E41 - The Future of AI is People: Legal Quants Founders Jamie Tso & Raymond Sun

I'm Zach Abramowitz and I am legally disrupted. What if AI doesn't just replace your old software? What if AI changes our very conception of what it means to be a software or a tech company? And what if in this new version, it's not the tech but actually the people that make the company valuable? On today's episode, episode 41 of Zach Abramowitz is legally disrupted. I'm talking with Jamie Tso and Raymond Sun. They are the founders of the legalquence community and they are the embodiment of the 2026 AI zeitgeist. Now recently Harvey, who I was writing about throughout all of 2025, announced an $11 billion valuation on a new raise. And for someone who has followed legal tech for as long as I have in Harvey in particular, this should have been like really key important news. But I have to tell you something. I kind of yawned because right now Jamie Ray and the legalquence community have my undivided attention. Now you have to hear about these guys backgrounds. Long before they were pioneering legal AI, they were both content creators who had some pretty significant viral experiences. So Jamie ran a fitness transformation channel on YouTube that racked up two million views. And during COVID, Ray built a machine learning app that compares your dance moves to professional K-pop dancers. And that went completely viral across South Korea, Japan, and China. By the way, both of them are corporate lawyers at big time law firms. But that content creator instinct perfectly set the stage for how they built this current movement. When Jamie started sharing posts on LinkedIn about vibe coding, different apps that he was building and different products that he was replicating, he just wanted to find their lawyers who thought like him, but very, very soon other lawyers started tagging him on LinkedIn, showing that they were inspired by what he was building and showing what they had built. So what started this this tiny five person WhatsApp group is organically exploded into a highly active grassroots community of a hundred lawyer developers, vibe lawyers, if you will, who are sharing insights every single day. Now at Killer Whale Strategies, we are keeping a very close eye on these two because they represent the real shift happening in our industry. The next big thing may not just be buying big vendor solutions, even ones that are AI first. What might be coming next is the vibe coding playbook. Jamie and Ray are showing us a future where lawyers take control and not just identify the pain point, but actually build their own bespoke just in time apps to solve immediate problems. They have an incredibly rare skill set that combines tech, law, and content. And we think that they are writing the blueprint for the lawyer developer error. I could not be more excited to share this conversation with you by the way. I should tell you in advance, we went for nearly an hour and a half. I am very excited about legal quants about what Ray and Jamie are building. So let's learn how to build. Let's hear the vision. Let's hear about the community. Let's get disrupted. So I was just saying to you guys offline how excited I am for this podcast. I've been doing this exercise recently with our team internally and then sharing a lot of this with our clients where I talk about the 2025 to 2026 vibe shift or is the way I've put it like the zeitgeist. What is the AI zeitgeist right now? And I feel like you two are like the 2026 AI zeitgeist right now within legal whereas in the last couple of years, Harvey and then Harvey Lagora and they've kind of like owned the mind share. And by the way, there are other big legal AI companies and other spaces. But if you were sort of if you travel in these circles, if you're kind of like following LinkedIn, if you're looking at Reddit threads, a lot of the energy was on these hyper scaling startups and just the other day, Harvey announced they've raised now at an 11 billion dollar valuation. And for someone like me who's been following legal AI and legal tech for many years, that should have been amazing like a legal profession, legal industry like Deca Corn. This is huge. And yet I kind of yawned because the two of you and legal quants right now is really the company that I'm following and I have been writing extensively about Harvey. I wrote about it so much in 2025 because it was such a big deal. But I think there's so much about what you guys are doing that is capturing where AI is moving right now and one of the big challenges and I want to start this off with the two of you. One of the big challenges in AI and I think this is getting to be more amplified recently is that AI is so fast moving that you know that sort of keeping up with keeping up with all the news, all the movements, all the advances is draining and it sometimes can cause a bit of anxiety whereas I think in people you know in 2025 there was a sense of oh I finally have got the hang of this. 2026 is like I'm feeling lost and behind Jamie why don't you start start us off with the reaction I want to get get raise raise reaction to that note. Mostly agree I feel that I have never felt so behind in AI and it feels like for the past few months we've seen Claude Opus 4.6 you know actually Gemini are free first is basically everyone's calling it AGI and then Claude Opus 4.6 and Claude code well where everyone just says well as they'll just say go wrong name to call the product because it's basically an operating system and a lot of non-technical people are using it and now you're with Claude Opus. It's just impossible to keep up to be frank I there's a lot of reading that I need to do and there's but I feel like I can't not keep up with the latest technology it's so fast. Yeah I'm good and you know I know a few years ago when people talked about the growth of AI will be exponential I knew what that meant rationally but I never really felt you know what then I feel like emotionally maybe 2020-24 themes were still manageable but you know since 25 like I think since deep seek and then that really kick startled a huge drive innovation across both sides of the world like every two weeks there's something that's come out and each week replaces like the development of the previous week and so we keep seeing that all the time so yeah it's definitely our challenge to try to keep up with everything but it's very exciting as well very stimulating. Yeah and I've got this uh this slide up on on the screen I think it's interesting to note like Andre Carpathy himself like a guy who coined the term vibe coding is admitting recently to feeling very much behind and to your point um Ray that there there is this this sense not only that like things are are moving moving fast but like we thought we could sort of like grasp exponential progress and what that meant like for me my feeling was just use that that point in 2020 late 2023 I began investing in Nvidia stock and the reason I started doing it then is because I felt like wait a second everyone else is saying it's too late to invest in Nvidia stock but if this is truly exponential progress there's no way that anyone could possibly have baked in what that means meaning the the market can't understand exponential progress because it's something beyond human psychology and in the human mind and on that like what I've told people and and ever and lawyers from the get go is is not to be so concerned with like which AI tool that you're using but just to be using AI and learning about AI and and I know that's like simple so you can't like charge a whole lot of money for giving that advice but I feel that one of the things that's happened and where you're starting to have this feeling of like feeling left behind at various levels is the value of early adoption is compounding not just in your capabilities and knowing like what prompts to push but just what your brain is ready to wrap itself around and be like oh okay like that's that's what's possible and and I think that if you don't sort of do that process early your brain will take a long time to adjust and I think this is like this you know there's this gap right now if you go and like you know since you know social or read it you'll see a lot of people who are saying like hey you know for everyone in the AI conversation just understand that like 99 percent of the population like have not heard of cursor and clawed code but there is like this this real feeling of like the AI halves and have nots what's your reaction yeah I think that the frontier technology is you know what we see or work perceive as a frontier technology is probably not what people inside us perceive because if we even think about what and the you know open AI and frock it internally what they see will probably be like years ahead of what we are actually seeing in the market and I think that is what was what scares me and for lawyers I feel that our I would In the generation of lawyers, I think I would encourage lawyers to try out whatever funds are a technology that they can get their hands on. I think that would inform them what the frontier looks like and how they can apply that on their day-to-day job. I feel there is the aspects of building tools for lawyers, but there is also another aspect where lawyers not only build tools for themselves, but also experiment with the possibilities of AI and knowing what is possible. That has been my kind of mission for the past few months. That's why I kind of showed how good Gemini Free is, how good clot opens is, and with whatever open source technology that we can stitch together, we can already achieve. I think that in itself is the kind of AI literacy that I encourage my fellow lawyers to counterfeit. Before you jump in on this, I wanted to highlight the very interesting podcast with the founder of Base 44, Maors Shalomah. I've been saying for the last year that that was the most important legal AI deal to watch. Of course, it's not a legal tech company. I don't think of Base 44 as legal technology, and I don't think they're even going to get into that space. I thought it was more important last year that people understand vibe coding, the concept of being lean, not needing a huge team to accomplish what you would have once needed a huge team in order to accomplish. There's so many takeaways from that deal, but he has a podcast he went on Harry Steppings podcast, and he was talking about the importance of showing users the magic. Actually, this may have been a podcast he did with Lenny Rashitsky even earlier, but one of the things he talked about in that podcast was he said, "When it came to AI, he'd realize that more important than building the most rigorous product was a product that had very, very little fiction between users seeing what AI could accomplish for them that they simply would never have thought possible." I think that there's a kind of anxiety, pre-AI, because you're thinking like, "Oh my gosh, am I going to waste my time with one of these tools and it doesn't even work?" Or you think to yourself, "Wait, I'm going to see something that's going to be terribly frightening, and then I'm going to realize I have to start making some life changes because, oh my gosh, look what's possible." Or third, you think, "I'm going to try this. I'm going to realize that it's amazing, but I'm also going to say I am too old to learn this." Meaning so you see the frightening thing, but you're not able to. I feel like that moment pre-AI is filled with anxiety, and I think what you've been doing on LinkedIn, and now with legalquence, is showing lawyers like, "This is not too difficult for you. This is not above you. You are capable of this." Not just the two of you creating these applications, but now inspiring other lawyers to say, "Hey, I saw what Jamie did, and now I'm going to. " Here's what I've created, and I think that that showing the magic has been something that you guys have been really amazing and impressive with. Ray, jump in the knowledge, Jamie Racktall, this as well. Yeah, definitely. Residinate with all of the above, and fundamentally it's a skill set shift, but deep down is a mindset shift. What AI brings, and consequential technologies bring afterward, is really forcing all of us to really rethink about the processing of work, the delivery of work, and the interfacing of work. Also, rethinking about what does legal actually mean now? What is legal service in the next few years? Fundamentally, lawyers have been the bridge between the law and broader society. We be the translators, right? And to fundamentally aid information-heavy profession, it's all based on words, and ultimately it's data, right? At the end of the day, we're also in a data space, because we're dealing with legal language, converting raw language, raw legal language into language that is understandable and useful to the client. And for the past 100 years, it has mostly been a manual effort of translation, right? Now with these tools, we can really take our game to a step up. How do we not only use AI to help make our existing processes efficient, but also get ourselves out of the box and really redefine how legal is delivered to the world society and make even more useful to clients, right? And so having legal requirements, for example, be more directly operationalizable and more embedded within what a product or service that the clients are trying to get your advice on. So really thinking about delivery form and I feel like that will be the next phase of thinking once people get used to the whole, yeah, this is the new way of doing my work, they'll be taking that step up, right? But I think this is where I feel like AI is much more challenging than previous like SaaS technologies, like a matter management system or a practice management system. Those didn't challenge fundamental beliefs about the world, about yourself, about the profession. And I've said this from early on, part of the reason lawyers don't like the conversations about AI and what I think a lot of people don't like is that it makes them reevaluate like what it means to be human, right? So in other words, it was, is what separates me from a computer, the ability to like reason? Maybe not. Now maybe you say, well, I, the computer doesn't have a motion and I do. This is, this is true, but again, it's, it's an uncomfortable conversation that Salesforce never forced anyone to have, right? It's work day, never forced anyone to kind of grapple with these deep philosophical questions. And I think that, you know, what does it mean to be a lawyer? It is something that we need to like now wrap our heads around and that's sort of an uncomfortable staring across the abyss, Jamie. Yeah, I totally feel you. I think AI is fundamentally a complicated technology. It's, I think of it almost like a very dangerous and fast car. It's off, you know, what is AI in, you know, in today's term, if we strip away all the UI, you express basically the model weights, right? And what do you do with that model? It is complicated. And I feel in the past few years, there are all these legal AI tools. There are bills just to create that kind of abstraction for lawyers so that the lawyers they can basically use these tools more easily and the harness the power of AI indirectly. But then there are a group, there's a group of lawyers that are willing to go one level deeper, right? To see AI in itself as the model and think, you know, what can I do with it? And this brings us to the concept control, right? If I can configure control and design my own AI system, how does it look like? And it goes back to, you know, Ray mentioned the deep seek moment, right? That's the moment when AI and LLM became a utility. And if AI becomes something that's free, it's complicated to use, then we can think of the people using it as basically the F1 drivers, you know, you're driving a fast car. How fast can you drive without crashing? I think there's a question. I think a lot of the AI tools are created. And technology. It's like Toyota and Honda. It's safe, you know, but then you don't actually need to know what it works. But what's the problem is that when you've, when you're providing your, and I've talked about this before, a lot of companies want to provide an AI with like guard rails and all kinds of stuff around it to the point where it basically neuters the AI and then people like, well, if you're going to give me like lame AI at work, I'm going to use my own AI. And I think that in 2025 that might have been like, if you're going to give me some like, you know, pair down branded version of, you know, let's just see like, you know, the Walmart, you know, Marty, your Walmart AI chatbot. And I'm just using like, and Walmart's awesome. They're over a trillion dollars now. So they're great company. But my point is if they're going to give like, if you're getting like the company version, and it can't do what your chat GPT at home can do, then like, forget about it. But I think there's going to be an interesting conversation that I think is already materializing inside of a lot of the AMLA 200, the global 500 firms, which is, wait, do I need to use this like legal AI specific tool that you guys purchased for us? Because I've been able to code something internally that really works much more how I work. And I can get better value out of this than I can of the firm approved tool. And I think that that is, that's kind of the new 2026 conversation. Curious like, like, and again, I, what you're, when you talk about like configurability, to me, like, I've been hesitant to tell firms to like rethink the build by calculus because I think in the past a lot of firms would build things that they absolutely should have bought. And I would like face palm a little bit over it and think like, oh gosh, this is another kind of like cringy law firm developed tech. And I can't think that way anymore at all because I don't feel like we're talking about products that are being developed by a firms like IT group that was like, you know, new some basic sequel. Like I think we're talking about like the actual users who can build this for themselves tailored. And if their workflow changes, they can literally throw it out and start from scratch and rebuild it and not always be in the business of configuring someone else's, you know, one to many products that may not work with their user habits. Yeah, precisely. And which is why I think when we had the vibe coding movement, but first I've five coding movement, it within the legal community. I think a lot of people are saying, well, the tools themselves are not production ready, you know, the apps are not secure. And I think that is missing the point. Of course, these are important problems. But I think that would be solved down the road by the by by the cloud providers. I think the more important point here is that we're looking at the skills of these lawyers rather than apps that they built. And you imagine how skillful a lawyer need to be to be building their own tools to be solving their own problems for the technology that way. And it's an extreme form of AI literacy. And I think that that should be their focus. And if we and the opposite, you know, the opposite would be a, you know, for a lawyer to just use something off the shelf, fully designed to do to do to let it work for every, you know, all the lawyers, you know, the learning curve is also pretty flat. But the downside for that is it's a black box. You can't really know if things happen. If if the AI product yields some bad results, there's no way for you to go back to fix it. You don't even know what went wrong. And you can't, you don't know what went wrong. It's really hard to improve it. And then you better results out of it. But then with five coding and with people finally understanding that using AI in a raw and fundamental way, it's a skill in itself. Then you kind of have that quick iterates of loop where you can test iterate and you know, design your own workflows, apps and skills. I think that is the, I think it's the whole point. Yeah. And I've heard you talk about this before. And I like to hear like Ray, Ray addressed this first and come back to you, Jamie. But I've heard you address before, you know, a lot of the questions that will come up is like, well, you can vibe code that. But is it really scalable and is it really secure? And is it, and I, what I've heard you say, and I think this is spot on, which is well, you don't need to be scalable necessarily in the way you would if you're building a one to one to many companies or to many customer style product. Like the reason they have scalability issues is because they're trying to solve for many, many customers, right? You can be what you just discussed right now is that you can really like the configurability value curve is off the charts for vibe coding. Security, as you described, seems to be better again. If you're not working, like the less third parties that you're working with and the more directly you're working with the models, I mean, the transform model itself is a black box. So like, no one kind of really knows. But is many, like as few black boxes as possible, right? And I think that so like there's, what I guess I'm trying to say is it doesn't feel like it's just a question of costs. Like, well, does it cost more to vibe code this or does it cost more to buy, you know, insert co-counsel, or what are any of these? The real question is like, which is actually more in line with firms goals? And again, I've, from someone who has discouraged law firms in the past to build a product that they could buy, I find myself. Vibeshifting, I guess. Ray, what's your thoughts? Yeah. I think I'll answer this in three quick points. I think first one, we've really seen a new layer of software which I like to call bespoke ware or customware, where it's different from, you know, SAS, where, as you mentioned, one to many, one size fits all. So, we're not at this point in time where you can create a very hyper personalized software and it can just be for one time use. Maybe there's a task, there you touch one time, you build this app one time and then you complete it and that's it. You don't even touch that app again. That's, that's the only one. Let me give you an example. Let me give you a very tangible example. You're starting a new M&A deal. We would typically say, okay, work with intro links to get the data room set up or work with Merrill to get the data room set up. How about spin up a data room? Here's the kind of deal we're working on. Here's the number, here are the parties that are going to be involved. Here's what each one is going to be looking at. I mean, tailored, individualized on the fly, per the contours of your deal, per the security requirements of all the parties involved and you're never going to use it again. Yeah, that's right. This, this, this, this, this emergence of custom where leads to my second point, which is around, you know, the build and buy equation that you talked about, Zach. I feel like that has now become more than nuanced question now, which is more like capability versus interface. A lot of, a lot of the existing tech vendors who, whose entire business mode is built upon having their interface being fulfill purpose for the customer. But there's no one better to understand the cost and problem than the customer themselves. If the customer can solve their own problem through these tools, they solve the interface problem that really is a challenge for a lot of the interface specific vendors. So a lot of this stuff is now, love the modes that we see is now shifting towards capability. The ones who are at the forefront of this are thinking, if in three to five years time we know that our own customers are building the most bespoke interfaces and we can't compete with them on that front. What about the capability that we can hone in to make sure that even if the customer makes the best custom app, they still need to rely on us for the underlying capability, whether that be the underlying reasoning model or the whatever infrastructure that they're going to provide. That's where that question comes in. So I feel like a lot of the modes will now become an infrastructural battle, less so an app layer battle, which would be a bit more, it's a bit less obvious and less fancy and less sexy and less visible to the everyday consumer, but there's a huge battle going on at the infrastructure layer. So I actually, by the way, I just invested in a company called Sandstone and Sandstone's entire thesis is exactly that is to be an infrastructure that connects to different data sources within a legal department that allows the legal department to get information from the business more easily than they would have up until now. So it's along that those lines, but I want to get back to like the mode question in a second, but I want to get Jamie's reaction first, unless you had another point to add their rate, because I know it would be like all over the place. I'll just quickly wrap up like the infrastructure point there, then we'll go back to this. So the infrastructure ecosystem, these are the two big pillars and I'm sure Jamie can speak more to that as we talk more about legal funds, but the third point is really just going back to the whole mindset shift. Lawyers are sort of trained to sort of work with the cards in their hand, right? You're given the facts, assess the risk, assess the liabilities, make the most of your facts, provide an honest advice. It's a type of training where you're trained to play within the box. You can maybe touch the borders, but you're fundamentally trained to be within the box. What I feel like and what Jamie has started with, the vibe coding sort of training is that it's inspired lawyers to actually break out of that box, right? You don't have to feel like you're limited by your tools. Actually, you can train new tools to solve your problem and that one switch, which I feel like it's such a deep emotional impact on so many of us. And that's what I want to highlight. That's the third point. So the surface area of legal practice, I mean, there's new surface area. For sure. That's the best analogy. So Jamie, I want you to comment on this, but based on what race that I think you're going to steal my point, but I'm looking forward to it. I think the future is where one where the clients put interface with technology, maybe, and to have the lawyers become the API. That's one way to think about it. The other way to think about it is, well, this software is a product, but then the lawyers become the interface. So to my mind, lawyering is a people business. There's a lot of soft skill. If you're doing a $5 billion merger, then you better talk to a lawyer, a human, because I think that's how. So we do trust. Everything could be AI enabled. But then would you do it the other way around as in, you just use whatever AI tool that was given to you by the lawyer to do the $5 billion merger? I don't think so. So I think that is-- I think in the future, a lot of things would be AI enabled at a fundamental layer. But that's still in the face. Perhaps more human. Perhaps more human. Well, to be frank, a lot of people are saying, well, do we have AI to AI negotiations? I think I was talking to a law firm partner the other day about whether there will be just AI to AI negotiations. And what if we have these negotiation calls where we can just spin up to 11 laps AI agents and negotiate with each another? That's not just not productive, right? That's just more AI slopes. So I do feel that maybe the human layer would still be there. So I have two thoughts about the human layer, but I want to first tell you what I thought you were going to say. But I think this gets to the legal quants point that Ray was mentioning before. But I have a second human factor that I wanted to discuss. When we were talking yesterday internally and preparing for this podcast, one of the things I was discussing with my team was, what is the value of a software company today? Now there is obviously a layer of software being built that's enabling all the builds, right? So the frontier models themselves, everyone who's taken the frontier model and tried to adapt it for coding and for building, this is obviously there's value in that. Although it's the same time, the value unlock has never been bigger. The ability to create that value unlock has never been easier. So where is the value potentially in a company that sells one to many that sells that scales? And I started thinking about this and it occurred to me, well, the value would be in the community of users around it. Meaning a software company is actually more human now because the software company gets-- like how would it be that the customers of the product would increase the value? And I think that when you develop a community around a software product, I think that's going to have extraordinary value moving forward. And now that I've put those goggles on, I went back and I started looking at legalquence and I'm sitting here going like, OK, now I think I'm really getting this because I think they've seen around a corner here. I think they understand that the value is in the community, not in the software, and that the future of valuable software ironically is in community. Reactions? 100%. Software cost is going to zero. But talents aren't, right? The whole mission of legalquence is to highlight the talents. The apps that they create is an extension of those talents. The apps, they build-- it's the manifestation of the new skills that they acquired the past two years. Today, it could be five coding, but three months from now, it could be class skills. You're from now, it could be designing your own agent. I mean, designing your own agent. I mean, that should happen a lot sooner. But then the thinking here is that how do we showcase these talents? How do we upskill them? How do we date the stuff for them? How do we find them? If we're talking about the future of law, I think personally that a lawyer needs to be excellent at tech to be excellent at law. So they need to be able to understand the technology, able to drive fast using that same technology, and to become a former loan driver. I think the community here that we're cultivating is basically one that is where everyone is just so excited. We can talk about that later about the things that we can do with that. But how we started is simply that when I started a vibe coding, I started seeing that a lot of people lay detect me in their post to show their builds. I've been yearning to connect with different lawyers with the same mindset, because I thought I was alone. Then I created this WhatsApp group. It was started with just five people. Then a slowly grew. Now is around 100 people. We have hundreds of messages per day, just discussing everything about applying front-sure technology, front-sure AI, or a day-to-day work. And it's just so fascinating to see that I'm learning so much from the group as well. People are talking about how to use the latest model, the latest technology, when Claude legal plug-in drops. Someone has already tested it and given them their feedback. So it was extremely active and really, the contributions that they made really blew my mind. Ray. Yeah, I just wanted to add that in this spot on with the community angle. And I just want to highlight how special this community is, because we're talking about a group of people with a combined skillset of law and tech. And when we say law and tech, I like out of all the possible best of both skillsets. I feel like this combination is the most lethal and dangerous, because we're talking about technology, which is like the language of production and change, combined with law, which is the language of power, combined those two together. You have like, and legal ones captures the global community of these exactly this particular skill set. You have people who not just know how to build, they can know how to like, bring it to scale, they know how to navigate through the regulatory and legal channels. They can do it all by themselves. And that's like, that's huge. Like that, that the caliber of talent in legalquence, it just blows my mind to be honest. Like, it's such a special community. Yeah. So the other way I was thinking about AI being more human and the, and the, and the interface shifting over to the human is, I think that law firms might be missing an opportunity. And here, this is like free advice. You don't even have to pay McKenzie. But the problem with the problem with like big law firms is like, like they want you to like charge them a lot of money for the advice so that we, we brought in McKenzie. But OK, this is free advice. Go do this. Have all this space in the firm. And they've been having this major challenge. How do we get people back into the office? And I think here's how you get all of your lawyers to want to be back in the office all the time. You build your office to attract clients. You've got all this space. Many law firms have leases that they're locked into. They, oh, if we knew about remote work, we never would have gotten them. OK, now ask yourself, what experiences could we create, whether it's boardroom experiences, whether it is clad-- whatever it is, OK? And I'm not saying it's like, well, defined. You don't have to think about it a little bit. But ask yourself this basic question. What would make our clients want to come into our office and see massive value in doing that regularly? And how could we create the office experience? Because you know what will happen when clients start coming to the office? All your lawyers will want to be in the office, too, because all your lawyers want to be around the clients. And I think that this AI revolution should leave a lot of firms rethinking and saying, what did it once upon a time mean to be an attorney before we got all into the documents, all into the digital? What was that actual human-to-human experience like? And how could we build on that? I'm not saying to go back to the early 1900s. But I think sometimes getting in touch with roots really, really critical for thinking in advance. And I've heard of firms that do the sort of like concierge service when you walk into their building. I think I've heard one of the UK firms is really good about this. I forgot which one. But I think that firms need to start thinking along those lines of like, OK, assume like you said, software development goes to zero. I think we probably, at some point, say document creation probably goes to zero. OK. So then like, where would the value be? Where would the value be beforehand? And I think that those would take this question seriously, who like think about it in the shower, who think about it, before they go to sleep, who wake up in the morning thinking about this challenge. Those are the ones that are going to solve it. But I think it also has to, you have to admit to yourself, software is going to go to zero and document creation. a lot of the outputs that we kind of think about today. day, even if they're not there perfectly, they will go to zero. That's where this is headed. Was that not an obvious enough question, guys? That's just not like a clip. That's not like just like a video like rant clip. Yeah, I think I can, Jamie helped happy for the jump in, but the value, right? Like the very definition of law is that it's meant to be open and accessible. Like the statutes has to be publicly and free to really available to read. The value of law has never been about information. The value of a law has always been about abstraction and the insights and the value of the law. So it goes back to the point, the translation piece. Like anyone can look up the registrar, look up the corporations acts, the words I'm front of you for free, but I do get it. Do you understand it? Do you know how to comply with it? That's where the law comes into. Apply the skills they translate that for you to guide you through the process, right? And all the tools out there are there to assist the lawyer to do that for you. And it doesn't matter whether you're in the litigation context, in the transaction context or advisory context is fundamentally the gathering of these raw legal information into a service that actually not just gives the client the answer, but the solution, right? So that's where the value is at. And so I feel like what's great about like the legal profession is that that value will, it's naturally changing. It evolves with the times, right? It's just the way we do things are different from period to period, but fundamentally it's still kind of that, right? And maybe there will be methods that in a make that translation piece more easy and requires less human effort, but fundamentally it's still still that piece. But of course, it's often like very much subject to the bait. I can't do anything. Yeah. Yeah. To actually appoint Ray, I think only lawyers can reinvent what legal service would look like in the future. That's not a software engineer's job. You can ask AI to, you know, build you law, the next generation law firm, you know, it will come up with something that is, you know, that is still thing of the past. And I think that's where imagination and creativity becomes so important. And AI is fundamentally not creative enough. So I can, well, I can think of a lot of ways where you can basically give and impress the clients, you know, and press upon them, you know, what the future of legal service looks like. And just to give you an example, I'm a fund lawyer. And and for a lot of the investors out there, bake institutional investors, they may be the higher at bunch of law firms, it helped them review a lot of fund investments. They've gathered a thousand different sandmatters. But the GC probably not tracking all the rights that they've negotiated. Maybe there is some notification rights that they're entitled to. Maybe there are some audit rights that, you know, they, they can actually exercise. But then because it's just impossible to go through those sandmatters on a day-to-day basis. So what if there's a van, you know, the lawyer advising this client, I'll basically build them an AI that would just screen your sellers as daily to find out, well, on this day, you can exercise these 100 different kinds of rights and automatically send an notification to the GP asking for them to be respected. And that's just, it's an alpha, right? In general, intelligence that way. And only lawyers can think of the ways that can advance the client's interest. So then this is, and this is where, this is where the prediction that lawyers might soon charge $10,000 an hour is not so crazy. Right? In other words, because if what you create in an app, let's say in a typically in an hour, you create a document. Again, I'm not saying it takes an hour for lawyer, no lawyer creates any document in an hour. But in other words, take that same hour and now apply it to a proactive system that you're going to periodically check in on and update and iterate and create new versions there. Like now you start, that doesn't sound so crazy that someone might charge $10,000 an hour for that because the output of what you get is just on a different scale. I would go further to say that the reason why they can charge that amount is there's no comparable in the market. It's almost like an infinite zero sum game. Imagine if we have the kind of service and there's only that lawyer who knows how the technology works and promises that, you know, this actually works. And if there's no comparable, then of course, the supply is low the mind high, that pushes it rise up. But then the reason why I say it's infinite is because once that kind of service appears in a market, people will try to copy and then people will think of new ways. So actually, if we're asked to try it, it will go on forever. And so I think that to bring back to the point about being a lawyer in a future requires a deep understanding of what technology enables. And if we know what technology enables, then we can think, you know, what kind of new alpha we can pursue. And you know, by using a new alpha, you know, it brings me back to the concept of legal quants. The reason why we say legal quants, but not legal engineer, is that you can think of legal engineer as the person who owns this strategy, who kind of thinks, you know, where the alpha is, how to pursue it. A legal engineer is given a task, a plan to execute. So it's two different things, I believe. So legal quants needs to be a lot more creative. They're practicing lawyers, think of new ways to deliver legal service. And creating apps is just wonderful. So yeah. Right. You want to. Anyway, before before rate, Jumston, I just want to like, I want to pick up on one thing and double click on it. You also work with hedge fund clients. And you've sort of the reason I know this is I have a partner who's in hedge funds. And I recognize some of the thinking patterns. Like how much of your thinking and the way you've been thinking about this product is influenced by to certain extent, being like a hedge fund guy. Only partly to be honest, I think I it actually came from my interest in finance. I did a whole, you know, CFA program and a complete all three levels. And that really taught me a lot about, you know, how to what is what is beta, what is alpha. And how do you actually find something that's not correlated to market? I think that is something that's really interesting because if we think that if we think of legal service, we kind of always try to distinguish ourselves from other people, right? You know, we try to see how we can be better. You know, it's a company. You know, it's all about competition and law firms need to want to want one another, right? So in that game of differentiation, it comes, you know, the desire to build a mult that is desire to actually chase things that will impress the client. And I think that is the whole idea behind thinking how to use AI and see create new service and how to swap new markets, new demands. And I think on the hedge fund side, I think that there are some hedge funds that are really, you know, hook up to, you know, well, Melania is one of them. And well, the way they do it is basically, what about high flyer? Yeah, definitely. I, you know, you know, I mean, these things is good and so on. But, you know, I haven't worked with them. But my point being, the strategies that I would think, you know, for PM for a portfolio manager, they own a strategy. They need to be responsible for their own P&L. I think that's not so much different from the law firm partner, right? They also need their own book, you know, they also need to serve their own class of clients. They also need to think about their niche hits. And I, I, I predict that, you know, a lot of the really, you know, you know, the rain makers of the future will be ones that will harness this technology really, really well. And then find new ways to, you know, serve their clients and try to make a market based on that, you know, jump in there and then I've got a specific question I want to ask you and then I want to come back to this topic in the last part of this conversation where I'm going to ask you guys what other lawyers should be doing right now in addition to just coding. So hold that thought, but Ray, jump in on the on the on Jamie's rant and then I've got a specific question for you. Yeah, I'll just validate Jamie's point around the legal engineer versus the legal quant. I was lucky enough to be one of the early legal engineers. I was actually, when I was still studying law school, I had a, I worked as a legal engineer job in, this was 2017. So it's quite early before language models know that. And back then, the legal engineer was basically just a person who knows how to code, but doesn't have a computer science degree. So they can't be called a software engineer, but because they can code and you're in law legal engineer, that was like, that was the title, right? There was a way to get around industrial war to know of that. I drew to what Jamie said, it was, it was purely implementation, right? I just followed the instructions of the PM. I coded it and that's it. Legal quants, that's completely different, right? And you're, you actually, you have the idea and then you try to, whether it's AR implementing for you or others helping you implement, regardless of what it's, it's the idea, right? And I think fundamentally, since Jamie mentioned finance, I'm also just mentioned like economics, right? Fundamentally, our economics has been based on, you know, the four factors production, land, labor, capital enterprise. I think lawyers have been used and the whole way we value our services is really in the labor camp, right? X time equals X dollars, right? X effort equals X dollars. I think the fundamental economic philosophy shift that we're facing is how do we then bring ourselves into the enterprise mindset where your, your value is, coming from the fact that you combined other factors production together to create a service, right? And that's something that, from the solar practitioner to the boutique war firm to the big end of town, to even the legal tech vendors, really grappling with that question, right? It's a new, it's a new economic paradigm for the legal and the infrefoot profession. And I feel like that's the biggest shift. So what I've got on the screen right now, then, for those who are listening, you know, in audio and not on, on watch on YouTube. This is an essay from Paul Graham from 14 years ago in 2012, but he applies us to startups, but I think it is absolutely the ethos of what you just described in terms of this next group of what you're calling legal quants. And just to read it, he says, the way to get startup ideas is not to try to think of startup ideas. It's look for problems. Preferably problems you have yourself. The very best startup ideas tend to three things in common. There's something the founders themselves want that they themselves can build and that few others realize are worth doing. Microsoft, Apple, Yahoo, Google and Facebook all began this way. And again, I think that is truly like the ethos that you were describing, you know, with the legal quants. And I think that it's exciting because today, you know, there was always this question, you know, lawyers would tell me like, Hey, I want to like get into legal tech or want to get into AI startups or and there was always like, okay, well, how are you going to transition, how are you going to make that move? How are you going to become a legal engineer like Rayson? Right? But today, you don't have to, you don't have to leave law to do something like exciting and entrepreneurial and new. And to me, anytime that lawyers get to go back and scratch the entrepreneurial itch and you don't just have to be someone who is, you know, sort of only, you know, less risk appetite by design, overly conservative, overly safe. Not every person who graduates law school has those, those, those character traits. And I think this is great for lawyers who want to be able to express themselves entrepreneurially without necessarily giving up on the profession of law. Exactly. I feel not many, not all businesses need to be built around the tools and not every tool deserves the business to be built around them. And lawyers should extend they can build things, use technology. They just become very, very good generalist. And I think generalist will definitely win in the positive age. They just need that safe space to express themselves. They just need to have the infrastructure that supports them. And right now, I do think that, you know, I've been talking to a lot of like putting lawyers, a lot of legal ones. Their common frustration is that, well, they don't seem to have been able to find that space. And either in a traditional offer or for those who are more fort thinking, they would apply to for different startups. And then the offers that came back or the roles that they're being given are also not that exciting. Right. So, which basic question? So these talents, you know, they are extremely valuable and would be instrumental to the future of law. What do we do with them? I think this is the biggest mislocation of talent. This is like a talent problem of our industry. And you know, how to put these people in places where they can shine, have the greatest leverage and make the greatest impact. I think it's a problem that I personally want to solve. So, yeah. Ray, I wanted to ask you before. And I want to come back to this. I've felt for a long time that AI coming right after the COVID shutdowns towards the end of 2022 is not just a coincidence, but that very much AI took off more because it came post COVID. Now, the way I've put that is many people were working alone. So the idea that you could now have someone to brainstorm with when you didn't have, you know, John or Mary or Jamie down the hall was hugely valuable. And I think it solved the loneliness problem. I mean, I think AI takes off no matter what whenever it hits. But I think the kindle for that initial fire was the fact that we had sort of just kind of come out of the COVID phenomenon. And this is what came from it. And what I was listening to, you know, on a different podcast, it occurred to me, there's another way in which this is true because there were a whole bunch of people who acquired skills during the course of COVID that they were then able to put into effect right away when AI hit. And in some cases, a lot of those were either technical or digital. So can you talk a little bit about your, what you did like during the, you know, during the lockdowns and during during the, you know, the, the, the time of 20 to 22, which is ended up being quite valuable in the age of AI. Yeah. So thanks for setting me up for that. So like I was, my branding was first more known as the guy who built AI to analyze K-pop dancers and how it started was, as you say, COVID time locked down. I do dancers. I actually, I, I, I dance a lot. But so lockdown, no more classes. What I do at home. I've been coding since high school. I've been making my own games and web apps. I just wanted to find a project. So I thought, you know, let's just learn machine learning. I was inspired by AlphaGo, like how it defeat Lisa, like dang, this is so cool. I'll always want to build something like that. Are you, are you a goal player? No, I'm a chess player. I actually played with chess during high school. But anything around board games, I love. And so, and then, you know, it was during COVID and BTS also blew up a lot as well. So K-pop AI, there's these were the training topics. And then I built a prototype where I, it's a, it's a basic machine learning script where you upload a video of a dancer and you upload yourself dancing the same dance. Elden compare how similar you guys dance and produce screenshots. So where you are out of sync. And I posted on YouTube as like a vlog. People liked it. And then I applied it to professional dancers. K-pop. I was trying to rank K-pop groups and it went like viral across Asia, like South Korea, Japan, China went crazy over these videos. For the next two years, I created like video series based on that. And then like along the way I was trying to refine the, the, the program and eventually had a build mobile apps. And so this whole AI dance program was now in an app that people can try out. And then along the way you pick up these skills, I've also picked up in it. When you deepen AR, you naturally come across questions around ethics. So I think around that time, self driving cars will also be a thing. And there was a moral ethical question, which is like the trolley problem. And once again to air ethics that leads into a rubber half of a regulation around 2021, the EAR act started to become, you know, theorized and like I never thought there would be a such thing called AI more like that sound was sort of crazy, but it's reality, right? And then you know, that I've been following that interest part. Let me tell you, I have been, we have been discussing this internally quite a bit because there is one reading of the EUA. AI Act, which would suggest that any AI used by a lawyer could be an issue because if it's the way the EUA Act puts it, is it if it's something that would influence a judge. So depending on how aggressive you want to be with what it is that influences a judge, I mean, you could theoretically say that drafting a contract at some point is meant to influence a judge if that contract then becomes the subject of a dispute. I mean, you could take a very wide reading of that, but even more, I work, they've been public about it, so I don't mind like talking about it. We work with the American Arbitration Association that has built the first AI arbitrator for a very limited set of use cases. Right now, documents only real estate construction or construction disputes, right? So they're very, very specific segment of cases, but the EUA Act has a lot to say about AI in the context of a judicial decision or even a decision in dispute resolution. It is by, it is for sure like taken more seriously, there's more anxiety around it, there's more governance and disclosures required around doing it, but it's very real. This is not something we're talking about in abstract anymore. Like this is actually in the market, not from some rogue startup, from the biggest name in dispute resolution. So these things are at the EUA Act is certainly coming up. And of course, people care about it more today. So you've obviously experienced both the impacts of AI and online content. One of the questions I struggle with because I'm investing a lot these days in podcasts and in media. My business is really a consulting business, but I started out as a content creator, as originally writing columns on above the law about legal tech in 2015 before it was really early then. But I worked at the SBN radio back in the day. I have the skill and know how to build a media company right now. And in the one sense, it does feel to me like, okay, this is like also along with my thesis of AI that enables more human experiences. Building an online following, building a media company feels like the right play. And then I sort of like take a step back and I kind of go, okay, but like, if I'm looking around corners here, right, is am I am I building right now for like 2021 or am I or am I building for 2030? And there's some part of me that wonders like will media will be the entire like, you know, influencer communities, are we going to be the beneficiaries of are we going to be these targets of disruption by AI? And I have to admit, I have a little like I continue doing it, but I've got a little bit of analysis paralysis. I'd like to develop better conviction. Help me peek around corners for a minute and tell me like how you're thinking about this. And I want to ask Jamie to do the same because I feel like part of the advantage of being early to AI like you guys talked about is like also being better able to see around corners. Yeah, like Jamie and I talk about this a lot like also in the content space, like I actually realized that I've been watching Jamie's fitness videos for a long time, but I didn't realize it was him like, oh wow, so that was a big revelation. But I think in short, I have a particular philosophy for this question, which is the three C's theory, which stands for code, content and community. So content, so what you just said above, it's just that forwards to the content side of things, but to really make that real is you have to supplement it with code, which represents the tech and also community which represents the ecosystem. And I feel like if people keep talking about the one man business and the one man startup in the future, I think fundamentally these are the core ingredients for that. And you don't have to turn into a business. The course would be a personal branding thing. The really successive ones have those three C's. And I feel like these three C's become the new definition of capital. I think traditionally capital has always been about you have the money by like in this day and age where things are getting more democratized, more accessible. Money is not the only barrier anymore. And that you know, there are other ways to achieve capital and it's through those three C's. Maybe the fourth C is compute. Yeah, I put that into code. Yeah, you can now, can I see does he want? Yeah. I think, yeah, compute, you know, code is a artifact of compute being these days. To raise point, I think Ray and I really bonded over our shared history of being a content creator quite eight to nine years ago. I think I have been writing music blocks. I wrote music blocks on Facebook and as a decent following, I stopped and three years later I created a YouTube channel just to show kind of Spanish content on my fitness journey. That got almost two million views. Amazing. A lot back in back in the days. So in other words, you're not like this is not like new for you. In other words, like I saw like, you know, you hit on LinkedIn and there was like a part of me was like, I wonder what this experience has been like for him like being famous because like it kind of just seems like a like, you know, a kind of brainiac lawyer. But no, you've got you guys have like both like been through this before. This is not your, this makes so much sense to me. Of course, this is not your first time. I mean, that was unintentional. I did not try to, you know, I did not post those content with a view that I want to become an influencer. I was just trying to to share knowledge. But then, you know, which was the same goal bet that when I looked at the, that's the next channel because I want to give back to the community because I brought so much, you know, these kind of fitness video, transformation video, and then I should do the same. So, but then I then I actually read a race article on, you know, there, there's no benefits if the five coding lawyers just start posting their content on a LinkedIn and all of a sudden it becomes like a rat race. And that really struck me one night because I found myself trying to think of another app to build and I'm running out of ideas. So, yeah, gosh, that feels like, yeah, that feels like the story that like Jerry Seinfeld and Larry David tell about like getting the renewal of the season of Seinfeld. And it was like, I think 22 episodes and Larry David says that he like went home, got in his bed and just began crying profusely because the idea like, how am I possibly going to create 22 more episodes like creating four was impossible, creating 13. I have no idea how we did it. Like, I can't do this. And there is like, you know, some, some, I think there's anxiety. I feel that daily, you know, because, you know, one of the things that we're doing is, as a company, Killer Well strategies is providing like the right, you know, intel and information to our clients so that they can make, you know, the right decisions. So luckily, like, AI moves, AI moves really fast. But it's not just the news. It's like, you know, coming up with like, here's what the strategy should be. And like, like having to constantly come up with new things, I think is an anxiety-ridden process. But I also do think that ultimately it is what we're headed towards in a world of AI where you don't necessarily get to rest on your laurels for as long as you thought. So we might see, you know, software companies shipping updates daily, even hourly. You might, you know, see, you know, content creators having to come up with new ideas more regularly because it's just easier to do. And so the rest, I think this is some of what people are like nervous about. What do you, when you think, like, obviously everyone should learn AI. That's, that's easy enough. But what do you guys, and I'll start with Ray and then go to Jamie? What should people be learning right now? And is there a version of, because I've been trying to teach myself a lot of like the theoreticals behind coast, like having chat GPT, like work with me on like my understanding of algorithms and, and, you know, a lot of the math involved thinking through like general concepts. But I, what would you recommend to lawyers to begin learning right now that, hey, listen, you should learn this because it's either very valuable or you should learn this because you thought that in order to learn this, this would take you four years actually today, it could take you three months. Like what would you, what would you suggest and how people go about that? Yeah. So my thinking is that learning theory in this day and age, it's like learning, it's like memorizing dictionary, right? You can learn all the words, but if you don't apply them, it's no point. I feel like applications, the most important thing. And I think the good thing about lawyers is that lawyers are sort of trained to learn anything on the spot. It can be a topic completely unfamiliar to them. But if they have to learn it, they'll learn it. I feel like that's the mindset and training that lawyers receive is really optimized for just absorption of knowledge. And that combination. By the way, this was the greatest argument for lawyers using AI initially. Not to draft your documents, but like, hey, anytime you have to learn something new, this is your tutor. That's right. Exactly, exactly. And I feel like my one advice is that, think about one thing that's really annoying in your work. Like just one pain point that is annoying enough to really cause Mickey angry, but also not annoying enough that other vendors have approached it, because it's not scalable, not commercial. That's the sweet spot, man. Like those really personal custom problems that you really dread and it makes you a job boring. Focus on that, right? And no one understands the problem better than you. And if you know the dynamics, the problem, the inputs and outputs that are needed, you'll naturally be able to research. And like lawyers are the ultimate researchers, right? They know how to research. And so just research, test things out, talk to people, and naturally you solve it. And I feel like a lot of the legal quants basically start off like that. Like just do it. I think it's all about execution, to be honest. If that's really key in this day and age execution, just an iteration and execution. Yeah, so I think this is action leads to information. That's right. Ray, you still have all my points. But I really want to double down on execution piece. The reason for that is for the legal quants that I've met for the past two months, all of them have extremely high agency. By agency, I mean, being able to take an idea, and execute on it, and then until the thing is perfect or near perfect. And I think a lot of people like to talk about AI, but not many people want to actually do things with it. And I think for a lawyer, to get into this, you kind of need to have that kind of motivation and that kind of grit to tolerate the bucketness of whatever AI tool that they are using. And the reason for that is it's almost like driving a dangerous car again using that terminology. You could hit the pedal to hard and then crash, but then do you want to back up and then try again? What's usually the one to say, "Calkway, well, maybe this is not for me." So I think that kind of high agency quality will be extremely important. And we do see that people that have high agency are usually better at an AI. So how to cultivate that agency is a, I think it's something that it's not easy to really have a concept for, but yeah. - So in other words, 2026 is not gonna be the year of AI agents. 2026 is gonna be the year of AI agency. - I like that. - All right. Yeah, I guess I'm one of those people that likes to talk about the AI. But I think this is spot on. I think that agency really is the right word here. And I think agency needs to triumph over anxiety. And I would tell those who have any anxiety, first of all, follow Jamie and Ray. You guys have a couple of ways that people can follow you. There's you, the LinkedIn group. You can follow them both on LinkedIn. You can also, how can people stay in touch with what's going on with legalquence right now? - So they can follow our, we're gonna launch our sub-tech pretty soon. So that's where we will maintain context. The community right now, it's right now, we are invitation only. So we wanna have Laurie's who basically, that we think are like mine to join us. But then we'll be, we also have our LinkedIn page. We just create a new one. - You guys learned this VIP concept from Harvey. I'll tell you, Jamie, I sent you a note. I tried to connect it to you on LinkedIn. And then, so I'm used to like, and this just shows how like permissionless this entire new world is. There was a time if I would like connect with a legal tech entrepreneur, I would get an immediate connection 'cause like I have invested in startups and a lot of times companies are looking or they know that I'm connected and can introduce them to this one and that one. I sent it a note to you, like it just sat there. I would like look at it every day. I'm like, pending. I'm like, come on, Jamie, connect with me. Please, but I have to tell you, I have to tell you. It made me even more interested. I'm like, this guy doesn't even need me. This guy does not at all. So I have to tell you something. Like I think the invitation only concept in developing a core community and being very deliberate with that is super smart. I obviously encourage people to stay in touch with you, but then I wanted to come back to this agency and anxiety point and on this note, Jeff Bezos has a thing about anxiety that I love. He says, he's used to suffer from terrible anxiety until he realized and reframed that anxiety was a little alarm that his body was sending him to let him know that there's something that he's supposed to be doing that he hasn't started doing yet. And he said, the minute I started, the minute I started, the anxiety would go away. And he said, now I don't look at anxiety as something bad. I just understand, oh, there's something I need to do. What is it the thing I need to do? Oh, that's what I need to do. Boom, done. Exactly. Anxiety gone. And I think that for folks who are feeling anxious, what the surveys tend to show over and over again, is that the people with the most negative sentiment towards AI are those who have not used it. So if you want to get rid of that anxiety, lean in, just go. I'm not telling you to productionize it for a client yet, but just kind of like get going and get learning. The sooner you let your brain wrap around this, the better it will be. Guys, I never go for like an hour and 15 with guests, but I recommend ever listen to like all parts of this podcast. This is just fabulous. Thanks so much for both of you for joining. And I look forward to following what you guys are doing. Thank you so much, Zach. Thank you, Zach. Appreciate it. Thank you for tuning in to Zach Abramowitz's "Legally Disrupted." If you enjoyed today's episode, please consider leaving us or rating in a review on your favorite podcast platform. Don't forget to subscribe, see you in never miss an episode. (singing)