They were doing the video. 66 00:04:28,732 --> 00:04:31,023 They also were doing robotics, right? 67 00:04:31,023 --> 00:04:37,863 They had a robotic hand because they were focused on AGI as they conceptualized it back in sort of 2016, 2017. 68 00:04:38,203 --> 00:04:42,743 They didn't know, they just had this idea there was going to be this super intelligent AI. 69 00:04:42,903 --> 00:04:43,893 right, human-like AI. 70 00:04:43,893 --> 00:04:46,644 They weren't quite sure how it was going to emerge. 71 00:04:46,644 --> 00:04:47,775 And they were looking at everything. 72 00:04:47,775 --> 00:04:51,386 They were looking at, like, sort of visual, video kind of stuff. 73 00:04:51,386 --> 00:04:53,166 They were looking at the physical realm. 74 00:04:53,166 --> 00:04:55,567 And they were also looking at language models. 75 00:04:55,707 --> 00:04:58,458 But those were all seen as, equally interesting. 76 00:04:58,458 --> 00:05:00,118 They didn't quite know which way it was going to go. 77 00:05:00,118 --> 00:05:05,810 And eventually, of course, once the LLM bit took off, they closed down the robotics bit or got rid of it. 78 00:05:05,810 --> 00:05:08,952 But now, interestingly, they've restarted it. 79 00:05:08,952 --> 00:05:10,392 going off on a tangent. 80 00:05:10,392 --> 00:05:15,832 I think the point I think that I mean, and the funny thing is all of that was going on unbeknownst to me. 81 00:05:15,832 --> 00:05:24,492 Well, I started running around trying to talk to people like Kira, know, no Waysburg and the Ebrevier and companies like that, talking to them about natural language processing, 82 00:05:24,492 --> 00:05:25,663 machine learning, la la la. 83 00:05:25,663 --> 00:05:36,634 And in an office somewhere in San Francisco, Palo Alto, were a bunch of people building the future with very, very little fanfare, very, very little fanfare. 84 00:05:36,634 --> 00:05:37,144 Yeah. 85 00:05:37,144 --> 00:05:37,564 Yeah. 86 00:05:37,564 --> 00:05:46,299 And, but, but you, came to the game, earlier than most and you know, I, paid attention to, and my listeners have, have heard me talk about this before. 87 00:05:46,299 --> 00:05:53,424 It was, when deep blue on Jeopardy was kind of my first introduction, right? 88 00:05:53,491 --> 00:05:54,565 It was IBM. 89 00:05:54,565 --> 00:05:54,946 Yep. 90 00:05:54,946 --> 00:05:59,650 And then alpha go, which was again, that was roughly the same time period. 91 00:05:59,650 --> 00:06:01,693 2017, 2016. 92 00:06:01,693 --> 00:06:02,025 wasn't it? 93 00:06:02,025 --> 00:06:04,865 Wasn't that, that was out of, yeah. 94 00:06:04,865 --> 00:06:10,539 deep mind, Dennis Hesabas that Google ended up buying deep mind. 95 00:06:10,539 --> 00:06:16,446 So Google has had ambitions and has made investments in AI for, for a very long time. 96 00:06:16,446 --> 00:06:20,129 And you know, their first model, couple of models really sucked. 97 00:06:20,129 --> 00:06:22,731 I mean, Bard was pretty bad. 98 00:06:22,731 --> 00:06:24,993 Um, they've really closed the gap. 99 00:06:24,993 --> 00:06:27,755 Gemini 2.5 is quite good. 100 00:06:27,755 --> 00:06:30,296 And I think they're on the verge of releasing. 101 00:06:30,318 --> 00:06:31,831 whatever that next version is. 102 00:06:31,831 --> 00:06:35,747 yeah, Google, Google was kind of slowly then suddenly. 103 00:06:36,779 --> 00:06:37,645 but they've caught up. 104 00:06:37,645 --> 00:06:38,525 thing, isn't it? 105 00:06:38,525 --> 00:06:41,825 You get these like plateau, kind of like watershed moments. 106 00:06:41,825 --> 00:06:49,445 But I mean, the interesting thing here is that the basic, you might say economic theory, the sort of techno economic theory, whatever you want to call it. 107 00:06:49,725 --> 00:06:59,205 But I sort of like, I would say I developed, but I certainly spent a ridiculously large amount of time working on it, you know, around, you know, moving away from the, I mean, 108 00:06:59,205 --> 00:07:02,661 all the key concepts were all there, well developed in the market. 109 00:07:02,661 --> 00:07:11,541 You might say that the only thing really artificial I did was just kind of like tie a little loose ends together and basically build almost like a worldview. 110 00:07:11,541 --> 00:07:13,521 This is the way it's going to be people, right? 111 00:07:13,521 --> 00:07:14,881 This is the way things are going to evolve. 112 00:07:14,881 --> 00:07:17,332 Well, they have to evolve if AI is going to be anything relevant. 113 00:07:17,332 --> 00:07:21,852 And it all came together like in a few months of 2016, right? 114 00:07:21,852 --> 00:07:23,812 Because it's not rocket science anyway, is it? 115 00:07:24,192 --> 00:07:29,808 And then the technology has moved on, but the basic ideas remain the same. 116 00:07:29,808 --> 00:07:33,611 It's rather like the green movement, know, batteries have improved. 117 00:07:33,770 --> 00:07:39,037 Now we have self-driving cars, which without humans are both safer and will be more efficient. 118 00:07:39,037 --> 00:07:40,759 Electric vehicles. 119 00:07:40,759 --> 00:07:51,488 Today they announced, I'm not sure which body, one of the large global energy bodies announced that this year was the first time ever that, well, actually you're going to have 120 00:07:51,488 --> 00:07:53,730 to edit this bit because I can't remember the facts. 121 00:07:53,730 --> 00:07:55,161 It's the... 122 00:07:55,282 --> 00:08:01,358 I think it was the first time that renewable energy overtook fossil fuels globally on global basis, something like that. 123 00:08:01,358 --> 00:08:01,999 Yeah. 124 00:08:01,999 --> 00:08:10,715 But the point is, that the tech moves forward, but the people who pioneered green energy had thought all of this through 20 years ago. 125 00:08:10,715 --> 00:08:11,896 Right. 126 00:08:11,896 --> 00:08:13,247 like, okay, so we're going to need wind. 127 00:08:13,247 --> 00:08:15,820 And then someone says, yes, but wind is very variable. 128 00:08:15,820 --> 00:08:19,233 So you're to need better battery technology. 129 00:08:19,233 --> 00:08:20,093 People go, yep, yep. 130 00:08:20,093 --> 00:08:20,664 Okay, right. 131 00:08:20,664 --> 00:08:21,784 Let's develop that. 132 00:08:21,784 --> 00:08:22,184 Right? 133 00:08:22,184 --> 00:08:27,264 And we're going to need a mix and we're going to have, you know, we're going to need better transmission systems and so forth. 134 00:08:27,264 --> 00:08:30,304 And we're going to massively have to improve battery technology. 135 00:08:30,644 --> 00:08:31,244 So on and so on. 136 00:08:31,244 --> 00:08:32,635 then we're to have to invest in some. 137 00:08:32,635 --> 00:08:42,495 And it's like the groundwork was laid out 20 years ago and we see this again and again in society that the, the, basic facts are pretty obvious for all to see. 138 00:08:42,515 --> 00:08:43,355 Right? 139 00:08:43,355 --> 00:08:46,986 It's just, they don't have the technology or the investment to make it real. 140 00:08:46,986 --> 00:08:49,657 Um, and then eventually the technology comes along. 141 00:08:49,657 --> 00:08:57,090 You know, it goes back to the, know, the old anecdotes about, you know, you know, revolutions needing the right conditions, right? 142 00:08:57,490 --> 00:09:02,660 You know, there's always revolutionaries in any society, any point in history, you all way back to ancient Greece. 143 00:09:02,660 --> 00:09:07,373 There's always a bunch of revolutionaries running around in their togas, you know, going, right. 144 00:09:07,373 --> 00:09:09,864 But it only takes off if you get the right conditions. 145 00:09:09,864 --> 00:09:11,121 And it's the same with technology. 146 00:09:11,121 --> 00:09:18,058 I mean, like I could have shouted until I was blue in the face until I keeled over about getting rid of the billable hour. 147 00:09:18,058 --> 00:09:27,573 by automating work streams, about rethinking the business model of law firms, about thinking about the end outputs, how that benefits society as a whole, all that kind of 148 00:09:27,573 --> 00:09:28,064 stuff, 149 00:09:28,064 --> 00:09:29,387 And we may never have got there. 150 00:09:29,387 --> 00:09:38,651 NLP would have just fizzled out, plateaued out, and you and I would still be, I mean, I don't know what I would be doing, but I'd still be doing legal tech. 151 00:09:38,651 --> 00:09:39,393 I'm not sure. 152 00:09:39,393 --> 00:09:41,653 Well, you know, this is it. 153 00:09:41,653 --> 00:09:46,253 I had a similar experience, similar, similar dynamic with cloud. 154 00:09:46,253 --> 00:09:57,833 So I was very early to pushing cloud in legal when I really didn't understand just how much friction towards change there was. 155 00:09:57,833 --> 00:10:00,033 So this is the early 2010s. 156 00:10:00,033 --> 00:10:08,865 And when something called wave 14 came through for office 365 and made, made it usable, like pre wave 157 00:10:08,865 --> 00:10:17,205 14 or maybe it was wave 15 office 365 now called Microsoft 365 was, was, was not great. 158 00:10:17,205 --> 00:10:24,185 Um, but I really thought to myself, wow, you know, I know what junk Microsoft junkies law firms are. 159 00:10:24,185 --> 00:10:25,345 This makes perfect sense. 160 00:10:25,345 --> 00:10:32,085 I know how little they like to invest in technology and uh, which has historically been the case. 161 00:10:32,085 --> 00:10:35,445 Um, moving to the cloud improves economics. 162 00:10:35,445 --> 00:10:38,025 It improves the security posture. 163 00:10:38,025 --> 00:10:38,978 Um, 164 00:10:38,978 --> 00:10:42,618 It certainly can scale up much more efficiently. 165 00:10:43,037 --> 00:10:47,498 And man, that didn't happen for about another eight years. 166 00:10:47,498 --> 00:10:49,138 It wasn't in the US. 167 00:10:49,138 --> 00:10:54,409 So law firms in the US really did not start to move in earnest to the cloud until a little bit after COVID. 168 00:10:54,409 --> 00:11:00,569 we, again, I beat my head against the wall, kind of like you did talking about AI. 169 00:11:00,769 --> 00:11:05,709 did 22 road shows in 2014 talking about the cloud. 170 00:11:05,765 --> 00:11:12,824 22 different cities and everybody's like, yeah, this is great and did absolutely nothing for another eight years. 171 00:11:12,986 --> 00:11:13,486 So. 172 00:11:13,486 --> 00:11:18,639 in some ways the AI situation was even more crazy in that cloud clearly did work. 173 00:11:18,639 --> 00:11:21,581 You could demonstrate it, but they still didn't want it. 174 00:11:21,581 --> 00:11:27,184 Well, it's not going to that maybe cloud is actually more crazy, but I suppose to some degree you could actually not blame the lawyers much because they were looking at the 175 00:11:27,184 --> 00:11:31,716 early NLP tools and going, man, this looks kind of complicated. 176 00:11:31,716 --> 00:11:33,608 I'm not sure I want to use this stuff. 177 00:11:33,608 --> 00:11:39,972 And then you've got all the economic macroeconomic, cultural, professional aspects that kind of stopped it. 178 00:11:39,972 --> 00:11:42,653 But yeah, I mean, it's 179 00:11:42,745 --> 00:11:43,866 I mean, this the thing, isn't it? 180 00:11:43,866 --> 00:11:52,333 Great, I mean, you cannot kill a great idea, but you can delay it for an enormous, enormous amount of time, right? 181 00:11:52,333 --> 00:11:59,659 I mean, and you see this again, you see this in politics, see this playing out over all kinds of different trains in our society. 182 00:11:59,659 --> 00:12:01,201 You cannot kill a great idea. 183 00:12:01,201 --> 00:12:04,904 People just instinctively look at it, they compare it to their own experiences. 184 00:12:04,904 --> 00:12:07,626 Maybe it's even like just like, know, it's gut instinct, right? 185 00:12:07,626 --> 00:12:08,277 It's intuition. 186 00:12:08,277 --> 00:12:10,625 They just go, that's right, that is right. 187 00:12:10,625 --> 00:12:12,785 And they're like, okay, let's have it. 188 00:12:13,125 --> 00:12:14,125 And then it doesn't happen. 189 00:12:14,125 --> 00:12:15,545 And then it doesn't happen. 190 00:12:15,545 --> 00:12:17,685 Although someone tries it, it doesn't really work very well. 191 00:12:17,685 --> 00:12:24,725 And then all the critics and the cynics start jumping on and going, ah, ha, ha, you were a fool to believe there is hope, young Skywalker. 192 00:12:24,785 --> 00:12:28,925 ha, ha, you know, come to the dark side and enjoy the empire. 193 00:12:28,925 --> 00:12:31,185 It's much nicer over here. 194 00:12:31,365 --> 00:12:34,025 We get much better food, et cetera, et cetera. 195 00:12:34,025 --> 00:12:34,876 office is lovely. 196 00:12:34,876 --> 00:12:39,062 And people give up or they just get disbanded and make off and. 197 00:12:39,062 --> 00:12:41,875 focus something that does work, right? 198 00:12:41,875 --> 00:12:45,377 Like, I don't know, Bitcoin. 199 00:12:45,495 --> 00:12:46,566 That's I was going to say. 200 00:12:46,566 --> 00:12:53,349 Digital currency is another great example of an idea that's been lingering now for almost 20 years. 201 00:12:53,349 --> 00:12:54,029 It was 2008-ish. 202 00:12:54,029 --> 00:12:55,368 But it makes it makes people money. 203 00:12:55,368 --> 00:12:56,020 mean, that's the thing. 204 00:12:56,020 --> 00:13:01,475 mean, I mean, you know, you can step away from the whole intellectual stuff if you want to and just say, look, does it make money? 205 00:13:01,816 --> 00:13:03,297 If it does make money? 206 00:13:05,659 --> 00:13:06,579 Does. 207 00:13:06,940 --> 00:13:08,894 Well, this is weird thing with crypto. 208 00:13:08,894 --> 00:13:17,349 mean, unless we have a collapse of all fiat currencies and we end up in a sort of Mad Max world, crypto currencies actually have no real functional value. 209 00:13:17,349 --> 00:13:18,737 They are simply 210 00:13:18,737 --> 00:13:26,284 gigantic momentum trades and the momentum is constantly up, of flips around a bit and then keeps going up again. 211 00:13:26,284 --> 00:13:28,008 But fundamentally, there's nothing there. 212 00:13:28,008 --> 00:13:29,353 There is literally nothing there. 213 00:13:29,353 --> 00:13:30,798 argue the same about gold. 214 00:13:30,798 --> 00:13:34,666 Gold's value is driven primarily by scarcity. 215 00:13:34,944 --> 00:13:37,686 True, But it has a tangible value. 216 00:13:37,686 --> 00:13:42,800 mean, you can pull out a lump of gold and people go, gold. 217 00:13:42,881 --> 00:13:45,464 People will always want gold. 218 00:13:45,464 --> 00:13:51,919 The same way that people will always want kebabs or burgers or, you know, it's just one of those things, right? 219 00:13:51,919 --> 00:13:52,671 People like it. 220 00:13:52,671 --> 00:13:59,848 And so, whereas cryptocurrency really is just a momentum trade, maybe that momentum will keep going forever and ever. 221 00:14:00,184 --> 00:14:06,597 Right, rather like a YouTube channel that just keeps on accumulating views and followers. 222 00:14:06,757 --> 00:14:07,758 It'll never, never stop. 223 00:14:07,758 --> 00:14:10,199 It just goes up and up and up and up. 224 00:14:10,199 --> 00:14:17,802 But if it ever did flip, like really flip, not just fluctuate, but completely flip and go reverse, there's nothing that you can't cash it in at the end. 225 00:14:18,174 --> 00:14:19,543 You what have I got left? 226 00:14:19,543 --> 00:14:20,935 I've got three electrons. 227 00:14:20,935 --> 00:14:22,160 Are they worth anything? 228 00:14:22,160 --> 00:14:24,396 It's like, no, I'm sorry, your electrons are not really worth anything. 229 00:14:24,396 --> 00:14:26,178 you've got a collection of prime numbers. 230 00:14:26,178 --> 00:14:35,747 um But the underlying technology of blockchain provides a tremendous opportunity and utility. 231 00:14:35,747 --> 00:14:37,468 And there's been a lot of talk in legal. 232 00:14:37,468 --> 00:14:39,490 It really hasn't gotten anywhere. 233 00:14:39,490 --> 00:14:47,014 the asynchronous ledger, distributed ledger concept is an interesting one in recording. 234 00:14:47,014 --> 00:14:47,674 is, is. 235 00:14:47,674 --> 00:14:50,085 mean, you know, David Fisher, know, integral ledger. 236 00:14:50,085 --> 00:14:52,997 He's been pushing that for ever since I got started. 237 00:14:52,997 --> 00:15:00,020 And, he still has a lot of people who really believe in it, you know, but the problem is, is it's really be useful. 238 00:15:00,020 --> 00:15:05,483 My personal view is you need a scenario where there's no trust, where trust is completely broken down. 239 00:15:05,483 --> 00:15:08,835 And most commercial lawyers don't operate in such a system. 240 00:15:08,835 --> 00:15:11,956 Probably this reason is because if trust had completely. 241 00:15:11,956 --> 00:15:19,960 utterly irrevocably broken down and no one can trust each other and nothing could be real, it'd be quite hard to be a lawyer because you've lost a rules-based system that underpins 242 00:15:19,960 --> 00:15:20,930 all of that. 243 00:15:21,711 --> 00:15:23,191 You're in Mad Max. 244 00:15:23,211 --> 00:15:30,294 And at that point, cryptocurrency and blockchain and everything really makes absolute sense because you cannot trust anybody to do the right thing. 245 00:15:30,486 --> 00:15:31,366 Yeah. 246 00:15:31,507 --> 00:15:42,519 Yeah, that's a, we could go down a rabbit hole with the cryptocurrency, but, let's talk about your, so you've got a couple of events coming up that, I wanted to make people aware 247 00:15:42,519 --> 00:15:48,815 of because your theme aligns really well with the theme of this podcast, which is legal innovation. 248 00:15:48,815 --> 00:15:53,189 So tell us about the, I think you have events coming up in New York and London. 249 00:15:53,189 --> 00:15:54,960 Tell us a little bit about those. 250 00:15:55,078 --> 00:16:00,580 Yeah, so just briefly, so Legal Innovators was a conference that grew out of Artificial Lawyer. 251 00:16:00,580 --> 00:16:06,044 It's organized by another company called Cosmonauts, but I kind of co-created it with them and I chair the events. 252 00:16:06,044 --> 00:16:07,982 London, we've been doing for years and years. 253 00:16:07,982 --> 00:16:09,515 It's quite a big conference now. 254 00:16:09,515 --> 00:16:12,868 We're hoping to get around about thousand people over three days. 255 00:16:12,868 --> 00:16:17,641 Law firm day, in-house day, litigation day, which is gonna be fun, first time we've done that. 256 00:16:17,641 --> 00:16:21,062 So that's the fourth, fifth and sixth of November in London. 257 00:16:21,184 --> 00:16:23,846 And then we're to be doing New York for very first time. 258 00:16:23,846 --> 00:16:28,100 We've been doing California for years and years in San Francisco, which is brilliant fun. 259 00:16:28,100 --> 00:16:28,427 love it. 260 00:16:28,427 --> 00:16:31,894 It's probably my favorite event, but we're doing New York and it's going be very interesting. 261 00:16:31,894 --> 00:16:42,883 Now, the reason why I delayed doing New York for so long is because I didn't feel that the big firms there who really rule, you know, the kingdom were really, really getting into AI 262 00:16:42,883 --> 00:16:49,290 and automation and rethinking business sufficiently to really make it an exciting event. 263 00:16:49,290 --> 00:16:51,731 I mean, Cleary was an outlier early on. 264 00:16:51,731 --> 00:16:54,833 They created Cleary X, did a lot of interesting work. 265 00:16:54,833 --> 00:16:57,073 But now things are changing. 266 00:16:57,073 --> 00:16:57,544 Definitely. 267 00:16:57,544 --> 00:17:04,067 I mean, when I was in New York for Legal Week earlier this year, I spoke to a lot of people that I got the feeling that things are shifting. 268 00:17:04,067 --> 00:17:07,268 You know, and there was some really good stuff going on at Gunderson as well, for example. 269 00:17:07,268 --> 00:17:12,360 ah It is, is, but in their New York office. 270 00:17:12,360 --> 00:17:13,170 it's true. 271 00:17:13,170 --> 00:17:15,980 But let's face it, Leif Mawatkins is a West Coast firm as well. 272 00:17:15,980 --> 00:17:17,604 You know, they're doing some interesting things. 273 00:17:17,604 --> 00:17:21,344 So it's kind of like, I think it's come of age finally. 274 00:17:21,644 --> 00:17:32,144 kind of, I mean, you expect Wilson, Sonsini and Cooley and others to have created startups and have really huge innovation teams, doing all kinds of fun stuff, working very closely 275 00:17:32,144 --> 00:17:33,484 with AI companies. 276 00:17:33,804 --> 00:17:39,584 New York firms, I think were slightly, they may have brought in a lot of people around knowledge management. 277 00:17:39,584 --> 00:17:41,044 They certainly did. 278 00:17:41,224 --> 00:17:44,884 But this idea that, hey, AI is actually gonna change the game for us. 279 00:17:44,884 --> 00:17:46,980 You know, just a little. 280 00:17:46,980 --> 00:17:48,000 Just a little. 281 00:17:48,161 --> 00:17:54,997 I think that was, even though people may have talked a good talk, I didn't really feel that it was actually happening until very recently. 282 00:17:54,997 --> 00:17:58,412 And I think it is actually starting to happen though. 283 00:17:58,412 --> 00:17:58,842 So. 284 00:17:58,842 --> 00:17:59,482 it is. 285 00:17:59,482 --> 00:18:04,415 And I think a leading indicator to that you alluded to, which is some of the hiring. 286 00:18:04,415 --> 00:18:05,916 mean, look at Simpson Thatcher. 287 00:18:05,916 --> 00:18:09,388 I mean, that's the epitome of a New York white shoe law firm. 288 00:18:09,388 --> 00:18:22,315 You know, they brought in Liz Grennan from McKinsey and they've made some other hires that again, that's how from the outside, if they're a client, I have got an inside view, but 289 00:18:22,315 --> 00:18:23,936 for firms who 290 00:18:24,289 --> 00:18:26,071 We don't do business with an info dash. 291 00:18:26,071 --> 00:18:30,694 look at their org chart and say, how are they, what moves are they making there? 292 00:18:30,694 --> 00:18:39,762 And you know, somebody like Liz isn't going to come from McKinsey without having a high level of commitment. 293 00:18:39,762 --> 00:18:46,988 Cause she's also an attorney and Connor Grennan's wife, who, you know, is a big thought leader in the AI space. 294 00:18:46,988 --> 00:18:52,192 think he's a NYU law school or I'm sorry, not as he'd law school. 295 00:18:52,676 --> 00:18:54,317 or maybe business school. 296 00:18:54,458 --> 00:19:04,788 So, you know, I see people who I know they're not going to move unless the firm has demonstrated a commitment and articulated some sort of a plan to move in a certain 297 00:19:04,788 --> 00:19:05,889 direction. 298 00:19:05,889 --> 00:19:10,815 And that tells me again, that's a leading indicator that they're starting to move in that direction. 299 00:19:10,815 --> 00:19:11,976 uh 300 00:19:11,976 --> 00:19:14,396 yeah, we shouldn't get carried away. 301 00:19:14,396 --> 00:19:23,747 I mean, while organizations like Salesforce out in California literally have their own team that builds AI products for their own use, right? 302 00:19:23,747 --> 00:19:31,567 You know, and some of the banks and some of the private equity funds are getting definitely very interested in what AI can do for them and doing stuff internally as well. 303 00:19:31,567 --> 00:19:35,767 And some of the big consultancies too, have big bases in New York as well. 304 00:19:35,767 --> 00:19:36,987 You know, 305 00:19:37,215 --> 00:19:41,158 Are the big New York firms really disrupting themselves? 306 00:19:41,158 --> 00:19:43,769 I don't think yet, and maybe not for a while. 307 00:19:43,769 --> 00:19:49,203 But as you say, they're genuinely thinking about it now and they're exploring and experimenting. 308 00:19:49,203 --> 00:19:53,546 They're bringing in these gen-ai productivity platforms and a whole bunch of other tools. 309 00:19:53,546 --> 00:19:57,588 It's going much, much more beyond knowledge management. 310 00:19:57,589 --> 00:20:01,431 It's not just like, okay, well make my life a little bit easier. 311 00:20:01,672 --> 00:20:05,455 Help me find the information I need so I can go off and do the deal that I need to do. 312 00:20:05,455 --> 00:20:06,605 It's more... 313 00:20:06,605 --> 00:20:16,459 now, well not more now, but in addition now, they're saying, okay, you know what, this may actually alter our business model at a very small level to begin with. 314 00:20:16,459 --> 00:20:18,100 But that, it's all relative, isn't it? 315 00:20:18,100 --> 00:20:18,881 It's all relative. 316 00:20:18,881 --> 00:20:24,454 know, that, that in itself, the fact that they're even asking that question, I think is a big deal. 317 00:20:24,454 --> 00:20:24,784 Yeah. 318 00:20:24,784 --> 00:20:32,432 Well, and let me ask you something along those lines when we talk about how law firms move in that direction. 319 00:20:32,432 --> 00:20:44,122 And by that direction, mean, you know, making fundamental changes to foundational pieces of their business, how legal service delivery, pricing, internal firm compensation, client 320 00:20:44,122 --> 00:20:45,022 engagement. 321 00:20:45,022 --> 00:20:51,267 These are foundational elements that are very firmly entrenched in law firms. 322 00:20:51,431 --> 00:20:54,733 and they have to change and they're going to change. 323 00:20:54,733 --> 00:21:00,896 Whether law firms like it or not, it's just at what speed and how. 324 00:21:00,896 --> 00:21:04,477 think one of the biggest ones is internal firm compensation. 325 00:21:04,477 --> 00:21:10,060 way lawyers are ruthlessly measured on billable hours today and that is such a deeply entrenched. 326 00:21:10,060 --> 00:21:16,133 I've talked to some leaders of innovation councils at some pretty big firms and they have cited that 327 00:21:16,255 --> 00:21:28,775 particular cultural element of the firm is one of the hardest to move the dial on where, because again, if you're AI really changes the internal firm compensation, when pricing 328 00:21:28,775 --> 00:21:31,917 changes, internal firm compensation changes, right? 329 00:21:31,917 --> 00:21:43,555 If you're in a AFA scenario where the firm is taking on risk, revenue in the door doesn't equate to profitability like it does today. 330 00:21:43,933 --> 00:21:46,891 you have to manage risk and efficiency. 331 00:21:46,984 --> 00:21:48,015 It's an interesting one, isn't it? 332 00:21:48,015 --> 00:21:52,658 Cause it's like, you know, the, I don't think there's anything wrong in using time as an internal metric. 333 00:21:52,658 --> 00:21:57,031 I don't think it's the best metric internally, but if you, if you want to use it, okay, fine. 334 00:21:57,031 --> 00:22:04,577 At least you've got some kind of yardstick, but I think for, know, there's Bob, he's only worked 20 hours this week. 335 00:22:04,577 --> 00:22:08,810 Ooh, you know, there's Jane, she's worked 60 hours or a hundred hours this week. 336 00:22:08,810 --> 00:22:09,251 she's great. 337 00:22:09,251 --> 00:22:10,031 Okay. 338 00:22:10,031 --> 00:22:10,922 Whatever. 339 00:22:11,534 --> 00:22:18,280 That kind of thinking is as old as time, But what I think is an issue is judging value by time. 340 00:22:18,380 --> 00:22:23,073 So, you know, here's Bob and Jane, they spend all weekend going through these documents, aren't they wonderful? 341 00:22:23,073 --> 00:22:26,366 That'll be, you know, $25,000, please. 342 00:22:26,366 --> 00:22:29,487 Oh, no, that AI system did it in three minutes. 343 00:22:29,848 --> 00:22:30,879 Now what are we gonna do? 344 00:22:30,879 --> 00:22:32,583 It's nuts, I personally think. 345 00:22:32,583 --> 00:22:36,918 So yeah, keep the art internally, but externally, you're gonna have to find a better method. 346 00:22:36,918 --> 00:22:38,080 Yeah, indeed. 347 00:22:38,080 --> 00:22:52,148 you know, and what, are your thoughts on, I've had this debate a lot and I would say from the people I've talked to, it's pretty split, maybe 60, 60, 40 in favor of, and 40 % think 348 00:22:52,148 --> 00:22:54,870 that this can't happen, but I'm a believer. 349 00:22:54,870 --> 00:23:04,099 think that law firms are actually going to have to leverage their data in a meaningful way as part of their AI strategy and 350 00:23:04,099 --> 00:23:11,239 and invest in R and D, which law firms historically have not done well or haven't done at all for a couple of reasons. 351 00:23:11,239 --> 00:23:16,539 One, the partnership model optimizes is optimized for profit taking at the end of the year. 352 00:23:16,539 --> 00:23:23,039 It's hard to accrue capital expenditure doesn't fit into that cash basis partnership model. 353 00:23:23,039 --> 00:23:27,459 Well, but also they're, they're not, they're not software companies. 354 00:23:27,459 --> 00:23:29,979 They're not used to building software. 355 00:23:30,239 --> 00:23:32,879 And I don't know, what are you? 356 00:23:32,879 --> 00:23:33,899 Yeah. 357 00:23:34,432 --> 00:23:42,458 think, you know, law firms want to build stuff, go for it, you know, the same way that someone might say to me, well, you don't have to buy that piece of technology to improve 358 00:23:42,458 --> 00:23:43,439 artificial wear. 359 00:23:43,439 --> 00:23:44,539 Why don't you build it yourself? 360 00:23:44,539 --> 00:23:45,454 It's really easy. 361 00:23:45,454 --> 00:23:48,702 You know, it's very, very, you know, pain by numbers. 362 00:23:48,983 --> 00:23:50,344 I might do it. 363 00:23:50,344 --> 00:23:54,738 I'll be honest, I'm probably more minded just to buy it and let someone else deal with it. 364 00:23:54,738 --> 00:23:57,250 I don't know, that for me is not the issue. 365 00:23:57,250 --> 00:24:03,626 It's the issue is can they start to think in a way that AI is the leverage? 366 00:24:03,626 --> 00:24:13,998 And I just want to bring in a law firm, an AI hybrid law firm, which I think is great, which is called Covenant based in New York, created by the ex-general council of WeWork 367 00:24:13,998 --> 00:24:15,443 and her partner. 368 00:24:15,443 --> 00:24:17,065 And it's a great model. 369 00:24:17,065 --> 00:24:17,665 It's a great model. 370 00:24:17,665 --> 00:24:21,408 They have a tech company, which is the AI bit. 371 00:24:21,408 --> 00:24:25,452 And they have a law firm, which has got multiple lawyers in it, not a huge number at the moment, just a small group. 372 00:24:25,452 --> 00:24:28,425 And they do private equity investments. 373 00:24:28,425 --> 00:24:30,887 I understand. 374 00:24:30,887 --> 00:24:31,769 And on the buy side. 375 00:24:31,769 --> 00:24:34,142 And there's a license agreement between the two. 376 00:24:34,142 --> 00:24:39,648 So they don't actually have to go away to Arizona to, you know, to do this. 377 00:24:39,648 --> 00:24:42,519 under special bar rules and it's, it's just great. 378 00:24:42,519 --> 00:24:51,171 then we were talking, we did this great video together and Jen, and she just came up with this phrase, just came out in the conversation and she said, you know, AI is the leverage. 379 00:24:51,331 --> 00:25:02,154 And I just thought, man, I wish I'd coined that phrase 10 years ago, because for me, it sums up everything that artificial law has been trying to say ever since I got started, 380 00:25:02,154 --> 00:25:05,435 that AI is the means of production. 381 00:25:05,635 --> 00:25:05,885 Right. 382 00:25:05,885 --> 00:25:08,636 Now there are some things which aren't sort of productivity based. 383 00:25:08,636 --> 00:25:16,396 You know, so, know, the classic cliche, know, it's Martin Lipton, you know, sitting on the sofa thinking about the poison pill and he has a eureka moment. 384 00:25:16,396 --> 00:25:18,356 goes, aha, I've got it. 385 00:25:18,356 --> 00:25:18,576 All right. 386 00:25:18,576 --> 00:25:21,816 That's not a productivity play, but a due diligence report. 387 00:25:21,816 --> 00:25:23,296 That's a productivity play. 388 00:25:23,336 --> 00:25:24,336 A lot of legal research. 389 00:25:24,336 --> 00:25:24,656 Okay. 390 00:25:24,656 --> 00:25:27,956 There's a degree of sophistication around that, but a lot of it is bulk. 391 00:25:28,076 --> 00:25:29,856 E-discovery, that's bulk. 392 00:25:29,856 --> 00:25:30,516 Okay. 393 00:25:30,516 --> 00:25:34,476 Other intelligent insights and knowledge we can make that more sophisticated. 394 00:25:34,476 --> 00:25:34,796 Better. 395 00:25:34,796 --> 00:25:35,656 Yeah, obviously. 396 00:25:35,656 --> 00:25:37,148 But there's a ton of bulk there. 397 00:25:37,148 --> 00:25:39,230 So that's all going down to productivity, right? 398 00:25:39,230 --> 00:25:45,353 Anything that relates to productivity, you throw at leverage, IAV associate body, right? 399 00:25:45,353 --> 00:25:49,515 And the way we need to think is AI is the leverage. 400 00:25:49,595 --> 00:25:56,789 And I think for me, that's a tipping point once law firms, the big established law firms start to look at Covenant and go, you know what? 401 00:25:56,789 --> 00:25:58,900 That's actually the best model we've ever seen. 402 00:25:59,221 --> 00:26:05,886 And the beautiful thing about it is they don't even have to get into complicated bar rules, right? 403 00:26:05,886 --> 00:26:11,159 They don't even have to do that because if they structure it in the right way, they can do this anyway. 404 00:26:11,159 --> 00:26:18,702 And besides, and even if they don't want to do that fancy, fancy structure, they can just buy in the technology as a license holder. 405 00:26:18,863 --> 00:26:26,206 I mean, it's not these, again, it goes back to this earlier conversation about the great idea is already here. 406 00:26:26,206 --> 00:26:27,798 The question is, will they do it? 407 00:26:27,798 --> 00:26:28,329 Right. 408 00:26:28,329 --> 00:26:29,289 Yeah. 409 00:26:29,289 --> 00:26:36,509 Did you hear, I'm assuming you read the story about Burford capital and the MSO model there. 410 00:26:36,509 --> 00:26:37,729 It's a private equity firm. 411 00:26:37,729 --> 00:26:39,609 was in the financial times. 412 00:26:39,727 --> 00:26:46,599 I mean, litigation financial financiers are doing pretty interesting stuff. 413 00:26:46,599 --> 00:26:47,780 Yeah. 414 00:26:47,780 --> 00:26:59,755 They essentially stand up these managed services organizations and deploy private capital into those, which really kind of take over some business of law functions. 415 00:26:59,755 --> 00:27:01,165 And it's a vehicle. 416 00:27:01,165 --> 00:27:05,758 It's an investment vehicle for private capital to get deployed in the US legal market. 417 00:27:05,758 --> 00:27:10,000 And it feels kind of a little bit like a back door at the moment. 418 00:27:10,000 --> 00:27:12,521 If things were 419 00:27:12,927 --> 00:27:20,743 If we had the ABS roles like we do in Arizona across the country, I'm not so sure that would be the ideal structure. 420 00:27:20,764 --> 00:27:26,348 But one of the things I wanted to get your take on is I talk about this a lot. 421 00:27:26,588 --> 00:27:32,073 Law firms today, I spent 10 years at Bank of America in mostly risk management roles. 422 00:27:32,073 --> 00:27:41,298 So I was in consumer risk, then I moved into global treasury, then I went to corporate audit, then I was in anti-money laundering and then compliance. 423 00:27:41,298 --> 00:27:51,203 really interesting work, but I got a really good sense for just how much rigor there is around risk management and financial services for obvious reasons. 424 00:27:51,544 --> 00:27:54,645 And looking at law firms, they have none. 425 00:27:54,965 --> 00:28:00,297 There's literally nothing between the lawyer and his or her client, right? 426 00:28:00,297 --> 00:28:05,750 It's I make requests, lawyer delivers work product or advice. 427 00:28:05,750 --> 00:28:06,870 There's no 428 00:28:07,018 --> 00:28:17,229 layers and when we get into a scenario like where we start to talk about automation and legal, you know, tech enabled legal service delivery, we have to start thinking about risk 429 00:28:17,229 --> 00:28:18,230 management. 430 00:28:18,230 --> 00:28:20,293 And what does that look like in a law firm? 431 00:28:20,293 --> 00:28:25,197 It would be lawyers listening to some risk manager saying, you can't do that. 432 00:28:25,298 --> 00:28:27,661 That's not going to go over very well, but I don't know. 433 00:28:27,661 --> 00:28:28,612 How do you think about 434 00:28:28,612 --> 00:28:29,313 two different things. 435 00:28:29,313 --> 00:28:36,418 think, I think one you sort of indirectly touched on absolutely key point, which is the lawyers are the means of production, right? 436 00:28:36,418 --> 00:28:38,030 They are the center of the universe, right? 437 00:28:38,030 --> 00:28:43,966 We have, we do not live in the Copernican, Copernican, Copernian system, right? 438 00:28:43,966 --> 00:28:46,348 know, lawyers are the center of the universe. 439 00:28:46,348 --> 00:28:51,372 They do not, they do not revolve around process and product and so forth. 440 00:28:51,372 --> 00:28:51,792 Right? 441 00:28:51,792 --> 00:28:53,053 The lawyers, 442 00:28:53,413 --> 00:28:54,514 control everything. 443 00:28:54,514 --> 00:28:56,315 They do the final sign off. 444 00:28:56,315 --> 00:28:58,278 They do the interaction with the clients. 445 00:28:58,278 --> 00:28:59,898 They are the workflow. 446 00:28:59,979 --> 00:29:02,702 They have a mind, they have a hand, they have a whole machine. 447 00:29:02,702 --> 00:29:07,529 And it's only when we start to think about automation in the true sense. 448 00:29:07,529 --> 00:29:09,551 So you don't do that anymore. 449 00:29:09,731 --> 00:29:11,614 The machine does that. 450 00:29:11,614 --> 00:29:13,875 Talking about agentic flows or anything else. 451 00:29:14,195 --> 00:29:15,375 That's when things change. 452 00:29:15,375 --> 00:29:17,675 And I think that's when it comes to your point about risk. 453 00:29:17,675 --> 00:29:25,195 I mean, as it currently stands, it doesn't make much difference because anything that as far as I understand, anything that goes out of the door of a law firm has to be signed off 454 00:29:25,195 --> 00:29:26,546 and is under their umbrella. 455 00:29:26,546 --> 00:29:32,937 What'll be interesting is as we start to build truly automated workflows, from start to finish. 456 00:29:32,937 --> 00:29:37,917 Now it may be a very, very narrow workflow, but they will grow. 457 00:29:37,917 --> 00:29:38,927 They will grow. 458 00:29:38,927 --> 00:29:41,979 It's going to get more and more powerful. 459 00:29:41,979 --> 00:29:44,220 That's when I think the whole risk and insurance thing comes in. 460 00:29:44,220 --> 00:29:47,762 But even so, could argue that law firms still have it under their umbrella. 461 00:29:47,762 --> 00:29:59,226 And it will be down to the law firm or any consultants they can bring in to do pen testing effectively, to make sure that it works completely fine. 462 00:29:59,226 --> 00:30:03,647 But yeah, for me, this has always been the battle in the... 463 00:30:03,829 --> 00:30:06,581 And it's totally understandable, because I probably would do the same. 464 00:30:06,581 --> 00:30:13,248 Most lawyers, most professionals, they see technology and they go, great, how can that add to what I do already? 465 00:30:13,348 --> 00:30:18,292 How can that finesse or take a little bit of a bother out of my life? 466 00:30:18,412 --> 00:30:22,116 All right, they're the center of the universe, right? 467 00:30:22,116 --> 00:30:27,380 If that is all we do with AI now, then nothing's going to change at all. 468 00:30:27,380 --> 00:30:32,564 It goes back to, don't know if we'll probably have this point before, it becomes the Ikea catalog situation. 469 00:30:32,588 --> 00:30:38,451 where you get your various shelving units and cushions and rugs and throwers and all of this kind of stuff. 470 00:30:38,451 --> 00:30:43,514 And it's very pretty and it's very nice and it greatly increases the comfort of that person. 471 00:30:43,514 --> 00:30:44,184 And why not? 472 00:30:44,184 --> 00:30:47,292 People like to be comfortable, but it doesn't change fundamentally anything. 473 00:30:47,292 --> 00:30:50,318 You're not Le Corbusier completely redesigning the building. 474 00:30:50,318 --> 00:30:56,181 You don't change your one bedroom flat into a machine for living, as Le Corbusier said. 475 00:30:56,181 --> 00:31:00,803 We're fundamentally still in the same world with some decorations from Ikea, right? 476 00:31:00,803 --> 00:31:02,804 Bought out of the catalog and then installed. 477 00:31:02,841 --> 00:31:07,102 Things only change once you start to automate whole streams. 478 00:31:07,422 --> 00:31:17,548 And I think that's, and I think this is incredibly difficult for professionals, particularly lawyers, to get their heads around because it's just like, yes, you are not 479 00:31:17,548 --> 00:31:20,609 gonna own everything any longer. 480 00:31:21,029 --> 00:31:26,532 You might be able to own the output and make money from it, but you will not own those workflows, right? 481 00:31:26,532 --> 00:31:28,572 And this is not like an assistant. 482 00:31:28,572 --> 00:31:30,467 I was thinking about how to... 483 00:31:30,467 --> 00:31:33,840 explain this in a very simple diagram for a thing I've got to do next week. 484 00:31:33,840 --> 00:31:38,645 So basically think about a partner, a partner has assistants, right? 485 00:31:38,645 --> 00:31:41,466 As they were kind of originally called rather than associates, right? 486 00:31:41,466 --> 00:31:46,720 That the assistant assists the partner in the things that the partner wants to achieve, right? 487 00:31:46,720 --> 00:31:50,211 So the partner says, right, I need you to go and look at that law book, go. 488 00:31:50,211 --> 00:31:52,612 I need you to prove that contract, go. 489 00:31:52,612 --> 00:31:55,883 You go and pick me up a sandwich, go, right? 490 00:31:55,883 --> 00:31:57,834 They're assistants, right? 491 00:31:57,840 --> 00:32:06,157 They may contribute meaningful aspects to the end product, but fundamentally they're completely under the command of this one person, right? 492 00:32:06,157 --> 00:32:08,019 They are the center of our universe, right? 493 00:32:08,019 --> 00:32:15,136 But in true automation and sort of an agentic flow, there is simply a person at one end of it, because it's all pre-built, right? 494 00:32:15,136 --> 00:32:19,168 Even if it can be uh go through iterations and you can have some feedback. 495 00:32:19,168 --> 00:32:24,583 But ultimately, there's one person at the end of that line and there's another person at the beginning of it. 496 00:32:24,583 --> 00:32:27,043 And that's it, right? 497 00:32:27,263 --> 00:32:31,474 You know, there is no assistant in this thing, right? 498 00:32:31,474 --> 00:32:36,174 It's a machine that makes something that comes out of the other end, to put it really simply. 499 00:32:36,474 --> 00:32:46,274 And I think this is very, very, difficult because it's the clash, and this is something that Susskind wrote about 25 million years ago, about the battle between the artisanal and 500 00:32:46,274 --> 00:32:47,414 the automation. 501 00:32:47,634 --> 00:32:52,642 when I started doing artificial law, I chose, I prefer the term like, 502 00:32:52,642 --> 00:32:59,002 you know, industrial, because I grew up in the Midlands in Britain, which is where parts of the industrial revolution started. 503 00:32:59,002 --> 00:33:05,533 And I've always thought about, you know, the industrial revolution and how machines changed everything. 504 00:33:05,533 --> 00:33:10,193 And for me, it it just seemed like so obvious, like this, this will happen to professional services. 505 00:33:10,193 --> 00:33:17,293 And I think when that, you know, going back to the early beginnings of NLP and machine learning, I think people could see making some small inroads, but they could say like, 506 00:33:17,293 --> 00:33:20,601 this is going to spread across the entire sector. 507 00:33:20,601 --> 00:33:22,641 No way, no way, it's not gonna happen. 508 00:33:22,741 --> 00:33:29,861 Now with large language models and language understanding and agentic flows, which are getting better like by the minute, right? 509 00:33:30,241 --> 00:33:32,441 It actually is doable now. 510 00:33:32,701 --> 00:33:33,861 I think it really is doable. 511 00:33:33,861 --> 00:33:35,241 Do we have accuracy problems? 512 00:33:35,241 --> 00:33:37,021 Yes, accuracy is still a problem. 513 00:33:37,021 --> 00:33:44,492 If you automate an inaccurate system, you just mass produce bad goods, right? 514 00:33:44,492 --> 00:33:47,800 You know, like if you've got like a lathe in your garage, 515 00:33:47,800 --> 00:33:52,380 and you're turning out widgets and you've got something wrong with the, with the tool bit. 516 00:33:52,380 --> 00:33:56,454 All you're going to do is turn out a thousand widgets that you can't use. 517 00:33:56,454 --> 00:34:01,887 yeah, but is that, I mean, okay, look at, look at, look at Waymo, right? 518 00:34:01,887 --> 00:34:06,049 If I, if I stick a Waymo on the streets and it's got something wrong with it, you're going to kill someone. 519 00:34:06,049 --> 00:34:08,140 I mean, literally you can kill someone, right? 520 00:34:08,140 --> 00:34:09,871 It's really serious stuff. 521 00:34:09,871 --> 00:34:12,723 They, they've managed to get on top of it. 522 00:34:12,723 --> 00:34:14,207 And now Waymo. 523 00:34:14,207 --> 00:34:17,232 from a day trial or is actually safer than human drivers. 524 00:34:17,232 --> 00:34:22,224 heard that it reduces side impact collisions by like higher than 95%. 525 00:34:22,224 --> 00:34:32,630 So from, you know, going from a hundred thousand a year to 5,000 a year, you know, because people don't look both ways before they cross an intersection. 526 00:34:32,630 --> 00:34:34,391 You see a green light, you go. 527 00:34:34,391 --> 00:34:39,564 These vehicles have the ability to, you know, they have uh adjacent visibility. 528 00:34:39,564 --> 00:34:39,975 So. 529 00:34:39,975 --> 00:34:45,146 I was in one in San Francisco, know, for the Legal Innovators California, and it was absolutely brilliant. 530 00:34:45,146 --> 00:34:46,506 Really fantastic. 531 00:34:46,506 --> 00:34:49,426 Shout out to Todd Smithline for paying for it. 532 00:34:49,426 --> 00:34:50,357 Thank you for that, Todd. 533 00:34:50,357 --> 00:34:51,277 It was so good. 534 00:34:51,277 --> 00:34:51,857 It was so good. 535 00:34:51,857 --> 00:34:54,828 Honestly, I got out of that and I just went, I'm a believer. 536 00:34:54,828 --> 00:34:58,872 It's funny, mean, it was funny having this, having this debate with a bunch of people. 537 00:34:58,872 --> 00:35:08,732 of this thing last week and they were talking about, know, oh my God, you know, it's so good that, um, you know, you have human pilots on a plane, um, because they can rescue you 538 00:35:08,732 --> 00:35:09,692 if anything happens. 539 00:35:09,692 --> 00:35:12,563 And I was just like, well, why don't you just have an extra AI system? 540 00:35:12,563 --> 00:35:17,863 Have two or two, have three AI systems, one after the other, you know, double fail safe. 541 00:35:18,283 --> 00:35:25,994 I mean, it's just, I mean, it's just, it's just humans hanging on to, um, just 542 00:35:25,994 --> 00:35:27,045 you know, things they got used to. 543 00:35:27,045 --> 00:35:38,165 mean, it's like, okay, so every time you go up in an elevator, right, there's a guy who stands in the elevator with a bunch of cushions in case anything goes wrong, right? 544 00:35:38,165 --> 00:35:41,127 It's just like, you just accept it, you go with it, right? 545 00:35:41,127 --> 00:35:45,854 You know, because after, literally, how long, when did they invent elevators? 546 00:35:45,854 --> 00:35:47,232 100 years ago? 547 00:35:48,394 --> 00:35:53,911 Gotta be house now, when they started building skyscrapers, so since the 1920s, right? 548 00:35:53,911 --> 00:35:55,213 But they worked out the kinks. 549 00:35:55,213 --> 00:35:56,614 I mean, and they've done it. 550 00:35:56,614 --> 00:36:01,136 First thing with Waymo, when it started, everyone said, this is ridiculous. 551 00:36:01,136 --> 00:36:02,036 It won't work. 552 00:36:02,036 --> 00:36:03,627 And then it kept on not working. 553 00:36:03,627 --> 00:36:04,991 And then people said, there you go. 554 00:36:04,991 --> 00:36:05,818 You're a bunch of idiots. 555 00:36:05,818 --> 00:36:07,238 You've wasted loads of money. 556 00:36:07,238 --> 00:36:08,379 Ha ha ha. 557 00:36:08,379 --> 00:36:11,440 And Tesla didn't help because they kept on saying, hey, we've got a self-driving car. 558 00:36:11,440 --> 00:36:13,042 And it obviously wasn't a self-driving car. 559 00:36:13,042 --> 00:36:14,744 People were like, Elon, you're not helping. 560 00:36:14,744 --> 00:36:15,584 And... 561 00:36:16,456 --> 00:36:21,860 Now they have, but he could have maybe toned it down a few years ago before he really actually had one. 562 00:36:21,860 --> 00:36:22,550 But they did it. 563 00:36:22,550 --> 00:36:23,388 They actually did it. 564 00:36:23,388 --> 00:36:27,835 I think that is, I don't understand why everyone like parades down every street in America. 565 00:36:27,835 --> 00:36:32,558 Because that is, that for me is as impressive as putting a man on the moon. 566 00:36:32,558 --> 00:36:33,738 It really is. 567 00:36:33,926 --> 00:36:41,926 You had somebody on your podcast recently that threw out a number that I'd love to bounce off of you. 568 00:36:41,926 --> 00:36:54,166 Speaking of automation, know, when shortly after chat GPT-35 was released, Goldman came out and said, like, I think it was 44 % of legal tasks were subject to automation by 569 00:36:54,166 --> 00:36:55,306 artificial intelligence. 570 00:36:55,306 --> 00:37:02,646 And you had somebody on that was talking about by 2027, 80 % of legal tasks could be automated. 571 00:37:02,646 --> 00:37:03,399 And that 572 00:37:03,399 --> 00:37:05,199 that number seems really high. 573 00:37:05,199 --> 00:37:09,499 Actually the 44 % number still seems really high to me. 574 00:37:09,619 --> 00:37:10,991 In 2027 that 575 00:37:10,991 --> 00:37:16,411 well, I mean, we're going to end up in a sort of like big semantic thing around, you know, what is a task? 576 00:37:16,411 --> 00:37:19,271 When, when does the task begin and end and so forth. 577 00:37:19,311 --> 00:37:30,091 I mean, if people want to listen to it, it's Richard, maybe, uh, who's the CEO of Juro, um, uh, legal tech company focused on in-house and it's on my law punks podcast. 578 00:37:30,091 --> 00:37:38,771 If you just type law punks P U N X not K S P U N X into artificial work, you'll find it. 579 00:37:38,791 --> 00:37:39,639 Um, 580 00:37:40,127 --> 00:37:42,807 Yeah, mean, mean, broadly, I agree. 581 00:37:43,047 --> 00:37:46,627 Will more and more and more tasks be automatable? 582 00:37:46,627 --> 00:37:47,467 Yeah. 583 00:37:47,947 --> 00:37:51,107 Secondly, the issue is will they get automated? 584 00:37:51,107 --> 00:37:57,158 Well, given all the things we've just talked about, probably at the point of an economic gun, right? 585 00:37:57,218 --> 00:37:58,578 So not anytime soon. 586 00:37:58,578 --> 00:37:59,938 Now for in-house, it is different. 587 00:37:59,938 --> 00:38:01,338 Obviously he's focused on in-house. 588 00:38:01,338 --> 00:38:06,738 So he's very much focused on what he's seeing day to day and what he sees. 589 00:38:06,738 --> 00:38:08,090 really does believe. 590 00:38:08,090 --> 00:38:10,010 there can be massive change very rapidly. 591 00:38:10,010 --> 00:38:18,361 And I think in-house probably does have, I'd hope so anyway, less of a barrier, but we've been saying that for years, right? 592 00:38:18,361 --> 00:38:19,501 We've been saying that for years. 593 00:38:19,501 --> 00:38:27,392 In-house teams, they obviously are gonna defeat the billable hour because why would they want it? 594 00:38:27,392 --> 00:38:28,252 But they do. 595 00:38:28,252 --> 00:38:35,264 Many, law firms say, actually, we suggested to blah, blah, General Counselor, large corporate, we'll stop doing billable hours if you want. 596 00:38:35,264 --> 00:38:37,464 I'm a general counsel, so no, no, no, please keep doing it. 597 00:38:37,464 --> 00:38:40,655 Makes it much, much easier for me to keep on track, know, keep track of things. 598 00:38:40,655 --> 00:38:42,534 So I don't know. 599 00:38:42,534 --> 00:38:43,775 I think for me it's for human aspect. 600 00:38:43,775 --> 00:38:45,575 I think technically he is right. 601 00:38:45,575 --> 00:38:47,615 I think technically he's absolutely right. 602 00:38:47,615 --> 00:38:49,035 Could it happen? 603 00:38:49,035 --> 00:38:50,835 Yes, probably. 604 00:38:51,175 --> 00:38:57,775 If we keep going the way we're going, will it happen because of human and economic factors and cultural factors? 605 00:38:57,775 --> 00:38:59,595 I think, I hate to it, probably not. 606 00:38:59,595 --> 00:39:02,575 My guess is, I think I might have said before on this program, 607 00:39:02,575 --> 00:39:06,606 my bet for real true transformation isn't about 12 years. 608 00:39:06,606 --> 00:39:07,966 Now you might say, well, that's terrible. 609 00:39:07,966 --> 00:39:10,666 That's a really long way, but I'm talking like real, real transformation. 610 00:39:10,666 --> 00:39:11,646 So is it worth it? 611 00:39:11,646 --> 00:39:17,026 For example, if I said to you, there'll be no oil and gas in 12 years, we'll have a completely green economy. 612 00:39:17,026 --> 00:39:22,886 There will be no more pollution, at least from, you know, the energy sector in 12 years. 613 00:39:22,886 --> 00:39:24,366 You'd like, well, I don't think that's going to happen. 614 00:39:24,366 --> 00:39:26,046 But if it did, that would be amazing. 615 00:39:26,046 --> 00:39:27,606 Completely transform a global economy. 616 00:39:27,606 --> 00:39:30,146 It would have geopolitical consequences. 617 00:39:30,366 --> 00:39:31,022 Right. 618 00:39:31,114 --> 00:39:36,816 like bye-bye Saudi Arabia, know, they're gonna have to move pretty quickly into solar power. 619 00:39:36,816 --> 00:39:38,497 It would have huge, huge effects. 620 00:39:38,497 --> 00:39:50,883 Now, I honestly do believe that in about 12 years that is gonna happen in professional services, knowledge management service, services, because if AI is where it is now, just 621 00:39:50,883 --> 00:39:59,458 less than three years after the launch of Jack GPT, when OpenAI was the only company seriously investing. 622 00:39:59,548 --> 00:40:05,428 in large language models and where and look now, there's more than you can even poke a stick at, right? 623 00:40:05,448 --> 00:40:07,199 And there's more every day. 624 00:40:07,199 --> 00:40:10,059 I mean, and look at the way it's building. 625 00:40:10,259 --> 00:40:12,659 So like look at lovable, right? 626 00:40:12,659 --> 00:40:18,210 You just type into like a prompt box, build me a job website. 627 00:40:18,210 --> 00:40:20,178 20 minutes later, there it is. 628 00:40:20,178 --> 00:40:21,858 It's incredible. 629 00:40:21,858 --> 00:40:24,598 then, you know, and it's something I remember talking to Zach about ages ago. 630 00:40:24,598 --> 00:40:30,278 He was just saying, you know, like when the internet began, people just use it for very basic search. 631 00:40:30,418 --> 00:40:33,318 And then you've got these like layers of an onion. 632 00:40:33,318 --> 00:40:36,318 They accreted, they built up and built up and built up. 633 00:40:36,318 --> 00:40:41,838 now, you know, well now the internet has become a source for LLMs. 634 00:40:41,838 --> 00:40:49,272 And through, and then through this bizarre sort of combination of data sources and LLMs and agents and all this kind of thing, it's all. 635 00:40:49,272 --> 00:40:57,747 merging together into this huge kind of ecosystem that surrounds us now in a way that we couldn't even have imagined when the internet first appeared. 636 00:40:57,788 --> 00:41:04,912 And on top of all of that will be new understandings and new uses for AI, which we haven't even started to grasp yet. 637 00:41:05,312 --> 00:41:13,469 know, that's the thing, you know, if you think about it, right, lovable could not have existed not in the way that it does today or cursor, for example, could not have existed 638 00:41:13,469 --> 00:41:14,209 the way it does today. 639 00:41:14,209 --> 00:41:18,722 If it hadn't been for work, people like Sam Altman and his group, right. 640 00:41:18,812 --> 00:41:21,104 back in the late teens. 641 00:41:21,104 --> 00:41:23,025 But look how fast things have moved. 642 00:41:23,465 --> 00:41:24,807 Where are we gonna be in five years? 643 00:41:24,807 --> 00:41:25,687 And then 10. 644 00:41:25,687 --> 00:41:35,411 And then you've got, you really, really wanna push the boat out, do you think we'll have some usable quantum computing technology in a decade? 645 00:41:35,572 --> 00:41:37,894 Maybe, not guaranteed. 646 00:41:37,894 --> 00:41:38,534 But you know what mean? 647 00:41:38,534 --> 00:41:45,157 It's like, even without that, even without that extra super powerful compute power, look at the chips they're knocking out now. 648 00:41:45,481 --> 00:41:55,201 Look at the investments, look at the gigantic investments from Cisco into OpenAI and the relationship with NVIDIA now, OpenAI is doing a deal with AMD. 649 00:41:55,661 --> 00:41:59,972 And there's a whole bunch of, there's just like billions upon billions flowing into this. 650 00:41:59,972 --> 00:42:00,543 Right. 651 00:42:00,543 --> 00:42:06,563 So yeah, I mean, it's, it's, I really do believe it's going to be a complete sort of tech ecosystem transformation. 652 00:42:06,743 --> 00:42:10,476 Very, very similar to what we saw from the late nineties onwards with the internet. 653 00:42:10,476 --> 00:42:12,598 And we're seeing, we're seeing examples of that. 654 00:42:12,598 --> 00:42:23,146 You know, the, the Garfields in the UK, the Crosby's in New York, you mentioned covenants, the UD is like, these are, this isn't theory anymore. 655 00:42:23,146 --> 00:42:24,879 This is, this is happening. 656 00:42:24,879 --> 00:42:27,501 And, um, and they're delivering real work. 657 00:42:27,501 --> 00:42:30,113 Like I know a little bit about Crosby. 658 00:42:30,113 --> 00:42:35,648 they had a Sequoia is one of their capital partners and they have a podcast called training data. 659 00:42:35,648 --> 00:42:38,121 And they had the founders of Crosby on. 660 00:42:38,121 --> 00:42:41,792 And the way, what they're doing essentially is like master services agreement. 661 00:42:41,792 --> 00:42:49,157 have a very narrow niche master services agreements for big, you know, like clay and stripe. 662 00:42:49,157 --> 00:42:54,000 And, they are delivering with a very small number of attorneys. 663 00:42:54,020 --> 00:42:56,211 So, I mean, this is, this is happening now. 664 00:42:56,211 --> 00:42:57,811 This isn't theory. 665 00:42:57,811 --> 00:42:58,422 So 666 00:42:58,615 --> 00:43:00,615 But again, it's scale, isn't it? 667 00:43:00,615 --> 00:43:00,995 It's scale. 668 00:43:00,995 --> 00:43:02,215 mean, this is the thing that look at Lovable. 669 00:43:02,215 --> 00:43:07,655 Lovable grew from like, you know, like $1 to tens of millions of dollars in about 12 months. 670 00:43:07,655 --> 00:43:12,455 It was instantly usable and there was no barrier, right? 671 00:43:12,475 --> 00:43:18,586 With legal tech has always suffered from, like I going back to the same point, it's got structural barriers, right? 672 00:43:18,586 --> 00:43:19,866 I Covenant is a great model. 673 00:43:19,866 --> 00:43:20,726 Crosby is a great model. 674 00:43:20,726 --> 00:43:21,797 Garfield is a great model. 675 00:43:21,797 --> 00:43:23,655 Why aren't there 200 of these? 676 00:43:23,655 --> 00:43:26,048 or why doesn't Crosby 677 00:43:30,644 --> 00:43:33,119 mean, they've only been in business a year, right? 678 00:43:33,119 --> 00:43:35,016 So it's, they're, they're brand new. 679 00:43:35,016 --> 00:43:37,257 I suppose it was really interesting calculation. 680 00:43:37,257 --> 00:43:45,197 I'm have to do it at some point, which is, I mean, if you look at the total revenue of all the legal tech companies on the planet, right? 681 00:43:45,217 --> 00:43:54,297 And this is not a zero sum type calculations, which makes it very, very complicated, but start off with what is the total revenue of all the legal tech companies in the world, 682 00:43:54,297 --> 00:43:54,917 right? 683 00:43:54,917 --> 00:43:57,277 And break it down into what they do. 684 00:43:57,477 --> 00:44:03,093 So if we look at sort of more productivity side, so let's put aside all project management, billing. 685 00:44:03,093 --> 00:44:10,156 all that kind of stuff, time tracking, take that out and just look at them all sort of productivity tools and look at the revenue that comes into them. 686 00:44:10,156 --> 00:44:10,636 Right. 687 00:44:10,636 --> 00:44:17,709 Now that is a measure to some degree of how far business of law has integrated these technologies. 688 00:44:17,709 --> 00:44:18,000 Right. 689 00:44:18,000 --> 00:44:22,942 And then once you've done that, and it's very, very complicated to figure that out, maybe McKinsey can do it. 690 00:44:22,942 --> 00:44:28,854 Then look at what a percentage that is of the total legal tech market. 691 00:44:29,296 --> 00:44:33,016 sorry, legal market, sorry, it's very late in the day here. 692 00:44:33,016 --> 00:44:38,656 What percentage is of the total revenue of the whole legal market, right? 693 00:44:38,656 --> 00:44:40,496 Now, these are not exact fractions, right? 694 00:44:40,496 --> 00:44:51,316 You can't say, you know, if someone spends 20 million on X legal tech company over a year, or a bunch of different people spend 20 million, doesn't mean that that has been extracted 695 00:44:51,316 --> 00:44:56,696 from the legal market, because it may actually be empowering the legal market to do more. 696 00:44:56,696 --> 00:44:57,646 The same way, 697 00:44:57,646 --> 00:45:04,437 that law firms spent a fortune on Microsoft Word and email, and that enabled them to make a hell of a lot more money, right? 698 00:45:04,437 --> 00:45:08,299 So I think I'd love to see some financial analysis around that. 699 00:45:08,299 --> 00:45:14,261 So like money that's coming out of a legal market into legal tech that then in turn drives productivity. 700 00:45:14,261 --> 00:45:22,563 And then how much of a sort of force multiplier effect that has on those lawyers. 701 00:45:22,563 --> 00:45:24,823 which enables the market to grow even larger. 702 00:45:24,823 --> 00:45:28,923 So it becomes like this sort of supporting cycle. 703 00:45:28,923 --> 00:45:30,943 And we just don't really have the data. 704 00:45:30,943 --> 00:45:32,583 We just don't have the data. 705 00:45:32,583 --> 00:45:41,294 And there's a generally very, you know, it's classic reductive view is, you if you spend a dollar on an AI tool, you're taking a dollar out of a lawyer's pocket, right? 706 00:45:41,294 --> 00:45:43,294 Which I don't think really follows at all. 707 00:45:43,294 --> 00:45:44,994 It could, it could in certain streams. 708 00:45:44,994 --> 00:45:49,485 And that's why it gets even more complicated because there are certain streams of work, which I think will go away. 709 00:45:49,485 --> 00:45:53,209 right, they will get automated out of existence in terms of the lawyers putting their fingers on it. 710 00:45:53,209 --> 00:45:55,391 It will become commoditized, right? 711 00:45:55,391 --> 00:45:56,432 That will happen. 712 00:45:56,432 --> 00:46:02,457 But there'll be other areas of work where the AI will actually massively increase what lawyers can do. 713 00:46:02,457 --> 00:46:06,319 And they may end up becoming way, way, way more wealthy than they are today. 714 00:46:06,320 --> 00:46:09,913 And it's quite difficult because you both of these, both of these pictures are true. 715 00:46:09,913 --> 00:46:14,537 And I think that's one of the things that's and people like clear pictures, right? 716 00:46:14,537 --> 00:46:16,548 know, AI will destroy all lawyers. 717 00:46:16,548 --> 00:46:17,533 No, that's not true. 718 00:46:17,533 --> 00:46:19,773 Okay, so AI is irrelevant. 719 00:46:20,053 --> 00:46:21,333 No, that's not true. 720 00:46:21,333 --> 00:46:30,893 So it's kind of like, you know, it's like, how can AI actually eat up probably what will be in 10 years, a significant chunk of what is today of a legal market. 721 00:46:31,373 --> 00:46:37,964 And yet at the end of the day, Christmas 3035, is that right? 722 00:46:37,964 --> 00:46:40,144 Yeah, Christmas 35, right? 723 00:46:40,624 --> 00:46:43,364 There'll be a lot of lawyers who are saying, my God, thank God for AI. 724 00:46:43,364 --> 00:46:44,616 I'm so much wealthier now. 725 00:46:44,616 --> 00:46:50,499 And I don't want to jump out of a window because that's really the mindset of a lot of lawyers today. 726 00:46:50,499 --> 00:46:56,425 It can be a very brutal profession and coming up the ranks is sometimes a painful process. 727 00:46:56,425 --> 00:46:57,699 uh 728 00:46:57,699 --> 00:47:06,279 that's, that's one of the things I mean, I I gave a talk in Oxford the other day at a, and I don't know if it me something in public meeting somebody called Gunnar Cook, which is a 729 00:47:06,279 --> 00:47:07,859 distributed law firm. 730 00:47:07,859 --> 00:47:08,099 Right. 731 00:47:08,099 --> 00:47:11,470 So they're all individuals, but they work under the umbrella of Gunnar Cook. 732 00:47:11,470 --> 00:47:12,210 Right. 733 00:47:12,330 --> 00:47:15,830 And they're using AI systems now, which is fantastic. 734 00:47:15,850 --> 00:47:24,070 And, you know, we were talking about the fact that, you know, it's, it's just, it's an, it's all, it's an all, you know, it's all positive for them because they don't have 735 00:47:24,070 --> 00:47:24,900 leverage. 736 00:47:24,900 --> 00:47:28,521 It's mostly made up of experienced lawyers, right? 737 00:47:28,521 --> 00:47:31,553 They don't have this huge leverage associate group beneath them. 738 00:47:31,553 --> 00:47:34,885 And having AI greatly increases what they can do. 739 00:47:34,885 --> 00:47:41,817 And it does work that quite often if a client comes to an individual in a group like this, one of these distributed law firms, they're not coming to them because they know they've 740 00:47:41,817 --> 00:47:44,288 got 25 associates working in the basement. 741 00:47:44,288 --> 00:47:49,820 They come to them because they know that person has real expertise and will really add value on a very human level. 742 00:47:49,820 --> 00:47:51,440 But of course, 743 00:47:51,766 --> 00:47:55,048 leverage based work will creep in. 744 00:47:55,048 --> 00:47:56,630 But now they can use AI to do that. 745 00:47:56,630 --> 00:47:58,872 So I mean, it's an absolute win-win for them. 746 00:47:58,872 --> 00:48:10,549 And again, for these hybrid AI companies and the ALSPs and all of these, it's, mean, this is the thing that is kind of tantalizing is that we really are starting to see more and 747 00:48:10,549 --> 00:48:14,040 more examples of organizations using AI. 748 00:48:14,356 --> 00:48:15,946 and it's just all positive. 749 00:48:15,946 --> 00:48:22,786 And simultaneously, we're seeing both some law firms and some in-house legal teams kind of gritting their teeth and going, oh, I'm not sure. 750 00:48:22,786 --> 00:48:23,826 I'm not sure about this. 751 00:48:23,826 --> 00:48:24,946 I'm not sure about this. 752 00:48:24,946 --> 00:48:26,446 Is this a good idea? 753 00:48:26,686 --> 00:48:28,826 And you're just like... 754 00:48:28,857 --> 00:48:32,459 Yeah, there's still a lot of pearl clutching going on. 755 00:48:32,459 --> 00:48:39,384 Well, we're almost out of time, but this has been a great conversation as it always is every time we have you on. 756 00:48:39,384 --> 00:48:46,589 Before we hop off, how do people find out more about legal innovators and the writing that you do? 757 00:48:46,936 --> 00:48:50,707 Go to chat GPT and I'm not even kidding. 758 00:48:50,707 --> 00:48:56,090 I get a lot of traffic from, from LLMs now uh perplexity and others, right? 759 00:48:56,090 --> 00:48:59,792 Go to chat GPT and ask it about artificial lawyer. 760 00:48:59,792 --> 00:49:02,592 And hopefully it'll say something nice about me and provide a link. 761 00:49:02,592 --> 00:49:08,495 But if you don't want to do that, just go into Google and type artificial lawyer.com and you'll, you'll find me. 762 00:49:08,495 --> 00:49:10,447 But yeah, on the New York event. 763 00:49:10,447 --> 00:49:20,886 if I imagine some of your listeners are in the US, the New York event, Legal Innovators New York, the website is literally legalinnovatorsnewyork.com and it will be on the 19th 764 00:49:20,886 --> 00:49:22,878 and 20th of November. 765 00:49:22,878 --> 00:49:31,330 Day one is law firms, day two is in-house and obviously we'd like to see everybody on both days but if you want to focus on one or the other then you can take your pick and it will 766 00:49:31,330 --> 00:49:32,630 be in Midtown. 767 00:49:33,041 --> 00:49:33,801 Okay. 768 00:49:34,021 --> 00:49:35,462 That's great to hear. 769 00:49:35,462 --> 00:49:38,545 And I'm a big fan of the artificial lawyer site. 770 00:49:38,545 --> 00:49:40,226 I'm very selective. 771 00:49:40,226 --> 00:49:42,316 I get a few newsletters that I read every day. 772 00:49:42,316 --> 00:49:43,207 Yours is one of them. 773 00:49:43,207 --> 00:49:46,188 So um I really appreciate your time. 774 00:49:46,188 --> 00:49:50,291 Best of luck with the conference and I'm sure we'll bump into each other sometime real soon. 775 00:49:50,291 --> 00:49:51,325 I hope so, definitely. 776 00:49:51,325 --> 00:49:51,872 Thank you. 777 00:49:51,872 --> 00:49:52,737 All right, take care. 778 00:49:52,737 --> 00:49:54,185 Bye bye. 00:00:02,334 Richard Trumans, how are you this afternoon? 2 00:00:02,597 --> 00:00:04,721 ah Good, good, good. 3 00:00:04,721 --> 00:00:12,334 Slightly frazzled around the edges, but I've been turned over and I'm a sunny side up again now, so I'm pretty good. 4 00:00:12,334 --> 00:00:14,156 uh 5 00:00:14,156 --> 00:00:14,527 good. 6 00:00:14,527 --> 00:00:16,123 That's better than sunny side down. 7 00:00:16,123 --> 00:00:17,593 um 8 00:00:17,593 --> 00:00:18,194 I don't like that. 9 00:00:18,194 --> 00:00:20,225 mean, yeah, it seems that will go scrambled. 10 00:00:20,225 --> 00:00:20,945 There you go. 11 00:00:20,945 --> 00:00:27,589 Yeah, I'm a over medium kind of guy, whatever floats your boat. 12 00:00:27,589 --> 00:00:35,326 So I know you, think most of our audience knows you, but for the ones that don't, just give us a quick introduction about who you are, what you do, and where you do it. 13 00:00:35,326 --> 00:00:38,238 Yeah, well, I'll give you the sensible two minute version. 14 00:00:38,238 --> 00:00:47,874 So working in the legal sector for about 25 years, started off as a proper journalist, went to journalism school for my sins, ended up at a magazine called Legal Week, which 15 00:00:47,874 --> 00:00:52,837 doesn't exist anymore, which was bought eventually by law.com, focused on the international world. 16 00:00:52,837 --> 00:00:55,278 Then I became a foreign correspondent. 17 00:00:55,278 --> 00:01:02,122 After a couple of years decided that I'd much rather focus on the business of law, ended up in the city. 18 00:01:02,170 --> 00:01:05,450 Chem and management consultant helped to merge law firms together. 19 00:01:05,450 --> 00:01:19,745 It was very much focused on research of the market, big macroeconomic picture, did a lot of work over the years on profitability, business structure, business culture, and all of 20 00:01:19,745 --> 00:01:20,506 those things. 21 00:01:20,506 --> 00:01:24,869 And really, if I had been happy doing that, you probably would never have met me. 22 00:01:24,869 --> 00:01:28,873 In fact, no one in the legal tech world would ever, ever have met me. 23 00:01:28,873 --> 00:01:30,734 Artificial law would not exist. 24 00:01:30,790 --> 00:01:38,960 and I would be living in Surrey in a nice little cottage and commuting into London to do my little bits of consulting work with big law firms. 25 00:01:38,960 --> 00:01:42,043 And I'd still be wearing a suit and all of that kind of stuff. 26 00:01:42,043 --> 00:01:48,479 But, fatefully, in 2015, I went and started getting interested in technology. 27 00:01:48,479 --> 00:01:53,793 And I became fascinated with AI and I've always been interested in science in my free time. 28 00:01:53,915 --> 00:01:59,247 And I, in 2016, it bubbled up to the point where I suddenly realized, my God, it's going to happen. 29 00:01:59,247 --> 00:02:00,768 It's really going to happen. 30 00:02:00,768 --> 00:02:02,169 AI is going to change everything. 31 00:02:02,169 --> 00:02:10,373 And I launched Artificial Lawyer, which combined a bit of my business analysis and macro picture skills with my good old fashioned journalistic skills. 32 00:02:10,373 --> 00:02:18,776 People kind of think that I went from being a journalist to running Artificial Lawyer, but there's like an enormous gap in the middle of running around the city in a suit, trying to 33 00:02:18,776 --> 00:02:21,667 advise people on business, which no one ever talks about. 34 00:02:21,667 --> 00:02:23,580 understandably, it wasn't very exciting. 35 00:02:23,580 --> 00:02:29,905 It was it was to begin with it is I mean, honestly, when you've been a reporter, going, tell me about your profitability. 36 00:02:29,905 --> 00:02:38,693 And then you get invited inside the inner sanctum and they show you your your you know, they show you the spreadsheet and you see everything you like, wow, I you get to talk to 37 00:02:38,693 --> 00:02:41,216 managing partners and practice heads and everything. 38 00:02:41,216 --> 00:02:44,019 You get a completely different sense of what's going on. 39 00:02:44,019 --> 00:02:44,579 And 40 00:02:44,579 --> 00:02:53,259 That for me is why I think those two things when AI came around, I just thought, yes, this is going to change everything. 41 00:02:53,259 --> 00:02:59,799 And then of course it didn't, but I hung in there like a fool. 42 00:02:59,799 --> 00:03:01,379 hung in there and guess what? 43 00:03:01,379 --> 00:03:03,902 In 2022, it became real. 44 00:03:03,902 --> 00:03:04,502 Yeah. 45 00:03:04,502 --> 00:03:09,482 And well, and you were, I've talked about this on a previous episode with you, you were early. 46 00:03:09,562 --> 00:03:23,351 So the, really the birth of generative AI as I know it was, you know, the attention is all you need paper from Google, which I think was 2017, right? 47 00:03:23,351 --> 00:03:30,591 Around that time, that time, we were just having things because I'm just reading, you know, book about Empire of AI, great book. 48 00:03:31,071 --> 00:03:33,591 Some bits that aren't so great, but broadly. 49 00:03:35,111 --> 00:03:39,271 Let's not get into the whole looking for victims kind of thing. 50 00:03:39,430 --> 00:03:45,051 But it reminds me of that South Park episode about Megan and Harry. 51 00:03:45,871 --> 00:03:49,091 CSI Special Victims Unit, you know. 52 00:03:49,320 --> 00:03:53,130 There's a lot of sour grapes and whining and that, but yeah. 53 00:03:53,130 --> 00:03:53,661 Yeah. 54 00:03:53,661 --> 00:03:54,071 right. 55 00:03:54,071 --> 00:03:54,891 We'll edit that out. 56 00:03:54,891 --> 00:03:56,143 This will be censored. 57 00:03:56,143 --> 00:04:03,291 This will be sent to the George Orwell School of video editing to be sanitized before publication. 58 00:04:03,291 --> 00:04:09,597 But the interesting thing is that Google were one of the, if not the one, that really helped to pioneer the transformer model. 59 00:04:09,597 --> 00:04:11,489 And they didn't really take it that far. 60 00:04:11,489 --> 00:04:17,352 And then OpenAI, Altman and the team around him and Musk as well. 61 00:04:17,352 --> 00:04:20,132 deserves a lot of credit for backing them early on. 62 00:04:20,132 --> 00:04:21,432 They just picked it up and wrapped them up. 63 00:04:21,432 --> 00:04:25,552 This is one of the cool things, right, about him I saw in the book, which is really fascinating to me. 64 00:04:25,552 --> 00:04:27,232 They started off doing everything. 65 00:04:27,232 --> 00:04:28,732 They were doing the video. 66 00:04:28,732 --> 00:04:31,023 They also were doing robotics, right? 67 00:04:31,023 --> 00:04:37,863 They had a robotic hand because they were focused on AGI as they conceptualized it back in sort of 2016, 2017. 68 00:04:38,203 --> 00:04:42,743 They didn't know, they just had this idea there was going to be this super intelligent AI. 69 00:04:42,903 --> 00:04:43,893 right, human-like AI. 70 00:04:43,893 --> 00:04:46,644 They weren't quite sure how it was going to emerge. 71 00:04:46,644 --> 00:04:47,775 And they were looking at everything. 72 00:04:47,775 --> 00:04:51,386 They were looking at, like, sort of visual, video kind of stuff. 73 00:04:51,386 --> 00:04:53,166 They were looking at the physical realm. 74 00:04:53,166 --> 00:04:55,567 And they were also looking at language models. 75 00:04:55,707 --> 00:04:58,458 But those were all seen as, equally interesting. 76 00:04:58,458 --> 00:05:00,118 They didn't quite know which way it was going to go. 77 00:05:00,118 --> 00:05:05,810 And eventually, of course, once the LLM bit took off, they closed down the robotics bit or got rid of it. 78 00:05:05,810 --> 00:05:08,952 But now, interestingly, they've restarted it. 79 00:05:08,952 --> 00:05:10,392 going off on a tangent. 80 00:05:10,392 --> 00:05:15,832 I think the point I think that I mean, and the funny thing is all of that was going on unbeknownst to me. 81 00:05:15,832 --> 00:05:24,492 Well, I started running around trying to talk to people like Kira, know, no Waysburg and the Ebrevier and companies like that, talking to them about natural language processing, 82 00:05:24,492 --> 00:05:25,663 machine learning, la la la. 83 00:05:25,663 --> 00:05:36,634 And in an office somewhere in San Francisco, Palo Alto, were a bunch of people building the future with very, very little fanfare, very, very little fanfare. 84 00:05:36,634 --> 00:05:37,144 Yeah. 85 00:05:37,144 --> 00:05:37,564 Yeah. 86 00:05:37,564 --> 00:05:46,299 And, but, but you, came to the game, earlier than most and you know, I, paid attention to, and my listeners have, have heard me talk about this before. 87 00:05:46,299 --> 00:05:53,424 It was, when deep blue on Jeopardy was kind of my first introduction, right? 88 00:05:53,491 --> 00:05:54,565 It was IBM. 89 00:05:54,565 --> 00:05:54,946 Yep. 90 00:05:54,946 --> 00:05:59,650 And then alpha go, which was again, that was roughly the same time period. 91 00:05:59,650 --> 00:06:01,693 2017, 2016. 92 00:06:01,693 --> 00:06:02,025 wasn't it? 93 00:06:02,025 --> 00:06:04,865 Wasn't that, that was out of, yeah. 94 00:06:04,865 --> 00:06:10,539 deep mind, Dennis Hesabas that Google ended up buying deep mind. 95 00:06:10,539 --> 00:06:16,446 So Google has had ambitions and has made investments in AI for, for a very long time. 96 00:06:16,446 --> 00:06:20,129 And you know, their first model, couple of models really sucked. 97 00:06:20,129 --> 00:06:22,731 I mean, Bard was pretty bad. 98 00:06:22,731 --> 00:06:24,993 Um, they've really closed the gap. 99 00:06:24,993 --> 00:06:27,755 Gemini 2.5 is quite good. 100 00:06:27,755 --> 00:06:30,296 And I think they're on the verge of releasing. 101 00:06:30,318 --> 00:06:31,831 whatever that next version is. 102 00:06:31,831 --> 00:06:35,747 yeah, Google, Google was kind of slowly then suddenly. 103 00:06:36,779 --> 00:06:37,645 but they've caught up. 104 00:06:37,645 --> 00:06:38,525 thing, isn't it? 105 00:06:38,525 --> 00:06:41,825 You get these like plateau, kind of like watershed moments. 106 00:06:41,825 --> 00:06:49,445 But I mean, the interesting thing here is that the basic, you might say economic theory, the sort of techno economic theory, whatever you want to call it. 107 00:06:49,725 --> 00:06:59,205 But I sort of like, I would say I developed, but I certainly spent a ridiculously large amount of time working on it, you know, around, you know, moving away from the, I mean, 108 00:06:59,205 --> 00:07:02,661 all the key concepts were all there, well developed in the market. 109 00:07:02,661 --> 00:07:11,541 You might say that the only thing really artificial I did was just kind of like tie a little loose ends together and basically build almost like a worldview. 110 00:07:11,541 --> 00:07:13,521 This is the way it's going to be people, right? 111 00:07:13,521 --> 00:07:14,881 This is the way things are going to evolve. 112 00:07:14,881 --> 00:07:17,332 Well, they have to evolve if AI is going to be anything relevant. 113 00:07:17,332 --> 00:07:21,852 And it all came together like in a few months of 2016, right? 114 00:07:21,852 --> 00:07:23,812 Because it's not rocket science anyway, is it? 115 00:07:24,192 --> 00:07:29,808 And then the technology has moved on, but the basic ideas remain the same. 116 00:07:29,808 --> 00:07:33,611 It's rather like the green movement, know, batteries have improved. 117 00:07:33,770 --> 00:07:39,037 Now we have self-driving cars, which without humans are both safer and will be more efficient. 118 00:07:39,037 --> 00:07:40,759 Electric vehicles. 119 00:07:40,759 --> 00:07:51,488 Today they announced, I'm not sure which body, one of the large global energy bodies announced that this year was the first time ever that, well, actually you're going to have 120 00:07:51,488 --> 00:07:53,730 to edit this bit because I can't remember the facts. 121 00:07:53,730 --> 00:07:55,161 It's the... 122 00:07:55,282 --> 00:08:01,358 I think it was the first time that renewable energy overtook fossil fuels globally on global basis, something like that. 123 00:08:01,358 --> 00:08:01,999 Yeah. 124 00:08:01,999 --> 00:08:10,715 But the point is, that the tech moves forward, but the people who pioneered green energy had thought all of this through 20 years ago. 125 00:08:10,715 --> 00:08:11,896 Right. 126 00:08:11,896 --> 00:08:13,247 like, okay, so we're going to need wind. 127 00:08:13,247 --> 00:08:15,820 And then someone says, yes, but wind is very variable. 128 00:08:15,820 --> 00:08:19,233 So you're to need better battery technology. 129 00:08:19,233 --> 00:08:20,093 People go, yep, yep. 130 00:08:20,093 --> 00:08:20,664 Okay, right. 131 00:08:20,664 --> 00:08:21,784 Let's develop that. 132 00:08:21,784 --> 00:08:22,184 Right? 133 00:08:22,184 --> 00:08:27,264 And we're going to need a mix and we're going to have, you know, we're going to need better transmission systems and so forth. 134 00:08:27,264 --> 00:08:30,304 And we're going to massively have to improve battery technology. 135 00:08:30,644 --> 00:08:31,244 So on and so on. 136 00:08:31,244 --> 00:08:32,635 then we're to have to invest in some. 137 00:08:32,635 --> 00:08:42,495 And it's like the groundwork was laid out 20 years ago and we see this again and again in society that the, the, basic facts are pretty obvious for all to see. 138 00:08:42,515 --> 00:08:43,355 Right? 139 00:08:43,355 --> 00:08:46,986 It's just, they don't have the technology or the investment to make it real. 140 00:08:46,986 --> 00:08:49,657 Um, and then eventually the technology comes along. 141 00:08:49,657 --> 00:08:57,090 You know, it goes back to the, know, the old anecdotes about, you know, you know, revolutions needing the right conditions, right? 142 00:08:57,490 --> 00:09:02,660 You know, there's always revolutionaries in any society, any point in history, you all way back to ancient Greece. 143 00:09:02,660 --> 00:09:07,373 There's always a bunch of revolutionaries running around in their togas, you know, going, right. 144 00:09:07,373 --> 00:09:09,864 But it only takes off if you get the right conditions. 145 00:09:09,864 --> 00:09:11,121 And it's the same with technology. 146 00:09:11,121 --> 00:09:18,058 I mean, like I could have shouted until I was blue in the face until I keeled over about getting rid of the billable hour. 147 00:09:18,058 --> 00:09:27,573 by automating work streams, about rethinking the business model of law firms, about thinking about the end outputs, how that benefits society as a whole, all that kind of 148 00:09:27,573 --> 00:09:28,064 stuff, 149 00:09:28,064 --> 00:09:29,387 And we may never have got there. 150 00:09:29,387 --> 00:09:38,651 NLP would have just fizzled out, plateaued out, and you and I would still be, I mean, I don't know what I would be doing, but I'd still be doing legal tech. 151 00:09:38,651 --> 00:09:39,393 I'm not sure. 152 00:09:39,393 --> 00:09:41,653 Well, you know, this is it. 153 00:09:41,653 --> 00:09:46,253 I had a similar experience, similar, similar dynamic with cloud. 154 00:09:46,253 --> 00:09:57,833 So I was very early to pushing cloud in legal when I really didn't understand just how much friction towards change there was. 155 00:09:57,833 --> 00:10:00,033 So this is the early 2010s. 156 00:10:00,033 --> 00:10:08,865 And when something called wave 14 came through for office 365 and made, made it usable, like pre wave 157 00:10:08,865 --> 00:10:17,205 14 or maybe it was wave 15 office 365 now called Microsoft 365 was, was, was not great. 158 00:10:17,205 --> 00:10:24,185 Um, but I really thought to myself, wow, you know, I know what junk Microsoft junkies law firms are. 159 00:10:24,185 --> 00:10:25,345 This makes perfect sense. 160 00:10:25,345 --> 00:10:32,085 I know how little they like to invest in technology and uh, which has historically been the case. 161 00:10:32,085 --> 00:10:35,445 Um, moving to the cloud improves economics. 162 00:10:35,445 --> 00:10:38,025 It improves the security posture. 163 00:10:38,025 --> 00:10:38,978 Um, 164 00:10:38,978 --> 00:10:42,618 It certainly can scale up much more efficiently. 165 00:10:43,037 --> 00:10:47,498 And man, that didn't happen for about another eight years. 166 00:10:47,498 --> 00:10:49,138 It wasn't in the US. 167 00:10:49,138 --> 00:10:54,409 So law firms in the US really did not start to move in earnest to the cloud until a little bit after COVID. 168 00:10:54,409 --> 00:11:00,569 we, again, I beat my head against the wall, kind of like you did talking about AI. 169 00:11:00,769 --> 00:11:05,709 did 22 road shows in 2014 talking about the cloud. 170 00:11:05,765 --> 00:11:12,824 22 different cities and everybody's like, yeah, this is great and did absolutely nothing for another eight years. 171 00:11:12,986 --> 00:11:13,486 So. 172 00:11:13,486 --> 00:11:18,639 in some ways the AI situation was even more crazy in that cloud clearly did work. 173 00:11:18,639 --> 00:11:21,581 You could demonstrate it, but they still didn't want it. 174 00:11:21,581 --> 00:11:27,184 Well, it's not going to that maybe cloud is actually more crazy, but I suppose to some degree you could actually not blame the lawyers much because they were looking at the 175 00:11:27,184 --> 00:11:31,716 early NLP tools and going, man, this looks kind of complicated. 176 00:11:31,716 --> 00:11:33,608 I'm not sure I want to use this stuff. 177 00:11:33,608 --> 00:11:39,972 And then you've got all the economic macroeconomic, cultural, professional aspects that kind of stopped it. 178 00:11:39,972 --> 00:11:42,653 But yeah, I mean, it's 179 00:11:42,745 --> 00:11:43,866 I mean, this the thing, isn't it? 180 00:11:43,866 --> 00:11:52,333 Great, I mean, you cannot kill a great idea, but you can delay it for an enormous, enormous amount of time, right? 181 00:11:52,333 --> 00:11:59,659 I mean, and you see this again, you see this in politics, see this playing out over all kinds of different trains in our society. 182 00:11:59,659 --> 00:12:01,201 You cannot kill a great idea. 183 00:12:01,201 --> 00:12:04,904 People just instinctively look at it, they compare it to their own experiences. 184 00:12:04,904 --> 00:12:07,626 Maybe it's even like just like, know, it's gut instinct, right? 185 00:12:07,626 --> 00:12:08,277 It's intuition. 186 00:12:08,277 --> 00:12:10,625 They just go, that's right, that is right. 187 00:12:10,625 --> 00:12:12,785 And they're like, okay, let's have it. 188 00:12:13,125 --> 00:12:14,125 And then it doesn't happen. 189 00:12:14,125 --> 00:12:15,545 And then it doesn't happen. 190 00:12:15,545 --> 00:12:17,685 Although someone tries it, it doesn't really work very well. 191 00:12:17,685 --> 00:12:24,725 And then all the critics and the cynics start jumping on and going, ah, ha, ha, you were a fool to believe there is hope, young Skywalker. 192 00:12:24,785 --> 00:12:28,925 ha, ha, you know, come to the dark side and enjoy the empire. 193 00:12:28,925 --> 00:12:31,185 It's much nicer over here. 194 00:12:31,365 --> 00:12:34,025 We get much better food, et cetera, et cetera. 195 00:12:34,025 --> 00:12:34,876 office is lovely. 196 00:12:34,876 --> 00:12:39,062 And people give up or they just get disbanded and make off and. 197 00:12:39,062 --> 00:12:41,875 focus something that does work, right? 198 00:12:41,875 --> 00:12:45,377 Like, I don't know, Bitcoin. 199 00:12:45,495 --> 00:12:46,566 That's I was going to say. 200 00:12:46,566 --> 00:12:53,349 Digital currency is another great example of an idea that's been lingering now for almost 20 years. 201 00:12:53,349 --> 00:12:54,029 It was 2008-ish. 202 00:12:54,029 --> 00:12:55,368 But it makes it makes people money. 203 00:12:55,368 --> 00:12:56,020 mean, that's the thing. 204 00:12:56,020 --> 00:13:01,475 mean, I mean, you know, you can step away from the whole intellectual stuff if you want to and just say, look, does it make money? 205 00:13:01,816 --> 00:13:03,297 If it does make money? 206 00:13:05,659 --> 00:13:06,579 Does. 207 00:13:06,940 --> 00:13:08,894 Well, this is weird thing with crypto. 208 00:13:08,894 --> 00:13:17,349 mean, unless we have a collapse of all fiat currencies and we end up in a sort of Mad Max world, crypto currencies actually have no real functional value. 209 00:13:17,349 --> 00:13:18,737 They are simply 210 00:13:18,737 --> 00:13:26,284 gigantic momentum trades and the momentum is constantly up, of flips around a bit and then keeps going up again. 211 00:13:26,284 --> 00:13:28,008 But fundamentally, there's nothing there. 212 00:13:28,008 --> 00:13:29,353 There is literally nothing there. 213 00:13:29,353 --> 00:13:30,798 argue the same about gold. 214 00:13:30,798 --> 00:13:34,666 Gold's value is driven primarily by scarcity. 215 00:13:34,944 --> 00:13:37,686 True, But it has a tangible value. 216 00:13:37,686 --> 00:13:42,800 mean, you can pull out a lump of gold and people go, gold. 217 00:13:42,881 --> 00:13:45,464 People will always want gold. 218 00:13:45,464 --> 00:13:51,919 The same way that people will always want kebabs or burgers or, you know, it's just one of those things, right? 219 00:13:51,919 --> 00:13:52,671 People like it. 220 00:13:52,671 --> 00:13:59,848 And so, whereas cryptocurrency really is just a momentum trade, maybe that momentum will keep going forever and ever. 221 00:14:00,184 --> 00:14:06,597 Right, rather like a YouTube channel that just keeps on accumulating views and followers. 222 00:14:06,757 --> 00:14:07,758 It'll never, never stop. 223 00:14:07,758 --> 00:14:10,199 It just goes up and up and up and up. 224 00:14:10,199 --> 00:14:17,802 But if it ever did flip, like really flip, not just fluctuate, but completely flip and go reverse, there's nothing that you can't cash it in at the end. 225 00:14:18,174 --> 00:14:19,543 You what have I got left? 226 00:14:19,543 --> 00:14:20,935 I've got three electrons. 227 00:14:20,935 --> 00:14:22,160 Are they worth anything? 228 00:14:22,160 --> 00:14:24,396 It's like, no, I'm sorry, your electrons are not really worth anything. 229 00:14:24,396 --> 00:14:26,178 you've got a collection of prime numbers. 230 00:14:26,178 --> 00:14:35,747 um But the underlying technology of blockchain provides a tremendous opportunity and utility. 231 00:14:35,747 --> 00:14:37,468 And there's been a lot of talk in legal. 232 00:14:37,468 --> 00:14:39,490 It really hasn't gotten anywhere. 233 00:14:39,490 --> 00:14:47,014 the asynchronous ledger, distributed ledger concept is an interesting one in recording. 234 00:14:47,014 --> 00:14:47,674 is, is. 235 00:14:47,674 --> 00:14:50,085 mean, you know, David Fisher, know, integral ledger. 236 00:14:50,085 --> 00:14:52,997 He's been pushing that for ever since I got started. 237 00:14:52,997 --> 00:15:00,020 And, he still has a lot of people who really believe in it, you know, but the problem is, is it's really be useful. 238 00:15:00,020 --> 00:15:05,483 My personal view is you need a scenario where there's no trust, where trust is completely broken down. 239 00:15:05,483 --> 00:15:08,835 And most commercial lawyers don't operate in such a system. 240 00:15:08,835 --> 00:15:11,956 Probably this reason is because if trust had completely. 241 00:15:11,956 --> 00:15:19,960 utterly irrevocably broken down and no one can trust each other and nothing could be real, it'd be quite hard to be a lawyer because you've lost a rules-based system that underpins 242 00:15:19,960 --> 00:15:20,930 all of that. 243 00:15:21,711 --> 00:15:23,191 You're in Mad Max. 244 00:15:23,211 --> 00:15:30,294 And at that point, cryptocurrency and blockchain and everything really makes absolute sense because you cannot trust anybody to do the right thing. 245 00:15:30,486 --> 00:15:31,366 Yeah. 246 00:15:31,507 --> 00:15:42,519 Yeah, that's a, we could go down a rabbit hole with the cryptocurrency, but, let's talk about your, so you've got a couple of events coming up that, I wanted to make people aware 247 00:15:42,519 --> 00:15:48,815 of because your theme aligns really well with the theme of this podcast, which is legal innovation. 248 00:15:48,815 --> 00:15:53,189 So tell us about the, I think you have events coming up in New York and London. 249 00:15:53,189 --> 00:15:54,960 Tell us a little bit about those. 250 00:15:55,078 --> 00:16:00,580 Yeah, so just briefly, so Legal Innovators was a conference that grew out of Artificial Lawyer. 251 00:16:00,580 --> 00:16:06,044 It's organized by another company called Cosmonauts, but I kind of co-created it with them and I chair the events. 252 00:16:06,044 --> 00:16:07,982 London, we've been doing for years and years. 253 00:16:07,982 --> 00:16:09,515 It's quite a big conference now. 254 00:16:09,515 --> 00:16:12,868 We're hoping to get around about thousand people over three days. 255 00:16:12,868 --> 00:16:17,641 Law firm day, in-house day, litigation day, which is gonna be fun, first time we've done that. 256 00:16:17,641 --> 00:16:21,062 So that's the fourth, fifth and sixth of November in London. 257 00:16:21,184 --> 00:16:23,846 And then we're to be doing New York for very first time. 258 00:16:23,846 --> 00:16:28,100 We've been doing California for years and years in San Francisco, which is brilliant fun. 259 00:16:28,100 --> 00:16:28,427 love it. 260 00:16:28,427 --> 00:16:31,894 It's probably my favorite event, but we're doing New York and it's going be very interesting. 261 00:16:31,894 --> 00:16:42,883 Now, the reason why I delayed doing New York for so long is because I didn't feel that the big firms there who really rule, you know, the kingdom were really, really getting into AI 262 00:16:42,883 --> 00:16:49,290 and automation and rethinking business sufficiently to really make it an exciting event. 263 00:16:49,290 --> 00:16:51,731 I mean, Cleary was an outlier early on. 264 00:16:51,731 --> 00:16:54,833 They created Cleary X, did a lot of interesting work. 265 00:16:54,833 --> 00:16:57,073 But now things are changing. 266 00:16:57,073 --> 00:16:57,544 Definitely. 267 00:16:57,544 --> 00:17:04,067 I mean, when I was in New York for Legal Week earlier this year, I spoke to a lot of people that I got the feeling that things are shifting. 268 00:17:04,067 --> 00:17:07,268 You know, and there was some really good stuff going on at Gunderson as well, for example. 269 00:17:07,268 --> 00:17:12,360 ah It is, is, but in their New York office. 270 00:17:12,360 --> 00:17:13,170 it's true. 271 00:17:13,170 --> 00:17:15,980 But let's face it, Leif Mawatkins is a West Coast firm as well. 272 00:17:15,980 --> 00:17:17,604 You know, they're doing some interesting things. 273 00:17:17,604 --> 00:17:21,344 So it's kind of like, I think it's come of age finally. 274 00:17:21,644 --> 00:17:32,144 kind of, I mean, you expect Wilson, Sonsini and Cooley and others to have created startups and have really huge innovation teams, doing all kinds of fun stuff, working very closely 275 00:17:32,144 --> 00:17:33,484 with AI companies. 276 00:17:33,804 --> 00:17:39,584 New York firms, I think were slightly, they may have brought in a lot of people around knowledge management. 277 00:17:39,584 --> 00:17:41,044 They certainly did. 278 00:17:41,224 --> 00:17:44,884 But this idea that, hey, AI is actually gonna change the game for us. 279 00:17:44,884 --> 00:17:46,980 You know, just a little. 280 00:17:46,980 --> 00:17:48,000 Just a little. 281 00:17:48,161 --> 00:17:54,997 I think that was, even though people may have talked a good talk, I didn't really feel that it was actually happening until very recently. 282 00:17:54,997 --> 00:17:58,412 And I think it is actually starting to happen though. 283 00:17:58,412 --> 00:17:58,842 So. 284 00:17:58,842 --> 00:17:59,482 it is. 285 00:17:59,482 --> 00:18:04,415 And I think a leading indicator to that you alluded to, which is some of the hiring. 286 00:18:04,415 --> 00:18:05,916 mean, look at Simpson Thatcher. 287 00:18:05,916 --> 00:18:09,388 I mean, that's the epitome of a New York white shoe law firm. 288 00:18:09,388 --> 00:18:22,315 You know, they brought in Liz Grennan from McKinsey and they've made some other hires that again, that's how from the outside, if they're a client, I have got an inside view, but 289 00:18:22,315 --> 00:18:23,936 for firms who 290 00:18:24,289 --> 00:18:26,071 We don't do business with an info dash. 291 00:18:26,071 --> 00:18:30,694 look at their org chart and say, how are they, what moves are they making there? 292 00:18:30,694 --> 00:18:39,762 And you know, somebody like Liz isn't going to come from McKinsey without having a high level of commitment. 293 00:18:39,762 --> 00:18:46,988 Cause she's also an attorney and Connor Grennan's wife, who, you know, is a big thought leader in the AI space. 294 00:18:46,988 --> 00:18:52,192 think he's a NYU law school or I'm sorry, not as he'd law school. 295 00:18:52,676 --> 00:18:54,317 or maybe business school. 296 00:18:54,458 --> 00:19:04,788 So, you know, I see people who I know they're not going to move unless the firm has demonstrated a commitment and articulated some sort of a plan to move in a certain 297 00:19:04,788 --> 00:19:05,889 direction. 298 00:19:05,889 --> 00:19:10,815 And that tells me again, that's a leading indicator that they're starting to move in that direction. 299 00:19:10,815 --> 00:19:11,976 uh 300 00:19:11,976 --> 00:19:14,396 yeah, we shouldn't get carried away. 301 00:19:14,396 --> 00:19:23,747 I mean, while organizations like Salesforce out in California literally have their own team that builds AI products for their own use, right? 302 00:19:23,747 --> 00:19:31,567 You know, and some of the banks and some of the private equity funds are getting definitely very interested in what AI can do for them and doing stuff internally as well. 303 00:19:31,567 --> 00:19:35,767 And some of the big consultancies too, have big bases in New York as well. 304 00:19:35,767 --> 00:19:36,987 You know, 305 00:19:37,215 --> 00:19:41,158 Are the big New York firms really disrupting themselves? 306 00:19:41,158 --> 00:19:43,769 I don't think yet, and maybe not for a while. 307 00:19:43,769 --> 00:19:49,203 But as you say, they're genuinely thinking about it now and they're exploring and experimenting. 308 00:19:49,203 --> 00:19:53,546 They're bringing in these gen-ai productivity platforms and a whole bunch of other tools. 309 00:19:53,546 --> 00:19:57,588 It's going much, much more beyond knowledge management. 310 00:19:57,589 --> 00:20:01,431 It's not just like, okay, well make my life a little bit easier. 311 00:20:01,672 --> 00:20:05,455 Help me find the information I need so I can go off and do the deal that I need to do. 312 00:20:05,455 --> 00:20:06,605 It's more... 313 00:20:06,605 --> 00:20:16,459 now, well not more now, but in addition now, they're saying, okay, you know what, this may actually alter our business model at a very small level to begin with. 314 00:20:16,459 --> 00:20:18,100 But that, it's all relative, isn't it? 315 00:20:18,100 --> 00:20:18,881 It's all relative. 316 00:20:18,881 --> 00:20:24,454 know, that, that in itself, the fact that they're even asking that question, I think is a big deal. 317 00:20:24,454 --> 00:20:24,784 Yeah. 318 00:20:24,784 --> 00:20:32,432 Well, and let me ask you something along those lines when we talk about how law firms move in that direction. 319 00:20:32,432 --> 00:20:44,122 And by that direction, mean, you know, making fundamental changes to foundational pieces of their business, how legal service delivery, pricing, internal firm compensation, client 320 00:20:44,122 --> 00:20:45,022 engagement. 321 00:20:45,022 --> 00:20:51,267 These are foundational elements that are very firmly entrenched in law firms. 322 00:20:51,431 --> 00:20:54,733 and they have to change and they're going to change. 323 00:20:54,733 --> 00:21:00,896 Whether law firms like it or not, it's just at what speed and how. 324 00:21:00,896 --> 00:21:04,477 think one of the biggest ones is internal firm compensation. 325 00:21:04,477 --> 00:21:10,060 way lawyers are ruthlessly measured on billable hours today and that is such a deeply entrenched. 326 00:21:10,060 --> 00:21:16,133 I've talked to some leaders of innovation councils at some pretty big firms and they have cited that 327 00:21:16,255 --> 00:21:28,775 particular cultural element of the firm is one of the hardest to move the dial on where, because again, if you're AI really changes the internal firm compensation, when pricing 328 00:21:28,775 --> 00:21:31,917 changes, internal firm compensation changes, right? 329 00:21:31,917 --> 00:21:43,555 If you're in a AFA scenario where the firm is taking on risk, revenue in the door doesn't equate to profitability like it does today. 330 00:21:43,933 --> 00:21:46,891 you have to manage risk and efficiency. 331 00:21:46,984 --> 00:21:48,015 It's an interesting one, isn't it? 332 00:21:48,015 --> 00:21:52,658 Cause it's like, you know, the, I don't think there's anything wrong in using time as an internal metric. 333 00:21:52,658 --> 00:21:57,031 I don't think it's the best metric internally, but if you, if you want to use it, okay, fine. 334 00:21:57,031 --> 00:22:04,577 At least you've got some kind of yardstick, but I think for, know, there's Bob, he's only worked 20 hours this week. 335 00:22:04,577 --> 00:22:08,810 Ooh, you know, there's Jane, she's worked 60 hours or a hundred hours this week. 336 00:22:08,810 --> 00:22:09,251 she's great. 337 00:22:09,251 --> 00:22:10,031 Okay. 338 00:22:10,031 --> 00:22:10,922 Whatever. 339 00:22:11,534 --> 00:22:18,280 That kind of thinking is as old as time, But what I think is an issue is judging value by time. 340 00:22:18,380 --> 00:22:23,073 So, you know, here's Bob and Jane, they spend all weekend going through these documents, aren't they wonderful? 341 00:22:23,073 --> 00:22:26,366 That'll be, you know, $25,000, please. 342 00:22:26,366 --> 00:22:29,487 Oh, no, that AI system did it in three minutes. 343 00:22:29,848 --> 00:22:30,879 Now what are we gonna do? 344 00:22:30,879 --> 00:22:32,583 It's nuts, I personally think. 345 00:22:32,583 --> 00:22:36,918 So yeah, keep the art internally, but externally, you're gonna have to find a better method. 346 00:22:36,918 --> 00:22:38,080 Yeah, indeed. 347 00:22:38,080 --> 00:22:52,148 you know, and what, are your thoughts on, I've had this debate a lot and I would say from the people I've talked to, it's pretty split, maybe 60, 60, 40 in favor of, and 40 % think 348 00:22:52,148 --> 00:22:54,870 that this can't happen, but I'm a believer. 349 00:22:54,870 --> 00:23:04,099 think that law firms are actually going to have to leverage their data in a meaningful way as part of their AI strategy and 350 00:23:04,099 --> 00:23:11,239 and invest in R and D, which law firms historically have not done well or haven't done at all for a couple of reasons. 351 00:23:11,239 --> 00:23:16,539 One, the partnership model optimizes is optimized for profit taking at the end of the year. 352 00:23:16,539 --> 00:23:23,039 It's hard to accrue capital expenditure doesn't fit into that cash basis partnership model. 353 00:23:23,039 --> 00:23:27,459 Well, but also they're, they're not, they're not software companies. 354 00:23:27,459 --> 00:23:29,979 They're not used to building software. 355 00:23:30,239 --> 00:23:32,879 And I don't know, what are you? 356 00:23:32,879 --> 00:23:33,899 Yeah. 357 00:23:34,432 --> 00:23:42,458 think, you know, law firms want to build stuff, go for it, you know, the same way that someone might say to me, well, you don't have to buy that piece of technology to improve 358 00:23:42,458 --> 00:23:43,439 artificial wear. 359 00:23:43,439 --> 00:23:44,539 Why don't you build it yourself? 360 00:23:44,539 --> 00:23:45,454 It's really easy. 361 00:23:45,454 --> 00:23:48,702 You know, it's very, very, you know, pain by numbers. 362 00:23:48,983 --> 00:23:50,344 I might do it. 363 00:23:50,344 --> 00:23:54,738 I'll be honest, I'm probably more minded just to buy it and let someone else deal with it. 364 00:23:54,738 --> 00:23:57,250 I don't know, that for me is not the issue. 365 00:23:57,250 --> 00:24:03,626 It's the issue is can they start to think in a way that AI is the leverage? 366 00:24:03,626 --> 00:24:13,998 And I just want to bring in a law firm, an AI hybrid law firm, which I think is great, which is called Covenant based in New York, created by the ex-general council of WeWork 367 00:24:13,998 --> 00:24:15,443 and her partner. 368 00:24:15,443 --> 00:24:17,065 And it's a great model. 369 00:24:17,065 --> 00:24:17,665 It's a great model. 370 00:24:17,665 --> 00:24:21,408 They have a tech company, which is the AI bit. 371 00:24:21,408 --> 00:24:25,452 And they have a law firm, which has got multiple lawyers in it, not a huge number at the moment, just a small group. 372 00:24:25,452 --> 00:24:28,425 And they do private equity investments. 373 00:24:28,425 --> 00:24:30,887 I understand. 374 00:24:30,887 --> 00:24:31,769 And on the buy side. 375 00:24:31,769 --> 00:24:34,142 And there's a license agreement between the two. 376 00:24:34,142 --> 00:24:39,648 So they don't actually have to go away to Arizona to, you know, to do this. 377 00:24:39,648 --> 00:24:42,519 under special bar rules and it's, it's just great. 378 00:24:42,519 --> 00:24:51,171 then we were talking, we did this great video together and Jen, and she just came up with this phrase, just came out in the conversation and she said, you know, AI is the leverage. 379 00:24:51,331 --> 00:25:02,154 And I just thought, man, I wish I'd coined that phrase 10 years ago, because for me, it sums up everything that artificial law has been trying to say ever since I got started, 380 00:25:02,154 --> 00:25:05,435 that AI is the means of production. 381 00:25:05,635 --> 00:25:05,885 Right. 382 00:25:05,885 --> 00:25:08,636 Now there are some things which aren't sort of productivity based. 383 00:25:08,636 --> 00:25:16,396 You know, so, know, the classic cliche, know, it's Martin Lipton, you know, sitting on the sofa thinking about the poison pill and he has a eureka moment. 384 00:25:16,396 --> 00:25:18,356 goes, aha, I've got it. 385 00:25:18,356 --> 00:25:18,576 All right. 386 00:25:18,576 --> 00:25:21,816 That's not a productivity play, but a due diligence report. 387 00:25:21,816 --> 00:25:23,296 That's a productivity play. 388 00:25:23,336 --> 00:25:24,336 A lot of legal research. 389 00:25:24,336 --> 00:25:24,656 Okay. 390 00:25:24,656 --> 00:25:27,956 There's a degree of sophistication around that, but a lot of it is bulk. 391 00:25:28,076 --> 00:25:29,856 E-discovery, that's bulk. 392 00:25:29,856 --> 00:25:30,516 Okay. 393 00:25:30,516 --> 00:25:34,476 Other intelligent insights and knowledge we can make that more sophisticated. 394 00:25:34,476 --> 00:25:34,796 Better. 395 00:25:34,796 --> 00:25:35,656 Yeah, obviously. 396 00:25:35,656 --> 00:25:37,148 But there's a ton of bulk there. 397 00:25:37,148 --> 00:25:39,230 So that's all going down to productivity, right? 398 00:25:39,230 --> 00:25:45,353 Anything that relates to productivity, you throw at leverage, IAV associate body, right? 399 00:25:45,353 --> 00:25:49,515 And the way we need to think is AI is the leverage. 400 00:25:49,595 --> 00:25:56,789 And I think for me, that's a tipping point once law firms, the big established law firms start to look at Covenant and go, you know what? 401 00:25:56,789 --> 00:25:58,900 That's actually the best model we've ever seen. 402 00:25:59,221 --> 00:26:05,886 And the beautiful thing about it is they don't even have to get into complicated bar rules, right? 403 00:26:05,886 --> 00:26:11,159 They don't even have to do that because if they structure it in the right way, they can do this anyway. 404 00:26:11,159 --> 00:26:18,702 And besides, and even if they don't want to do that fancy, fancy structure, they can just buy in the technology as a license holder. 405 00:26:18,863 --> 00:26:26,206 I mean, it's not these, again, it goes back to this earlier conversation about the great idea is already here. 406 00:26:26,206 --> 00:26:27,798 The question is, will they do it? 407 00:26:27,798 --> 00:26:28,329 Right. 408 00:26:28,329 --> 00:26:29,289 Yeah. 409 00:26:29,289 --> 00:26:36,509 Did you hear, I'm assuming you read the story about Burford capital and the MSO model there. 410 00:26:36,509 --> 00:26:37,729 It's a private equity firm. 411 00:26:37,729 --> 00:26:39,609 was in the financial times. 412 00:26:39,727 --> 00:26:46,599 I mean, litigation financial financiers are doing pretty interesting stuff. 413 00:26:46,599 --> 00:26:47,780 Yeah. 414 00:26:47,780 --> 00:26:59,755 They essentially stand up these managed services organizations and deploy private capital into those, which really kind of take over some business of law functions. 415 00:26:59,755 --> 00:27:01,165 And it's a vehicle. 416 00:27:01,165 --> 00:27:05,758 It's an investment vehicle for private capital to get deployed in the US legal market. 417 00:27:05,758 --> 00:27:10,000 And it feels kind of a little bit like a back door at the moment. 418 00:27:10,000 --> 00:27:12,521 If things were 419 00:27:12,927 --> 00:27:20,743 If we had the ABS roles like we do in Arizona across the country, I'm not so sure that would be the ideal structure. 420 00:27:20,764 --> 00:27:26,348 But one of the things I wanted to get your take on is I talk about this a lot. 421 00:27:26,588 --> 00:27:32,073 Law firms today, I spent 10 years at Bank of America in mostly risk management roles. 422 00:27:32,073 --> 00:27:41,298 So I was in consumer risk, then I moved into global treasury, then I went to corporate audit, then I was in anti-money laundering and then compliance. 423 00:27:41,298 --> 00:27:51,203 really interesting work, but I got a really good sense for just how much rigor there is around risk management and financial services for obvious reasons. 424 00:27:51,544 --> 00:27:54,645 And looking at law firms, they have none. 425 00:27:54,965 --> 00:28:00,297 There's literally nothing between the lawyer and his or her client, right? 426 00:28:00,297 --> 00:28:05,750 It's I make requests, lawyer delivers work product or advice. 427 00:28:05,750 --> 00:28:06,870 There's no 428 00:28:07,018 --> 00:28:17,229 layers and when we get into a scenario like where we start to talk about automation and legal, you know, tech enabled legal service delivery, we have to start thinking about risk 429 00:28:17,229 --> 00:28:18,230 management. 430 00:28:18,230 --> 00:28:20,293 And what does that look like in a law firm? 431 00:28:20,293 --> 00:28:25,197 It would be lawyers listening to some risk manager saying, you can't do that. 432 00:28:25,298 --> 00:28:27,661 That's not going to go over very well, but I don't know. 433 00:28:27,661 --> 00:28:28,612 How do you think about 434 00:28:28,612 --> 00:28:29,313 two different things. 435 00:28:29,313 --> 00:28:36,418 think, I think one you sort of indirectly touched on absolutely key point, which is the lawyers are the means of production, right? 436 00:28:36,418 --> 00:28:38,030 They are the center of the universe, right? 437 00:28:38,030 --> 00:28:43,966 We have, we do not live in the Copernican, Copernican, Copernian system, right? 438 00:28:43,966 --> 00:28:46,348 know, lawyers are the center of the universe. 439 00:28:46,348 --> 00:28:51,372 They do not, they do not revolve around process and product and so forth. 440 00:28:51,372 --> 00:28:51,792 Right? 441 00:28:51,792 --> 00:28:53,053 The lawyers, 442 00:28:53,413 --> 00:28:54,514 control everything. 443 00:28:54,514 --> 00:28:56,315 They do the final sign off. 444 00:28:56,315 --> 00:28:58,278 They do the interaction with the clients. 445 00:28:58,278 --> 00:28:59,898 They are the workflow. 446 00:28:59,979 --> 00:29:02,702 They have a mind, they have a hand, they have a whole machine. 447 00:29:02,702 --> 00:29:07,529 And it's only when we start to think about automation in the true sense. 448 00:29:07,529 --> 00:29:09,551 So you don't do that anymore. 449 00:29:09,731 --> 00:29:11,614 The machine does that. 450 00:29:11,614 --> 00:29:13,875 Talking about agentic flows or anything else. 451 00:29:14,195 --> 00:29:15,375 That's when things change. 452 00:29:15,375 --> 00:29:17,675 And I think that's when it comes to your point about risk. 453 00:29:17,675 --> 00:29:25,195 I mean, as it currently stands, it doesn't make much difference because anything that as far as I understand, anything that goes out of the door of a law firm has to be signed off 454 00:29:25,195 --> 00:29:26,546 and is under their umbrella. 455 00:29:26,546 --> 00:29:32,937 What'll be interesting is as we start to build truly automated workflows, from start to finish. 456 00:29:32,937 --> 00:29:37,917 Now it may be a very, very narrow workflow, but they will grow. 457 00:29:37,917 --> 00:29:38,927 They will grow. 458 00:29:38,927 --> 00:29:41,979 It's going to get more and more powerful. 459 00:29:41,979 --> 00:29:44,220 That's when I think the whole risk and insurance thing comes in. 460 00:29:44,220 --> 00:29:47,762 But even so, could argue that law firms still have it under their umbrella. 461 00:29:47,762 --> 00:29:59,226 And it will be down to the law firm or any consultants they can bring in to do pen testing effectively, to make sure that it works completely fine. 462 00:29:59,226 --> 00:30:03,647 But yeah, for me, this has always been the battle in the... 463 00:30:03,829 --> 00:30:06,581 And it's totally understandable, because I probably would do the same. 464 00:30:06,581 --> 00:30:13,248 Most lawyers, most professionals, they see technology and they go, great, how can that add to what I do already? 465 00:30:13,348 --> 00:30:18,292 How can that finesse or take a little bit of a bother out of my life? 466 00:30:18,412 --> 00:30:22,116 All right, they're the center of the universe, right? 467 00:30:22,116 --> 00:30:27,380 If that is all we do with AI now, then nothing's going to change at all. 468 00:30:27,380 --> 00:30:32,564 It goes back to, don't know if we'll probably have this point before, it becomes the Ikea catalog situation. 469 00:30:32,588 --> 00:30:38,451 where you get your various shelving units and cushions and rugs and throwers and all of this kind of stuff. 470 00:30:38,451 --> 00:30:43,514 And it's very pretty and it's very nice and it greatly increases the comfort of that person. 471 00:30:43,514 --> 00:30:44,184 And why not? 472 00:30:44,184 --> 00:30:47,292 People like to be comfortable, but it doesn't change fundamentally anything. 473 00:30:47,292 --> 00:30:50,318 You're not Le Corbusier completely redesigning the building. 474 00:30:50,318 --> 00:30:56,181 You don't change your one bedroom flat into a machine for living, as Le Corbusier said. 475 00:30:56,181 --> 00:31:00,803 We're fundamentally still in the same world with some decorations from Ikea, right? 476 00:31:00,803 --> 00:31:02,804 Bought out of the catalog and then installed. 477 00:31:02,841 --> 00:31:07,102 Things only change once you start to automate whole streams. 478 00:31:07,422 --> 00:31:17,548 And I think that's, and I think this is incredibly difficult for professionals, particularly lawyers, to get their heads around because it's just like, yes, you are not 479 00:31:17,548 --> 00:31:20,609 gonna own everything any longer. 480 00:31:21,029 --> 00:31:26,532 You might be able to own the output and make money from it, but you will not own those workflows, right? 481 00:31:26,532 --> 00:31:28,572 And this is not like an assistant. 482 00:31:28,572 --> 00:31:30,467 I was thinking about how to... 483 00:31:30,467 --> 00:31:33,840 explain this in a very simple diagram for a thing I've got to do next week. 484 00:31:33,840 --> 00:31:38,645 So basically think about a partner, a partner has assistants, right? 485 00:31:38,645 --> 00:31:41,466 As they were kind of originally called rather than associates, right? 486 00:31:41,466 --> 00:31:46,720 That the assistant assists the partner in the things that the partner wants to achieve, right? 487 00:31:46,720 --> 00:31:50,211 So the partner says, right, I need you to go and look at that law book, go. 488 00:31:50,211 --> 00:31:52,612 I need you to prove that contract, go. 489 00:31:52,612 --> 00:31:55,883 You go and pick me up a sandwich, go, right? 490 00:31:55,883 --> 00:31:57,834 They're assistants, right? 491 00:31:57,840 --> 00:32:06,157 They may contribute meaningful aspects to the end product, but fundamentally they're completely under the command of this one person, right? 492 00:32:06,157 --> 00:32:08,019 They are the center of our universe, right? 493 00:32:08,019 --> 00:32:15,136 But in true automation and sort of an agentic flow, there is simply a person at one end of it, because it's all pre-built, right? 494 00:32:15,136 --> 00:32:19,168 Even if it can be uh go through iterations and you can have some feedback. 495 00:32:19,168 --> 00:32:24,583 But ultimately, there's one person at the end of that line and there's another person at the beginning of it. 496 00:32:24,583 --> 00:32:27,043 And that's it, right? 497 00:32:27,263 --> 00:32:31,474 You know, there is no assistant in this thing, right? 498 00:32:31,474 --> 00:32:36,174 It's a machine that makes something that comes out of the other end, to put it really simply. 499 00:32:36,474 --> 00:32:46,274 And I think this is very, very, difficult because it's the clash, and this is something that Susskind wrote about 25 million years ago, about the battle between the artisanal and 500 00:32:46,274 --> 00:32:47,414 the automation. 501 00:32:47,634 --> 00:32:52,642 when I started doing artificial law, I chose, I prefer the term like, 502 00:32:52,642 --> 00:32:59,002 you know, industrial, because I grew up in the Midlands in Britain, which is where parts of the industrial revolution started. 503 00:32:59,002 --> 00:33:05,533 And I've always thought about, you know, the industrial revolution and how machines changed everything. 504 00:33:05,533 --> 00:33:10,193 And for me, it it just seemed like so obvious, like this, this will happen to professional services. 505 00:33:10,193 --> 00:33:17,293 And I think when that, you know, going back to the early beginnings of NLP and machine learning, I think people could see making some small inroads, but they could say like, 506 00:33:17,293 --> 00:33:20,601 this is going to spread across the entire sector. 507 00:33:20,601 --> 00:33:22,641 No way, no way, it's not gonna happen. 508 00:33:22,741 --> 00:33:29,861 Now with large language models and language understanding and agentic flows, which are getting better like by the minute, right? 509 00:33:30,241 --> 00:33:32,441 It actually is doable now. 510 00:33:32,701 --> 00:33:33,861 I think it really is doable. 511 00:33:33,861 --> 00:33:35,241 Do we have accuracy problems? 512 00:33:35,241 --> 00:33:37,021 Yes, accuracy is still a problem. 513 00:33:37,021 --> 00:33:44,492 If you automate an inaccurate system, you just mass produce bad goods, right? 514 00:33:44,492 --> 00:33:47,800 You know, like if you've got like a lathe in your garage, 515 00:33:47,800 --> 00:33:52,380 and you're turning out widgets and you've got something wrong with the, with the tool bit. 516 00:33:52,380 --> 00:33:56,454 All you're going to do is turn out a thousand widgets that you can't use. 517 00:33:56,454 --> 00:34:01,887 yeah, but is that, I mean, okay, look at, look at, look at Waymo, right? 518 00:34:01,887 --> 00:34:06,049 If I, if I stick a Waymo on the streets and it's got something wrong with it, you're going to kill someone. 519 00:34:06,049 --> 00:34:08,140 I mean, literally you can kill someone, right? 520 00:34:08,140 --> 00:34:09,871 It's really serious stuff. 521 00:34:09,871 --> 00:34:12,723 They, they've managed to get on top of it. 522 00:34:12,723 --> 00:34:14,207 And now Waymo. 523 00:34:14,207 --> 00:34:17,232 from a day trial or is actually safer than human drivers. 524 00:34:17,232 --> 00:34:22,224 heard that it reduces side impact collisions by like higher than 95%. 525 00:34:22,224 --> 00:34:32,630 So from, you know, going from a hundred thousand a year to 5,000 a year, you know, because people don't look both ways before they cross an intersection. 526 00:34:32,630 --> 00:34:34,391 You see a green light, you go. 527 00:34:34,391 --> 00:34:39,564 These vehicles have the ability to, you know, they have uh adjacent visibility. 528 00:34:39,564 --> 00:34:39,975 So. 529 00:34:39,975 --> 00:34:45,146 I was in one in San Francisco, know, for the Legal Innovators California, and it was absolutely brilliant. 530 00:34:45,146 --> 00:34:46,506 Really fantastic. 531 00:34:46,506 --> 00:34:49,426 Shout out to Todd Smithline for paying for it. 532 00:34:49,426 --> 00:34:50,357 Thank you for that, Todd. 533 00:34:50,357 --> 00:34:51,277 It was so good. 534 00:34:51,277 --> 00:34:51,857 It was so good. 535 00:34:51,857 --> 00:34:54,828 Honestly, I got out of that and I just went, I'm a believer. 536 00:34:54,828 --> 00:34:58,872 It's funny, mean, it was funny having this, having this debate with a bunch of people. 537 00:34:58,872 --> 00:35:08,732 of this thing last week and they were talking about, know, oh my God, you know, it's so good that, um, you know, you have human pilots on a plane, um, because they can rescue you 538 00:35:08,732 --> 00:35:09,692 if anything happens. 539 00:35:09,692 --> 00:35:12,563 And I was just like, well, why don't you just have an extra AI system? 540 00:35:12,563 --> 00:35:17,863 Have two or two, have three AI systems, one after the other, you know, double fail safe. 541 00:35:18,283 --> 00:35:25,994 I mean, it's just, I mean, it's just, it's just humans hanging on to, um, just 542 00:35:25,994 --> 00:35:27,045 you know, things they got used to. 543 00:35:27,045 --> 00:35:38,165 mean, it's like, okay, so every time you go up in an elevator, right, there's a guy who stands in the elevator with a bunch of cushions in case anything goes wrong, right? 544 00:35:38,165 --> 00:35:41,127 It's just like, you just accept it, you go with it, right? 545 00:35:41,127 --> 00:35:45,854 You know, because after, literally, how long, when did they invent elevators? 546 00:35:45,854 --> 00:35:47,232 100 years ago? 547 00:35:48,394 --> 00:35:53,911 Gotta be house now, when they started building skyscrapers, so since the 1920s, right? 548 00:35:53,911 --> 00:35:55,213 But they worked out the kinks. 549 00:35:55,213 --> 00:35:56,614 I mean, and they've done it. 550 00:35:56,614 --> 00:36:01,136 First thing with Waymo, when it started, everyone said, this is ridiculous. 551 00:36:01,136 --> 00:36:02,036 It won't work. 552 00:36:02,036 --> 00:36:03,627 And then it kept on not working. 553 00:36:03,627 --> 00:36:04,991 And then people said, there you go. 554 00:36:04,991 --> 00:36:05,818 You're a bunch of idiots. 555 00:36:05,818 --> 00:36:07,238 You've wasted loads of money. 556 00:36:07,238 --> 00:36:08,379 Ha ha ha. 557 00:36:08,379 --> 00:36:11,440 And Tesla didn't help because they kept on saying, hey, we've got a self-driving car. 558 00:36:11,440 --> 00:36:13,042 And it obviously wasn't a self-driving car. 559 00:36:13,042 --> 00:36:14,744 People were like, Elon, you're not helping. 560 00:36:14,744 --> 00:36:15,584 And... 561 00:36:16,456 --> 00:36:21,860 Now they have, but he could have maybe toned it down a few years ago before he really actually had one. 562 00:36:21,860 --> 00:36:22,550 But they did it. 563 00:36:22,550 --> 00:36:23,388 They actually did it. 564 00:36:23,388 --> 00:36:27,835 I think that is, I don't understand why everyone like parades down every street in America. 565 00:36:27,835 --> 00:36:32,558 Because that is, that for me is as impressive as putting a man on the moon. 566 00:36:32,558 --> 00:36:33,738 It really is. 567 00:36:33,926 --> 00:36:41,926 You had somebody on your podcast recently that threw out a number that I'd love to bounce off of you. 568 00:36:41,926 --> 00:36:54,166 Speaking of automation, know, when shortly after chat GPT-35 was released, Goldman came out and said, like, I think it was 44 % of legal tasks were subject to automation by 569 00:36:54,166 --> 00:36:55,306 artificial intelligence. 570 00:36:55,306 --> 00:37:02,646 And you had somebody on that was talking about by 2027, 80 % of legal tasks could be automated. 571 00:37:02,646 --> 00:37:03,399 And that 572 00:37:03,399 --> 00:37:05,199 that number seems really high. 573 00:37:05,199 --> 00:37:09,499 Actually the 44 % number still seems really high to me. 574 00:37:09,619 --> 00:37:10,991 In 2027 that 575 00:37:10,991 --> 00:37:16,411 well, I mean, we're going to end up in a sort of like big semantic thing around, you know, what is a task? 576 00:37:16,411 --> 00:37:19,271 When, when does the task begin and end and so forth. 577 00:37:19,311 --> 00:37:30,091 I mean, if people want to listen to it, it's Richard, maybe, uh, who's the CEO of Juro, um, uh, legal tech company focused on in-house and it's on my law punks podcast. 578 00:37:30,091 --> 00:37:38,771 If you just type law punks P U N X not K S P U N X into artificial work, you'll find it. 579 00:37:38,791 --> 00:37:39,639 Um, 580 00:37:40,127 --> 00:37:42,807 Yeah, mean, mean, broadly, I agree. 581 00:37:43,047 --> 00:37:46,627 Will more and more and more tasks be automatable? 582 00:37:46,627 --> 00:37:47,467 Yeah. 583 00:37:47,947 --> 00:37:51,107 Secondly, the issue is will they get automated? 584 00:37:51,107 --> 00:37:57,158 Well, given all the things we've just talked about, probably at the point of an economic gun, right? 585 00:37:57,218 --> 00:37:58,578 So not anytime soon. 586 00:37:58,578 --> 00:37:59,938 Now for in-house, it is different. 587 00:37:59,938 --> 00:38:01,338 Obviously he's focused on in-house. 588 00:38:01,338 --> 00:38:06,738 So he's very much focused on what he's seeing day to day and what he sees. 589 00:38:06,738 --> 00:38:08,090 really does believe. 590 00:38:08,090 --> 00:38:10,010 there can be massive change very rapidly. 591 00:38:10,010 --> 00:38:18,361 And I think in-house probably does have, I'd hope so anyway, less of a barrier, but we've been saying that for years, right? 592 00:38:18,361 --> 00:38:19,501 We've been saying that for years. 593 00:38:19,501 --> 00:38:27,392 In-house teams, they obviously are gonna defeat the billable hour because why would they want it? 594 00:38:27,392 --> 00:38:28,252 But they do. 595 00:38:28,252 --> 00:38:35,264 Many, law firms say, actually, we suggested to blah, blah, General Counselor, large corporate, we'll stop doing billable hours if you want. 596 00:38:35,264 --> 00:38:37,464 I'm a general counsel, so no, no, no, please keep doing it. 597 00:38:37,464 --> 00:38:40,655 Makes it much, much easier for me to keep on track, know, keep track of things. 598 00:38:40,655 --> 00:38:42,534 So I don't know. 599 00:38:42,534 --> 00:38:43,775 I think for me it's for human aspect. 600 00:38:43,775 --> 00:38:45,575 I think technically he is right. 601 00:38:45,575 --> 00:38:47,615 I think technically he's absolutely right. 602 00:38:47,615 --> 00:38:49,035 Could it happen? 603 00:38:49,035 --> 00:38:50,835 Yes, probably. 604 00:38:51,175 --> 00:38:57,775 If we keep going the way we're going, will it happen because of human and economic factors and cultural factors? 605 00:38:57,775 --> 00:38:59,595 I think, I hate to it, probably not. 606 00:38:59,595 --> 00:39:02,575 My guess is, I think I might have said before on this program, 607 00:39:02,575 --> 00:39:06,606 my bet for real true transformation isn't about 12 years. 608 00:39:06,606 --> 00:39:07,966 Now you might say, well, that's terrible. 609 00:39:07,966 --> 00:39:10,666 That's a really long way, but I'm talking like real, real transformation. 610 00:39:10,666 --> 00:39:11,646 So is it worth it? 611 00:39:11,646 --> 00:39:17,026 For example, if I said to you, there'll be no oil and gas in 12 years, we'll have a completely green economy. 612 00:39:17,026 --> 00:39:22,886 There will be no more pollution, at least from, you know, the energy sector in 12 years. 613 00:39:22,886 --> 00:39:24,366 You'd like, well, I don't think that's going to happen. 614 00:39:24,366 --> 00:39:26,046 But if it did, that would be amazing. 615 00:39:26,046 --> 00:39:27,606 Completely transform a global economy. 616 00:39:27,606 --> 00:39:30,146 It would have geopolitical consequences. 617 00:39:30,366 --> 00:39:31,022 Right. 618 00:39:31,114 --> 00:39:36,816 like bye-bye Saudi Arabia, know, they're gonna have to move pretty quickly into solar power. 619 00:39:36,816 --> 00:39:38,497 It would have huge, huge effects. 620 00:39:38,497 --> 00:39:50,883 Now, I honestly do believe that in about 12 years that is gonna happen in professional services, knowledge management service, services, because if AI is where it is now, just 621 00:39:50,883 --> 00:39:59,458 less than three years after the launch of Jack GPT, when OpenAI was the only company seriously investing. 622 00:39:59,548 --> 00:40:05,428 in large language models and where and look now, there's more than you can even poke a stick at, right? 623 00:40:05,448 --> 00:40:07,199 And there's more every day. 624 00:40:07,199 --> 00:40:10,059 I mean, and look at the way it's building. 625 00:40:10,259 --> 00:40:12,659 So like look at lovable, right? 626 00:40:12,659 --> 00:40:18,210 You just type into like a prompt box, build me a job website. 627 00:40:18,210 --> 00:40:20,178 20 minutes later, there it is. 628 00:40:20,178 --> 00:40:21,858 It's incredible. 629 00:40:21,858 --> 00:40:24,598 then, you know, and it's something I remember talking to Zach about ages ago. 630 00:40:24,598 --> 00:40:30,278 He was just saying, you know, like when the internet began, people just use it for very basic search. 631 00:40:30,418 --> 00:40:33,318 And then you've got these like layers of an onion. 632 00:40:33,318 --> 00:40:36,318 They accreted, they built up and built up and built up. 633 00:40:36,318 --> 00:40:41,838 now, you know, well now the internet has become a source for LLMs. 634 00:40:41,838 --> 00:40:49,272 And through, and then through this bizarre sort of combination of data sources and LLMs and agents and all this kind of thing, it's all. 635 00:40:49,272 --> 00:40:57,747 merging together into this huge kind of ecosystem that surrounds us now in a way that we couldn't even have imagined when the internet first appeared. 636 00:40:57,788 --> 00:41:04,912 And on top of all of that will be new understandings and new uses for AI, which we haven't even started to grasp yet. 637 00:41:05,312 --> 00:41:13,469 know, that's the thing, you know, if you think about it, right, lovable could not have existed not in the way that it does today or cursor, for example, could not have existed 638 00:41:13,469 --> 00:41:14,209 the way it does today. 639 00:41:14,209 --> 00:41:18,722 If it hadn't been for work, people like Sam Altman and his group, right. 640 00:41:18,812 --> 00:41:21,104 back in the late teens. 641 00:41:21,104 --> 00:41:23,025 But look how fast things have moved. 642 00:41:23,465 --> 00:41:24,807 Where are we gonna be in five years? 643 00:41:24,807 --> 00:41:25,687 And then 10. 644 00:41:25,687 --> 00:41:35,411 And then you've got, you really, really wanna push the boat out, do you think we'll have some usable quantum computing technology in a decade? 645 00:41:35,572 --> 00:41:37,894 Maybe, not guaranteed. 646 00:41:37,894 --> 00:41:38,534 But you know what mean? 647 00:41:38,534 --> 00:41:45,157 It's like, even without that, even without that extra super powerful compute power, look at the chips they're knocking out now. 648 00:41:45,481 --> 00:41:55,201 Look at the investments, look at the gigantic investments from Cisco into OpenAI and the relationship with NVIDIA now, OpenAI is doing a deal with AMD. 649 00:41:55,661 --> 00:41:59,972 And there's a whole bunch of, there's just like billions upon billions flowing into this. 650 00:41:59,972 --> 00:42:00,543 Right. 651 00:42:00,543 --> 00:42:06,563 So yeah, I mean, it's, it's, I really do believe it's going to be a complete sort of tech ecosystem transformation. 652 00:42:06,743 --> 00:42:10,476 Very, very similar to what we saw from the late nineties onwards with the internet. 653 00:42:10,476 --> 00:42:12,598 And we're seeing, we're seeing examples of that. 654 00:42:12,598 --> 00:42:23,146 You know, the, the Garfields in the UK, the Crosby's in New York, you mentioned covenants, the UD is like, these are, this isn't theory anymore. 655 00:42:23,146 --> 00:42:24,879 This is, this is happening. 656 00:42:24,879 --> 00:42:27,501 And, um, and they're delivering real work. 657 00:42:27,501 --> 00:42:30,113 Like I know a little bit about Crosby. 658 00:42:30,113 --> 00:42:35,648 they had a Sequoia is one of their capital partners and they have a podcast called training data. 659 00:42:35,648 --> 00:42:38,121 And they had the founders of Crosby on. 660 00:42:38,121 --> 00:42:41,792 And the way, what they're doing essentially is like master services agreement. 661 00:42:41,792 --> 00:42:49,157 have a very narrow niche master services agreements for big, you know, like clay and stripe. 662 00:42:49,157 --> 00:42:54,000 And, they are delivering with a very small number of attorneys. 663 00:42:54,020 --> 00:42:56,211 So, I mean, this is, this is happening now. 664 00:42:56,211 --> 00:42:57,811 This isn't theory. 665 00:42:57,811 --> 00:42:58,422 So 666 00:42:58,615 --> 00:43:00,615 But again, it's scale, isn't it? 667 00:43:00,615 --> 00:43:00,995 It's scale. 668 00:43:00,995 --> 00:43:02,215 mean, this is the thing that look at Lovable. 669 00:43:02,215 --> 00:43:07,655 Lovable grew from like, you know, like $1 to tens of millions of dollars in about 12 months. 670 00:43:07,655 --> 00:43:12,455 It was instantly usable and there was no barrier, right? 671 00:43:12,475 --> 00:43:18,586 With legal tech has always suffered from, like I going back to the same point, it's got structural barriers, right? 672 00:43:18,586 --> 00:43:19,866 I Covenant is a great model. 673 00:43:19,866 --> 00:43:20,726 Crosby is a great model. 674 00:43:20,726 --> 00:43:21,797 Garfield is a great model. 675 00:43:21,797 --> 00:43:23,655 Why aren't there 200 of these? 676 00:43:23,655 --> 00:43:26,048 or why doesn't Crosby 677 00:43:30,644 --> 00:43:33,119 mean, they've only been in business a year, right? 678 00:43:33,119 --> 00:43:35,016 So it's, they're, they're brand new. 679 00:43:35,016 --> 00:43:37,257 I suppose it was really interesting calculation. 680 00:43:37,257 --> 00:43:45,197 I'm have to do it at some point, which is, I mean, if you look at the total revenue of all the legal tech companies on the planet, right? 681 00:43:45,217 --> 00:43:54,297 And this is not a zero sum type calculations, which makes it very, very complicated, but start off with what is the total revenue of all the legal tech companies in the world, 682 00:43:54,297 --> 00:43:54,917 right? 683 00:43:54,917 --> 00:43:57,277 And break it down into what they do. 684 00:43:57,477 --> 00:44:03,093 So if we look at sort of more productivity side, so let's put aside all project management, billing. 685 00:44:03,093 --> 00:44:10,156 all that kind of stuff, time tracking, take that out and just look at them all sort of productivity tools and look at the revenue that comes into them. 686 00:44:10,156 --> 00:44:10,636 Right. 687 00:44:10,636 --> 00:44:17,709 Now that is a measure to some degree of how far business of law has integrated these technologies. 688 00:44:17,709 --> 00:44:18,000 Right. 689 00:44:18,000 --> 00:44:22,942 And then once you've done that, and it's very, very complicated to figure that out, maybe McKinsey can do it. 690 00:44:22,942 --> 00:44:28,854 Then look at what a percentage that is of the total legal tech market. 691 00:44:29,296 --> 00:44:33,016 sorry, legal market, sorry, it's very late in the day here. 692 00:44:33,016 --> 00:44:38,656 What percentage is of the total revenue of the whole legal market, right? 693 00:44:38,656 --> 00:44:40,496 Now, these are not exact fractions, right? 694 00:44:40,496 --> 00:44:51,316 You can't say, you know, if someone spends 20 million on X legal tech company over a year, or a bunch of different people spend 20 million, doesn't mean that that has been extracted 695 00:44:51,316 --> 00:44:56,696 from the legal market, because it may actually be empowering the legal market to do more. 696 00:44:56,696 --> 00:44:57,646 The same way, 697 00:44:57,646 --> 00:45:04,437 that law firms spent a fortune on Microsoft Word and email, and that enabled them to make a hell of a lot more money, right? 698 00:45:04,437 --> 00:45:08,299 So I think I'd love to see some financial analysis around that. 699 00:45:08,299 --> 00:45:14,261 So like money that's coming out of a legal market into legal tech that then in turn drives productivity. 700 00:45:14,261 --> 00:45:22,563 And then how much of a sort of force multiplier effect that has on those lawyers. 701 00:45:22,563 --> 00:45:24,823 which enables the market to grow even larger. 702 00:45:24,823 --> 00:45:28,923 So it becomes like this sort of supporting cycle. 703 00:45:28,923 --> 00:45:30,943 And we just don't really have the data. 704 00:45:30,943 --> 00:45:32,583 We just don't have the data. 705 00:45:32,583 --> 00:45:41,294 And there's a generally very, you know, it's classic reductive view is, you if you spend a dollar on an AI tool, you're taking a dollar out of a lawyer's pocket, right? 706 00:45:41,294 --> 00:45:43,294 Which I don't think really follows at all. 707 00:45:43,294 --> 00:45:44,994 It could, it could in certain streams. 708 00:45:44,994 --> 00:45:49,485 And that's why it gets even more complicated because there are certain streams of work, which I think will go away. 709 00:45:49,485 --> 00:45:53,209 right, they will get automated out of existence in terms of the lawyers putting their fingers on it. 710 00:45:53,209 --> 00:45:55,391 It will become commoditized, right? 711 00:45:55,391 --> 00:45:56,432 That will happen. 712 00:45:56,432 --> 00:46:02,457 But there'll be other areas of work where the AI will actually massively increase what lawyers can do. 713 00:46:02,457 --> 00:46:06,319 And they may end up becoming way, way, way more wealthy than they are today. 714 00:46:06,320 --> 00:46:09,913 And it's quite difficult because you both of these, both of these pictures are true. 715 00:46:09,913 --> 00:46:14,537 And I think that's one of the things that's and people like clear pictures, right? 716 00:46:14,537 --> 00:46:16,548 know, AI will destroy all lawyers. 717 00:46:16,548 --> 00:46:17,533 No, that's not true. 718 00:46:17,533 --> 00:46:19,773 Okay, so AI is irrelevant. 719 00:46:20,053 --> 00:46:21,333 No, that's not true. 720 00:46:21,333 --> 00:46:30,893 So it's kind of like, you know, it's like, how can AI actually eat up probably what will be in 10 years, a significant chunk of what is today of a legal market. 721 00:46:31,373 --> 00:46:37,964 And yet at the end of the day, Christmas 3035, is that right? 722 00:46:37,964 --> 00:46:40,144 Yeah, Christmas 35, right? 723 00:46:40,624 --> 00:46:43,364 There'll be a lot of lawyers who are saying, my God, thank God for AI. 724 00:46:43,364 --> 00:46:44,616 I'm so much wealthier now. 725 00:46:44,616 --> 00:46:50,499 And I don't want to jump out of a window because that's really the mindset of a lot of lawyers today. 726 00:46:50,499 --> 00:46:56,425 It can be a very brutal profession and coming up the ranks is sometimes a painful process. 727 00:46:56,425 --> 00:46:57,699 uh 728 00:46:57,699 --> 00:47:06,279 that's, that's one of the things I mean, I I gave a talk in Oxford the other day at a, and I don't know if it me something in public meeting somebody called Gunnar Cook, which is a 729 00:47:06,279 --> 00:47:07,859 distributed law firm. 730 00:47:07,859 --> 00:47:08,099 Right. 731 00:47:08,099 --> 00:47:11,470 So they're all individuals, but they work under the umbrella of Gunnar Cook. 732 00:47:11,470 --> 00:47:12,210 Right. 733 00:47:12,330 --> 00:47:15,830 And they're using AI systems now, which is fantastic. 734 00:47:15,850 --> 00:47:24,070 And, you know, we were talking about the fact that, you know, it's, it's just, it's an, it's all, it's an all, you know, it's all positive for them because they don't have 735 00:47:24,070 --> 00:47:24,900 leverage. 736 00:47:24,900 --> 00:47:28,521 It's mostly made up of experienced lawyers, right? 737 00:47:28,521 --> 00:47:31,553 They don't have this huge leverage associate group beneath them. 738 00:47:31,553 --> 00:47:34,885 And having AI greatly increases what they can do. 739 00:47:34,885 --> 00:47:41,817 And it does work that quite often if a client comes to an individual in a group like this, one of these distributed law firms, they're not coming to them because they know they've 740 00:47:41,817 --> 00:47:44,288 got 25 associates working in the basement. 741 00:47:44,288 --> 00:47:49,820 They come to them because they know that person has real expertise and will really add value on a very human level. 742 00:47:49,820 --> 00:47:51,440 But of course, 743 00:47:51,766 --> 00:47:55,048 leverage based work will creep in. 744 00:47:55,048 --> 00:47:56,630 But now they can use AI to do that. 745 00:47:56,630 --> 00:47:58,872 So I mean, it's an absolute win-win for them. 746 00:47:58,872 --> 00:48:10,549 And again, for these hybrid AI companies and the ALSPs and all of these, it's, mean, this is the thing that is kind of tantalizing is that we really are starting to see more and 747 00:48:10,549 --> 00:48:14,040 more examples of organizations using AI. 748 00:48:14,356 --> 00:48:15,946 and it's just all positive. 749 00:48:15,946 --> 00:48:22,786 And simultaneously, we're seeing both some law firms and some in-house legal teams kind of gritting their teeth and going, oh, I'm not sure. 750 00:48:22,786 --> 00:48:23,826 I'm not sure about this. 751 00:48:23,826 --> 00:48:24,946 I'm not sure about this. 752 00:48:24,946 --> 00:48:26,446 Is this a good idea? 753 00:48:26,686 --> 00:48:28,826 And you're just like... 754 00:48:28,857 --> 00:48:32,459 Yeah, there's still a lot of pearl clutching going on. 755 00:48:32,459 --> 00:48:39,384 Well, we're almost out of time, but this has been a great conversation as it always is every time we have you on. 756 00:48:39,384 --> 00:48:46,589 Before we hop off, how do people find out more about legal innovators and the writing that you do? 757 00:48:46,936 --> 00:48:50,707 Go to chat GPT and I'm not even kidding. 758 00:48:50,707 --> 00:48:56,090 I get a lot of traffic from, from LLMs now uh perplexity and others, right? 759 00:48:56,090 --> 00:48:59,792 Go to chat GPT and ask it about artificial lawyer. 760 00:48:59,792 --> 00:49:02,592 And hopefully it'll say something nice about me and provide a link. 761 00:49:02,592 --> 00:49:08,495 But if you don't want to do that, just go into Google and type artificial lawyer.com and you'll, you'll find me. 762 00:49:08,495 --> 00:49:10,447 But yeah, on the New York event. 763 00:49:10,447 --> 00:49:20,886 if I imagine some of your listeners are in the US, the New York event, Legal Innovators New York, the website is literally legalinnovatorsnewyork.com and it will be on the 19th 764 00:49:20,886 --> 00:49:22,878 and 20th of November. 765 00:49:22,878 --> 00:49:31,330 Day one is law firms, day two is in-house and obviously we'd like to see everybody on both days but if you want to focus on one or the other then you can take your pick and it will 766 00:49:31,330 --> 00:49:32,630 be in Midtown. 767 00:49:33,041 --> 00:49:33,801 Okay. 768 00:49:34,021 --> 00:49:35,462 That's great to hear. 769 00:49:35,462 --> 00:49:38,545 And I'm a big fan of the artificial lawyer site. 770 00:49:38,545 --> 00:49:40,226 I'm very selective. 771 00:49:40,226 --> 00:49:42,316 I get a few newsletters that I read every day. 772 00:49:42,316 --> 00:49:43,207 Yours is one of them. 773 00:49:43,207 --> 00:49:46,188 So um I really appreciate your time. 774 00:49:46,188 --> 00:49:50,291 Best of luck with the conference and I'm sure we'll bump into each other sometime real soon. 775 00:49:50,291 --> 00:49:51,325 I hope so, definitely. 776 00:49:51,325 --> 00:49:51,872 Thank you. 777 00:49:51,872 --> 00:49:52,737 All right, take care. 778 00:49:52,737 --> 00:49:54,185 Bye bye. -->