Episode Transcript
[00:00:00] Speaker A: Welcome to the Relay, the legal show for personal injury law firm owners, presented by Lexamica, the number one attorney referral network. I'm your host, Gabriel Steeritz. If you've been anywhere near legal tech this week, you know the News Anthropic launched Claude for legal. 20 connectors that plug Claude straight into the software law firms are already running on, plus a stack of practice area specific skills. And the whole industry is talking about it for good reason. But most of that conversation is around big law. And I want to cut through that because underneath the noise is news that's making a very quiet point, which is that a connector is the thing that lets an AI agent reach into a system and do something inside of it. The headline isn't that AI got better at legal. The headline is that the agent layer is being built into Claude natively, meeting legal teams where they are and meeting everyone else inside of your chatbot. And for the last 15 years, there's been nothing like this, and I'm really excited today. My guest, Tyler Stillwell, founder of banejack Solutions, is one of of the best people in the industry to talk about this. Tyler has been migrating law firms, moving them into the cloud for the past decade, and he has seen more of this change upfront and personal than anyone else in the industry. Tyler, it's great to have you here. I'm excited to talk about what this means and what the heck is going on. But first of all, are you calling me from the woods?
[00:01:20] Speaker B: Dude, I am. I took a break from my very luxurious walk in the park to be on this show. So thanks for having me and hopefully my cellular connection last, last for the duration of this podcast. But yeah, man, I appreciate you having me on here.
[00:01:35] Speaker A: Absolutely. And beard. Recognized beard. If anything, if you don't have an answer, just stroke the beard. We'll all think that you're smart.
[00:01:41] Speaker B: Yeah. Or, or at least you'll think that I'm thinking, which is great.
[00:01:44] Speaker A: A hundred percent. And who, who actually knows.
[00:01:47] Speaker B: Right.
[00:01:47] Speaker A: This anthropic thing is really, really interesting to me. I. We use Claude at Lexamica every day. We've been using MCPS internally to connect to our tools before anything else. First of all, what is this for the people who are in our industry, they may not know what an MCP is or what a cloud skill is. Why is, what is this and why is it such a big deal that Anthropic released it?
[00:02:11] Speaker B: Yeah, I mean, golly, man, how many hours do we have to talk about this? I think that, you know, to Divert your question a little bit. What everybody's talking about is how do we deliver better outcomes, right? Law firms are talking about how do we deliver better outcomes for our clients. And then on, on our end on the professional services industry, we're trying to figure out how do we deliver better outcomes for our clients. Who are the, the law firms themselves. And I think at the end of the day, all of these new tools that are rolling out are tools that are trying to find ways to help people deliver better outcomes for the people that they serve. If you've ever tried to talk to your parents about AI, you know, at the end of the day I tried to distill it down and say it's just they're just new tools, right? They are new tools in our tool belt that we're all trying to figure out how do we use them to get better. And so when you ask what the heck are these, what are they about? And I wish that I had a dissertation that I could share that says here are how to implement these tools and here are the concrete ways to deliver better outcomes for the organizations that you're trying to serve. But we don't really know. And I think that that is maybe a part of the dirty little secret is we all have ideas and we all have a hypothesis or a hypothesis.
[00:03:21] Speaker A: And let me, and let me just step back real quick. Tyler, you're already talking as someone who's way in the weeds here. For some of our audience, an MCP is a way that your chatbot, OpenAI, Claude, whatever you want to call it, it's a way for it to talk directly to the tool that you're using. Whether that is a Google search or that could be your Clio or that could be your Supio or your even up or your Eve. Whatever the tool is, it's a way for them to interface directly with other. Right. And that's what the big release was about, was Claude can now talk to these tools in legal tech that were not publicly available before. And it's like a one click setup which means that it's really easy to enable that. That's what this release was about, is now Claude is talking directly to legal tools. What you're saying is all very true, but I want to make sure that for our, we're not leaving our audience behind and talking about it. So that's what this release was, which was your, your trillion dollar company Anthropic is, is now actually building in the legal tech space. That's why this is such a big deal. Before it was all generic tooling and now it's anthropic. It's like the eye of Sauron turning to Frodo and saying, I'm going to build tooling over here. And it's a really, it's really a big deal.
[00:04:35] Speaker B: Yeah, it's, for lack of a better word, it's, it's disruption. Right. And it's, and I mean that really in a positive way, as I was talking about earlier, it is making it easier for these organizations to disrupt in a positive way the service offering for their clients.
I think this is going to be a game changer. And what it's going to allow organizations to do, including ourselves, including like Samica, including the law firms, is to iterate faster. And we take the, the process of idea to execution and we're making it so much faster because we don't have to integrate all of these other various tools. We're able to work directly or interact directly with these systems, which means that we can see outcomes faster and which also means that we can throw ideas away faster. We're much sooner going to be able to figure out are the investments that we're making, are they paying off or not?
[00:05:21] Speaker A: And let me ask you this because you're, you're in an interesting position because you're not selling software directly to law firms. You're doing implementation, you're doing migrations. There's two parts to this. One is, are you seeing your customers adopt the generic AI tools into their law firms? Are they asking you to do that and help them do that? Part A and part B is does this announcement move the needle in that direction? Regardless of where everyone is right now? What, what are you seeing?
[00:05:47] Speaker B: Yeah. So two questions are one, am I seeing it? And two, do I think this announcement is going to change that one? Yes, we're seeing it. We took a little bit different of an approach as an organization whenever this, this AI craze started. And everyone was racing to figure it out. And we've been monitoring it, we've been working with it slowly internally, but we wanted to take some time to flesh it out. And I think that is, that decision is paying off well for us right now. I think we chased less rabbit holes, if you will. Sorry, I'm from Texas for all of our listeners, so forgive my backwoods analogies, but I think we're seeing the same things with law firms right now. Everyone is excited about it and everyone is trying to find ways. And so I've seen every law firm is playing with it. Every single law firm. I don't want to speak that flippantly, but almost every client that we interact with is experimenting with these various tools right now. And they have various degrees of it integrated into their platforms or into their operating processes as an organization. But everyone's talking about it, not everyone's doing it yet. I think this moves the needle because I think it makes it easier and we.
Not lower the bar, but it's. It's easier to bring these ideas to light. I keep coming back to that because I think that's what the real power is.
[00:06:59] Speaker A: My. My perception is a lot of law firms are using Claude and OpenAI over here on the one hand to do work, and then they have their more specialized legal tools like their case management system or their agentic workflows or the demand writing. Those things largely, I don't think are touching each other. Part of what's interesting about this release is this is the first time we've really seen that connection between those tools. And correct me if I'm wrong, Tyler, because I don't. I tend to see things a little bit further out and you're way deep inside of law firm workflows. But I don't hear people saying, I'm using my chat over here to, like, connect into my litify. It's like there's still an element of, hey, these things don't talk to each other directly. And what's interesting about this release is it's. It is a movement toward, like, I don't see it necessarily as being a replacement of the tools so much as, hey, the. For the first time, these tools will talk directly with each other. Do you think that's true, or. And how are you seeing law firms think about, do these tools work together? Do they view them as zero sum? They're replacing one with another?
[00:08:03] Speaker B: Yeah, great. Great question. So, you know, as we're recording this, this announcement is, I don't know how many hours old.
[00:08:10] Speaker A: It's like 24 hours. This is the day after they released it.
[00:08:13] Speaker B: Yeah. So it's pretty recent. So in 72 hours, I might have a different answer after I review some client calls from this week. But if you've done any level of experimentation with any AI platform, you've quickly learned that what you get out of it is only as good as what you put into it. And so we are all desperately trying to figure out, how do we get more out of our tools? And I keep coming back to this idea that it's only as good as how we prompt it. That's not Revolutionary. I'm not the first person person to say that. I think people have been saying that and I think that's what we're all experiencing. The beauty of this happening is now you get to in one place, when they're all connected, these tools are going to have a more widespread understanding and knowledge of what's going on in all of these different pockets, which makes the tool more effective, assuming that you can prompt it well. So it's not, you know, it's not like all of a sudden your prompts are going to get way easier to write, but they're going to be more effective. You're going to get better information because now you're dialed in and connected with the various systems that, that you're using inside of your organization. So, you know, if we were to record this three weeks from now, I would probably have a lot more information about how it's actually being done and what the payoff is. Because the short answer to that is I don't, I don't know yet. I think it's going to be powerful. I think it's going to enable a lot, but who knows?
[00:09:31] Speaker A: Yeah, I. Great answer because we actually don't know because this is all new. Two or three months ago, we hooked up all of our internal tools via MCP to Claude and we pushed Claude out through our entire company. And we didn't replace the tools that are more specialized like we didn't replace HubSpot, we haven't replaced Postgres for our data analytics, we haven't replaced Slack or Gmail or whatever. But what an MCP allows is for you to pull all of that stuff out into your cloud chat, do work there, push it back in and do all of that extraordinarily seamlessly. And so I think that for plaintiff lawyers, like that's the opportunity that we're seeing here is the ability to do that. And I think what'll be interesting to see is how much having that native connected experience replaces the need for like the AI co pilots that are more built into the legal software. And I could almost see a bit of a, like what the core value prop is for the legal AI tooling that exists as standalone right now is going to be, I think, more about creating good context, creating good citations, creating the right frameworks. Having a copilot isn't necessarily that much of a better experience than having Claude run an MCP that accesses all of your cases, all of your settlement values, all of your client information. But you do need a database and you need a way to structure all of that so the Claude can do work really, really well here and then push it back over there. And I think. And again, like you could take all this and say it's just like a PR stunt by Anthropic, but that's not a very interesting storyline. Let's assume for the sake of our conversation that it actually means something that like this is a shift that we're going to actually see technologically. I think that's where that would be is I think that the legal techs spend less time building the copilot experience and they spend more time, like you said, figuring out how to serve data up to AI so that those MCPs are highly effective so that people are able to do that work more efficiently and effectively.
[00:11:34] Speaker B: Yes, yeah. And I think that I could take this a bunch of different directions, but if you follow the Salesforce landscape at all, you know, four, six weeks ago, whatever it was, Benioff announced this, this headless model which basically layman's understanding is that you don't have to log into the UI anymore in order to use Salesforce. And so if you've been involved in any industry for the last 20 years, you know, everyone's been trying to consolidate platforms. And then what I've seen happen with this AI boom is there've been a bunch of specialty tools that have been buil on the backbone of AI. And so our industry, the one that you and I are both in, has understandably so now picked up more tools. So we spent decades trying to consolidate Right. In this idea of let's put everything together and have fewer places to interact with our data. Kind of got blown up, if you will, by the design and the rolling out of all of these specific tools. And then that also creates a different challenge for all of the various products in the space because now they're left. Prior to this announcement and prior to where both you and I think share envision where this is going, they were left to try to solve problem after problem after problem and expand and expand and expand, which expanding isn't a bad thing. But it's challenging to create the gold standard solution for multiple different problem sets. Right, that's, that's not rocket science. It's hard to do a lot of things really well. But what this will allow us to do is now say, great, I might have 10 different specific. I might have some homegrown agents, I might have some specific tools that I'm using, but now I can have one place to do some of my interactions with and consolidate all of that information versus having to interact with 10 different tools. And I think.
[00:13:12] Speaker A: And dude, the. What you said about Salesforce being headless, like I read an article about legal tech becoming more invisible, more in the background that I feel like that is the kernel that's going to grow into something very, very large. Like the idea that Marc Benioff, one of the most successful tech entrepreneurs of all time, people lived in Salesforce for the past 25 years. Man, like that user interface is burned into my brain from the day I started in the professional workforce to say that now you use Salesforce headlessly on the one hand, it's like, yeah, but that starts to feel natural because we spend so much time in chat interfaces. But does it mean that Salesforce is less valuable? No, because all of your information lives there and it structures it well. And when we built our MCP is one of the first things that we did that's been the most powerful across our company is we hooked it into our postgres database, relational database, exactly what Salesforce is at its core. And the ability to use a chat bot to write the queries, the reports, to go and control our database for the exact right piece of information at the right time and go synthesize it. That's what I want my Salesforce experience to be like. That's what I want my case management experience to be like, is ask the question, it goes across the tools, it gives me the report right here and bam, I have full context.
And I think that you pulling that out is exactly what we're going to see. Probably be the most important thing is the ability for your Claude or your OpenAI to go grab context across all of your different tools. Because the reality is, Tyler, like everyone who's using alidify or a filevine or even UP or any of these tools, like, you still have email, right? You still have either teams or Slack, you still have your Google Calendar or your Exchange Calendar, and you have your case management system and you want to be able to pull across all of those data sources and either you have one chat interface that's trying to go integrate with all of those, or everyone starts to slot in and build an mcp. That's highly effective.
[00:15:09] Speaker B: Yeah. And I think to mention, or to call back to what I was saying earlier, I think without this, I think in a lot of situations, the clients and I mean the law firms, they actually lose if every product has to become everything. I think that's a. A large challenge that, you know, is really hard to. To do. Well. I Can't sum it up any more simple than that.
[00:15:32] Speaker A: Well, well, just drill down into that for me, Tyler, because you've seen the transition from needles like old school on prem software to the cloud based systems now to the AI systems, to the agentic AI systems. What are you seeing as far as the number of tools being used? And look, I am not of the. I get that we're being more and more productive, but like, there's probably like a sweet spot that you've seen of like the bundling and unbundling of tooling. And like where, like where do you think that we should be as we move into the AI era with the universality? Like, is it one tool to rule them all or is it 12 tools that are connected in certain ways? Like how. How do you see this shaping up over the next.
[00:16:13] Speaker B: Yeah, I don't think it's one tool to rule them all, but I do think it is the ultimate goal is one interface to rule them all. And I think to, to your point with this topic, that's why this has the power to bring so much value, is because if you think back to what I was talking about earlier, we spent, you know, decades trying to figure out how do I stop logging into 30 systems, right? Well, now you can log into whatever your various case management system is and they're all connected, they're all integrated. But now we have enough tools, we have enough specialized tools that you can't even fit them all on your window, right? Your case manager, your paralegal, your attorney. If they're trying to interact with seven, eight different unique tools, where do you even begin, right? So imagine popping up a window in your case management system and you have eight different integrations, right? That's a pain. How do you train? How do you teach people which one to interact with it? And so I'll go back to your question. I don't think it's one system. I do think it's one interface.
And you know, admittedly, when. When I saw the Benioff announcement, I sat and thought about it for, for quite some time. I was just thinking like, you know, what. What is going on? Because like you, the idea and the vision of Salesforce, Salesforce, Salesforce is just imprinted in my brain. And so, yeah, to think of a world that is not that I'm not interacting with that really. I sat back in my chair and thought, what is it? What does this even mean? What the heck is this even about? But as I said on it, I realized the same things we've been talking about in this conversation, that, that's really, that's really where I think we're headed. You, you, you had a great line earlier, something to the effect of like how does case management or how do these systems become invisible? And I don't remember exactly what it was, but it was good. And I think that that's, that's really true. How do we get to this? Everyone's been searching for this uni unified system and I don't think the answer is one system to rule them all. I think it's one interface and that's a distinction that we could argue about what that actually means.
[00:18:06] Speaker A: And the interface is going to largely be chat driven, I think. Like, and, and I had a conversation earlier today too and let's just go way off topic here and talk about deterministic and non deterministic systems to go like way nerd zone here. AI is non deterministic. It's not going to give you the same answer twice. It's going to always give you something slightly different. And that's great if you want to have like a great chat interaction. But deterministic systems where it's A follows B follows C follows D every single time is a necessary part of running a law firm as well. And so even with an mcp and we're starting to see more of this with like scheduled tasks with Claude, but then they still are just running a prompt. You, you can't run a law firm like that, right? Like, you just can't do that. And so there are necessarily just going to be different types of tools as well. But then what do we want to interact with? And if I'm interacting with a client, I want to have full context and like that's a really, really powerful thing to have. Or if I'm going to draft something, I want to be able to like pull from all the knowledge sources that I know exist into one place. That's you know, case law, citations, former client interactions, all the documents. Like that's like you said in one interface that can wrap all of that and then do work there and then push it back out to where it needs to go. That's the part where I think there's a lot of possibility that hasn't yet come to bear across even the best in class tools in legal tech. And some of them are being built and are and are doing a lot of that like right now. But I also think that for law firms that are more like, hey, there's always the question of build versus buy, like specific tooling And I think that the build is becoming interesting as these MCPs are being released. And I'm curious to see, like, from, from your perspective, Tyler, do you think that over the next six, nine months you're going to be seeing more requests from law firms to build using generic tools that are really powerful, or do you think that we're going to see, hey, we're still in the legal tech specific space and the core systems of record are going to be legal tech specific.
[00:20:09] Speaker B: Yeah. So if you're asking me, build versus buy, we've been in a cycle for the last 24, 36 months where the answer has been buy. People have bought and bought some really phenomenal tools. But have you been spying on our internal meetings?
This, this thesis that we're headed back into a cycle of build is, is one that we have at Banjax. Now, I'm not saying that the existing AI tools and platforms are going away. I don't think that that's the case. But you used to have to go search to find someone who was doing 80% of what you wanted. And the last 20% was not really worth trying to build your own thing over.
Um, and we're starting at the smaller scale. So I'm not saying that you're going to go build an entire engine to replace your AI demand tool today. I don't think that we're going to see that be widespread and rampant. But these smaller tasks, these smaller ideas that each organization has, that makes them their own, those things are going to be worth tackling. And so I think we're going to see more building, but I don't think the buying is going to stop either. Does that make sense?
[00:21:16] Speaker A: 100%. And the reason you can build and buy is because the tools are so much more powerful than they were. If you can, yes, someone's caseload, well, you can afford to build it and you can afford to buy it because you both are going to provide an outsized return. And we have an internal word for those little tooly bits that like, you haven't bought it and you're building it. And it's something that you can use and reuse to get things done, like an artifact or a prompt or a project in Claude or whatever it might be a Google Chrome plugin that one of our devs created. We call them jigs, like from, from woodworking. It's like you need to make the same cut ten times. You screw two pieces of wood together, it's good enough to go and like make the same cut over and over. But eventually you might throw it away. You might take it apart and build something else. And that kind of idea of like, is it a tool or is it a piece of wood that you kind of sl. It's both.
[00:22:04] Speaker B: Right?
[00:22:04] Speaker A: There's this kind of, like a tritable tooling product concept that has never existed before that I think is really powerful. I mean, y' all are in an interesting position as kind of internal builders for law firms to create these jigs where you're like, is it product? Yes. But also, like, don't expect anyone to support this, because a week from now, there'll be something better. So you shouldn't spend much time on it, but it's still going to make your life better. And these jigs have become this core part of what we do. Like, we build this thing and it's like, hey, we're going to build 10 things. Two of them will exist in six months. They'll be wildly phenomenal. It won't exist anymore, and we won't care because we spent three hours on each one. And so you're just kind of like, your. The level of prolificness, it's very cool. I think we're in a golden area of tool creation right now that could continue to get better and better, or it can harden in some way as the tools just become so sophisticated. And we'll look back and be like, you know, we're. Everyone was talking about prompt engineers as, like, the role of the future 24 months ago. Like, how many prompt engineers exist today? Not like, nobody hires for prompt engineers. This may also fade away, or it could become something bigger. But the one thing that I want to come back to is you did say prompt was a big skill, something that was buried in the blog post or actually wasn't in. The blog post was buried in the GitHub repo that Anthropic released alongside the legal connectors was 12 tools that Anthropic built specifically for personal injury law firms. So demand letter writing, medical chronologies, initial intake, evaluation. I thought that was very interesting. And Skills are just very, very fancy prompts to go make agents in your clutch at do stuff.
I don't know, and I would love to hear someone tell me, because I don't. I'm not going to go in and play with all of them. Is that a PR stunt, or is that genuinely another kind of jig that's going to let law firms just plug their clot in and start running a law firm in? Because there's. Claude Skills now to write, demand Letters. Yeah, take.
[00:23:57] Speaker B: Yeah, my hot take. And I have not played with all of them either. I haven't hands on played with any of them at this point. But my hot take is that I think it is. It's not totally a PR stunt, but I do think it is a way to entice organizations to start experimenting with Claude more than they would have previously. Because now let's say that this release gets a law firm 40% of the way there. Well, that's 40% of the work that's removed. And sure, getting to the 60%, the rest of the 60% on your own might not be feasible right now. I don't know the answer to that question. But they're at least getting more people to play with the system and they're at least getting more people to envision a world where why couldn't I use cloud for this? And I'm not suggesting if any of our clients are listening to this. I'm not suggesting that you go start doing that.
[00:24:46] Speaker A: Right.
[00:24:48] Speaker B: I think that in this hype cycle, there have been many organizations that have bought AI tools in a hurry and not established what they're actually trying to get out of buying this system. Like, what are the objective results that I expect to see as a return for me buying this and if you don't have that and if you don't implement it well, which we could talk for two hours about why the, that process matters and why it's so important. What are you buying and why are you buying it? I know I took that a little bit different direction than in your question, but I think it's part real, part hype, and I think it's a way to get a way to, to get law firms to start interacting more with. With anthropic.
[00:25:29] Speaker A: Yeah, the metaphor that is been really grounding for me as we've looked at AI systems and what ultimately becomes human in the loop or what's AI in the loop and what's copilot is as they get more and more powerful, 40%, 80%, 99%, the reliability of some of the tasks or many of the tasks inside of law firms are like the brakes on your car. Which is to say that even if they work 99 times out of a hundred or 999 times out of a thousand, a failure is catastrophic by nature. And it is true in law that a lot of tasks are like that. Some are not to be clear, but there's, there are plenty that are like that. And that creates a very, it's a very different use case than if I am like selling flowers or running a coffee shop, where the worst that happens and it's not really true. Like, the worst happens, I sell a few extra coffees or something like that, or I misplaced an order. Like with law, it's like you lose your bar license, you lose a hundred thousand dollars settlement, you miss a medical record. Like, the bar is extremely high and the margin for error very low. Do people make mistakes? Yes, a hundred percent they make mistakes, and that's not the point. But it's like these systems, you can't just say, oh, cool, well, Claude released a bunch of skills, so now we're all going to write demand letters. With Claude, it's much more difficult. And I do think because of that, there's a level of value to the systems built on top of those systems for the guardrails, the citations, that will have a lot more durability than other industries where, like, at this point, it's like you really don't need anything except Claude to run your nail salon or to run a library. Like, you literally could just use Claude because it'll do all of that work. It's not going to be true of like doctors and lawyers and other professions maybe ever, because there's really specific stuff to engineer around, man.
[00:27:18] Speaker B: Totally. And the stakes are too high in the legal landscape. Using your break analogy. And for many of the organizations or for many of the clients who are working with these law firms, this is quite literally the genesis of their case is one of the worst events in their life. It's traumatic. It's, you know, it's terrible. Right? What, what, what they're dealing with. And you can't afford to get that wrong, both from a business standpoint, but also as a human. Nobody wants to have, because of some tool that they used, missed a statute of limitations.
[00:27:50] Speaker A: Right?
[00:27:50] Speaker B: Like how beyond the business impact, how like, as humans, we don't want to allow that to happen. And so I mentioned it earlier, but we were relatively slow to start. Even though we're not a law firm, we interact with a bunch of law firms. And so by proxy, we have filled many of the same level of responsibility for this. And I know it's not quite the same, but we take it personal. And so we've been slow to adopt it because we can't afford to get it wrong.
And I encourage people to be slow to adopt, quick to shop, quick to evaluate, but slow to adopt and start small before you get big. And that's not the attractive answer. But as we're all realizing the AI boom has been going on for quite some time now. If you're not an entirely AI native organization, you're still alive.
[00:28:36] Speaker A: Right.
[00:28:36] Speaker B: You know, 36 months ago, the talk was if you're not fully AI native by 2025, you're dead. And that makes for a great headline. But it's not reality. And I, I believe.
[00:28:46] Speaker A: Well, and in, in an injury law and contingency practice, like, there's actually a lot of interesting economic protections because even if you're a billable hour attorney, like, efficiency is destructive to enterprise value. Efficiency is constructive to enterprise value as long as the take rates remain the same on your, your fees. Right. So there's like actually some larger protections structurally for businesses in plaintiff law that don't exist elsewhere. So sidebar that. But it's good news for injury lawyers.
[00:29:15] Speaker B: Yeah, no, it.
So what a wild world we're living in, Gabe. You know, and we could record this episode 72 hours from now and we might have totally different takes.
[00:29:24] Speaker A: I mean, I guarantee you that OpenAI is going to do something similar next week. And so maybe the main takeaway here is like, don't switch to Claude, because I guarantee you OpenAI is doing the same thing tomorrow because they're just going back and forth on all of this.
[00:29:38] Speaker B: But yeah, and I think that ultimately, regardless of what tool that you use, and I think it might be different from tool to tool or which model you use, I believe that, and I'm not suggesting that this is a novel take, but I believe that the organizations that are really going to win a decade from now are going to be the organizations who are able to precisely put the human in the loop at the right time. And so if you view it as whatever you're working on, let's talk about the legal system. If you're looking at a, you know, there's all these various points where you could get in trouble if a tool misses something. And so the organizations that are accurately and precisely able to predict when a human needs to be involved, I think that's where some of the magic is. And there's some no brainers, right? There's.
I think someone tried it actually in a different state or maybe that was an Onion article. But no, no robot is going to trial right now. Like that's a, that's a no brainer.
[00:30:28] Speaker A: But, well, do not, do not pay tried to send someone to trial with a chatbot in their ear like 18 months ago. And it became headlines in a, in a negative way.
But to, to some extent also Though it's a little bit like self driving cars. So there's the brake analogy on the one side, but on the other it's self driving cars. And at some point AI gets better than people at the job. And so at what point does it become ethically obligatory to use AI if not for an oversight layer or if not for the entire case, at least for an oversight layer? And I think we're well past that point. Like, if you're not using AI, you're definitely doing a worse job for your clients in the same way that if you don't have a car that can stop before it wrecks, like, you're more likely to kill people. And so are we running fully agentic AI? No, but we are absolutely in a world where you have to be. And I think too that the most successful law firms in five years are not the ones that can see for this down the road, but they're the ones that are trying the most things while being anchored to what you said, which is their ethical obligations and to doing right by their client. If you're anchored to those things, everything else should be changing very, very rapidly around that North Star. But like, yes, you're going to, you should be adopting new tools every month, you should be trying new tools every fricking week. And you should. The only thing you're not compromising on is are we iterating toward better outcomes for our clients? And that, I think that is what is going to lead to the most successful. Because everything else is up in the air right now, right?
[00:31:58] Speaker B: Yeah, if that, what you just said reminds me of the book. And if you haven't read it, game you should. The Infinite Game by Simon Sinek. And to your point, if your compass is set right, the various tools, there's hundreds of different options out there. But if you iterate and work on it and move through different ideas rapidly, as long as your compass is set properly, you're, you're going to get better as an organization. And I think having your compass set correctly is, is going to have a big, it's going to be a big part of determining how successful we are. Like Ocker, business is still business and you still have to deliver great outcomes. Now we have, we have sharper knives right in our tool belt. But it's still about delivering great legal work and great outcomes for your clients in the legal industry. None of these tools are going to change that. Nor should they.
[00:32:44] Speaker A: They don't change the North Star. No, they change how you do it. They change what you do. But they'll never change why we're doing what we're doing.
It is. It is exciting. And I would say as a final takeaway, if you use Claude, go try out the skills, see if they're any good. See how close they are. They're certainly not bad. They certainly do something. And at minimum, you'll learn something about how to run your law firm better and what AI can do. Right now, skills are, I would say, one of them. The less used Claude features outside of there's connectors and chat and projects, but skills still relatively underused. Tyler, thank you so much for taking time to talk today from your beautiful walk in the woods and the. The cell coverage held up. So it's been great and I'm sure we'll talk again soon in the midst of all of the changes that are happening. But I really appreciate you being on the show.
[00:33:35] Speaker B: Yeah, of course. Gabe, thanks for having me, man. Always great to talk to you and get your perspective and looking forward to our next conversation.
